My favorite quote from an old IBM mainframe computer service manual: "Never attempt to repair or replace equipment which is working to the customer's satisfaction." Or in colloquial terms: If it ain't "broke" don't try to "fix" it.
This reminds me of the robot that created the celebrated light painting in Asimov's collection of robot stories. The maker performed an unauthorised service on the robot and the wanted malfunction was removed.
I wonder if people insisting on wholesale replacement have ever stopped to really think. What is behind the fact that they replace a 10 cubic inch capacitor with a “modern” 1 cubic inch version? In comparison with resistors we can conclude that a smaller package does not dissipate as much heat as the large one does at the same temperatures. Or, if the same heat generated has to be dissipated from the small package, its temperature is higher. And higher temperatures are what is a main detriment of the component life time. Just check any document that refers to the Activation Energy. Why are the new capacitors so much smaller than the really old ones? Some time after WW2 the manufacturers started increasing the aluminum foil surface area by etching (roughening) the surface and reducing all the materials, increasing the per unit stresses. That of course included the amount of electrolyte. If you open a true long life electrolytic capacitor and compare the amount of liquid it contained to a “standard” capacitor, you see a big difference. Checking the data sheet of most any modern capacitor you find a reference life time of 10,000 hours at whatever reference temperature. How long is that time in more common terms? It is a little over 1 year... If the oldies would have had a 10,000 hour life specification, none of them would be any good after 65 years. Simply, they were made with smooth foil and plenty of liquid for self healing. And then the larger case kept the temperature down.
I found out that high quality capacitors have same size as old ones I was looking for 8200 and there's Mundorf a German company makes hi end version which is same size as one from the 70s 80s, while cheaper ones have gotten smaller.
They will tell you the ESR is lower so the cap doesn't heat as much. They never stop to question if they ESR being lower will impact the overall circuit as it was originally designed. The audio guys really get me, they restore amps because they want that old sound back and then put in all these low ESR caps and often modern replacement transistors. Does it work, sure...but it doesn't sound original anymore
@@AureliusR So then why not just go with a modern amp which is probably going to have better specs and be far far less work then fixing up old crap that doesn't sound original anymore? I'm of the mindset that people like the imperfection's which of course only come from the 'original' circuit. Tubes are a good example. Second order harmonics anyone....
"Why are the new capacitors so much smaller than the really old ones?" - So that devices that use them don't take up a quarter to half of your living room, that's why. With that kind of technology a satellite receiver (if possible to be built at all) would take half of your bedroom and would probably be bigger than the TV sets from around the same time. You can't have smaller sizes, bigger capacity and higher longevity at the same time, it's a compromise.
Main reason for the need to replace caps in more modern equipment is because cap plague was largely caused by [incomplete] stolen electrolyte formulas that made them fail. Those of us that grew up using caps that were made before this theft took place are aware of the fact that they still perform very well. TLDR: Get over it!
@@AmaroqStarwind Basically the "replace every cap" thought process is based on inferior caps. There are many cases where the caps in older equipment simply perform as well as they ever did, and replacement simply isn't necessary. The problem comes where higher speed switching supplies and such took over and the inadequate ESR and such became an actual issue. In things designed for use with the slower frequencies (50 and 60 Hz) as well as any use where there's no detrimental operation because of minor drift in tolerance, it's simply not necessary. I'm not saying caps don't age and eventually need replacement, just saying that they don't need it "just because they're old". It matters what they do, and what they're designed for.
Keep it up Marc! Do whatever the hell you decide is best. It's YOUR equipment after all. If it breaks it's yours, if it lasts a lifetime it's still yours. Only you have a say in what you do with it.
as time goes on i find that recapping is more and more necessary, 1950s caps are fine, 1960s caps are generally fine, 1970s-1990s, i almost always recap
@@memecoinmafia2732 The thing is with 50s and 60s capacitors is that it was rare to make compact capacitors since it wasnt really possible at that time. 70s-present there have been more and more pushes to make things smaller. With the improvements things got smaller and only slightly worse when it comes to durability
@@AlpineTheHusky Do not forget insane dimensions of old devices. With very free air circulation. Now everything is smd and every manufacturer produce devices as small as possible. 60c is new "normal", as time goes devices gradually increases incase temperature to 80c. Due to degrading components, dust and bad case location without airflow.
What a lot of cargo cultists don't understand is that any electrolytic that has been unused for a long period of time is likely to need reforming. This includes new old stock. Typically you will find that capacitors in need of reforming will exhibit higher than normal leakage and also higher than normal capacitance owing to thinning of the oxide layer. While reforming, it's wise to gradually increase the voltage up to the rated working voltage while keeping the current at or below 1 mA in order to minimise internal heating. It's also wise to monitor for increasing, rather than decreasing current. As noted by Marc, increasing leakage current is not a good sign. Note that a 100V capacitor that has spent its life at, say, 10V will likely still need reforming if reused in an application that will see it at 100V.
oldblokeh - Please provide a link to your “science.” I’ve been doing this work for nearly 50 years. Your technical, material science ideas must be in some peer reviewed publications with actual data that this old Ph.D. adviser can understand. My God. All those grad students I need to call and tell them we were all wrong about charge tunneling, polymorphic surface layer structure changes, and oh Lord, simple old atmospheric penetration! Please help me.
That sounds wrong to me. I have successfully reformed old electrolytics for years (since the '70s) with a good success rate. oldblokeh experience is similar to mine. I'm not really sure about your credentials, but I do actually have a bachelor's in EE. We *never* had *any* material about reforming electrolytics, or any other restoration work. That is something that hobbyists and restoration people do. Vey much not covered in college, especially even for more advanced degrees. Troll or BS artist, I'm not sure which...
@@boonedockjourneyman7979 How come you never came back and shared your thoughts about the paper posted here? Or are you still busy chasing down former grad students and informing them? LMAO
Never had to replace 80's Sprague caps in pinball machines. Mallory caps are meant to last until mankind's extinction also. Don't know how electrolytic caps were made back in the day but theses things are orders of magnitude more reliable than modern ones, not to speak about Chinese ready-to-dump craps. Keeping old but working caps also looks better in vintage machines.
But it’s a Mallory capacitor... they’re bulletproof! Seriously though, I replace capacitors in 30 year old Macs because they are cheap. Given the historical significance of this teletype terminal, I would only replace those items that have been tested and found to be faulty or dangerous. “If it ain’t broke …”
Plus we had the infamous Capacitor Plague which many people remember because computers built with those improperly built parts would fail, and essentially the entire production runs from those manufacturers need to be replaced. That sustains some of the belief in wholesale capacitor replacement, broadening suspicion from the era of crap parts, 1999 to 2007, to every capacitor ever made.
If I was going to replace the caps in this thing, I'd probably open up the Mallory caps and turn them into a shell using a smaller (physically) modern cap inside.
Hi! I LOVE your channel! In this video, you mentioned a capacitor that exploded, and nearly caught-fire inside an NAD amplifier. I was a bench tech for years in a shop that sold those. NAD issued a service bulletin, including a fix, for that problem. They'll probably pay you for your time and cost of parts. :) If memory serves, that capacitor has reversed voltage across it at power-up for some reason, and will fail again if the fix isn't done.
Hi Marc, I totally agree with what you're saying in this video, personally the only time I would ever seriously consider replacing a capacitor is if it shows a significant-enough amount of leakage of maybe a milliamp or something like that, but not something like 560 micro-amps as you're seeing with those 1953 vintage caps which were obviously very well-made.
Thank you for this confirmation, Because I keep hearing caps last around 20 years but I have equipment that's running for 40 + years just fine. But I have a tv that blew two caps after one year. So I don't want to replacing old caps that are working for newer caps what will not last two years.
I enjoy computing history, and love understanding how we got here, so a few years back I started that journey. I started with the machines which made a large impact on my educational experience, so the first machine I acquired as a restore project was a VT100. Most of the components checked out fine. There were a few tricky caps that were bulging from the bottom in the power supply, tricky Dicks! In the end I replaced them all anyway. In fact the entire power supply board is pretty much rebuilt 90% of the components. I have to go through and replace more components on the video board resistors/etc. Which I am planning on doing this year sometime. My reason is the same reason I replace the items on my car. I don't wait for a tire to fail, or the battery to die, or the belts to snap before I replace them when I know, they are 3-4 years old and have a lot of wear and tear on them. For fairly vintage equipment, I don't want to risk damage when some of the components on my VT100 for example, I am not sure if I could get them replaced without scavenging. Just my two cents.
The components are part of the historical context of when that particular device was made. I only replace failed components to preserve the originality. Replacing resistors as "routine maintenance" is truly crazy. By the 1970s, industry was using metal film and metal oxide resistors which are highly stable, and are the same type of resistors used today.
@@gregorymalchuk272 I actually use my museum pieces for fun and work, so I am more concerned over failure than your ordinary collector. Preventive maintenance is imperative.
That's a disturbing and insightful comment. I'm not near retirement age yet but sometimes I listen to and watch the ever rising tide of rote bullshit I start to think to myself maybe I need to start thinking about putting my feet up voluntarily.
My only objection to the earlier video was the interpretation fo the results from the Heathkit capacitor tester. Due to the range selection, you were testing at only 75% of the rated voltage, and the magic eye only opened to a sliver of a gap. I’m no expert, but I have two of the C-3s and two IT-28s. When testing a good cap at it’s full rated voltage (where possible) I typically see the window open nearly instantly to it’s full width. I would have judged that cap as marginal or “maybe good enough” at best. The test you did in this video shows that the leakage is improving with time at its operating voltage. If it took over 60 years to reach that level of leakage, I suppose it will last awhile longer. As a bit of a novice, it does make me think when experts such as yourself and your team have opinions that are diametrically opposed to others, Mr. Carlson’s Lab, for example. Thinking is a good thing, of course. Before I saw these videos, I was leaning towards the “just replace them all” camp, but now I’m not so sure. When I saw Mr. Carlson’s explanation, I was wondering why he was concerned that a fused input transformer would be damaged by a filter cap drawing excessive current. Maybe we can’t always rely on the fussing being properly designed, especially in consumer products. Looking at your transformer, it looks like you’d be safe. I have seen (I can’t remember where) demonstrations of coupling caps causing tubes to run hot, but that’s not applicable in a power supply. Great video series, by the way. I actually saw much later versions of the teletype in my USMC 5938 training in the early 1980s. Our equipment was a weird mix of things from the 1950s and state of the art digital automatic weather stations and satellite receivers. Thanks for your, and your teams, efforts. I am generally amazed. It’s great there are people putting in the effort to preserve this kind of history.
That's a good way to put it. We thought it was a "marginal good enough" cap, which would be fine in a filtering application. And probably not for a coupling cap, as Carl recently pointed out to me. I like following Mr. Carlson (should we call it Professor Carlson?). In general he does older stuff than us, with much higher voltages and consumer paper caps and all, so I am not sure that his experience directly translates to our more recent and professional stuff.
"Maybe we can’t always rely on the fussing being properly designed, especially in consumer products" Fusing exists to protect wires, not the transformer. You can easily destroy a winding without exceeding the mains current rating.
@@userPrehistoricman That is not true in most cases. Cap-changers are the type of people who don't seem capable of doing proper diagnosis and repair if an individual component fails.
@@gregorymalchuk272 Thankfully i wasnt one of those cases, my Macintosh LC had electrolytic all over the board coming from ever capacitor. thankfully no traces were damaged. but that is the issue with people just replacing capacitors even if they are good, a lot of people would say to recap my dell dimension 4100 but im not going to because all of the capacitors are fine and people say the capacitors from 1999 to around 2007 are always bad, my dell dimension 9100 and 4100 are both fine just the 9100 has 1 bulging cap but what do you expect it was running for probably 3 years straight
These capacitors have lasted 65 years and you have proved they are still working well, it would be crazy to replace them with new capacitors with an unknown lifespan.
The rule I use is, if it's encased in wax it should go, though if I check it and it has no voltage across it I will probably not replace it, but wax coated paper/electrolytic caps are nearly always bad, if not they will definitely go bad eventually. The reason is because the wax let's moisture in and will degrade the electrolytic, the paper also goes acidic.
I just converted a tube radio into an AM transmitter, since somebody already stripped half of it (lytics included) I didn't have any filter caps to use only a 50 year old elwa branded filter, I installed it and the transmitter is working fine to this day. I only change old filters like that if they are damaged on the outside too much, are dry or have noticeable faults. The capacitor itself reforms when in use so those are actually gonna get better with time.
I'll start with the most important point: It's your equipment, do what you want with it. As I understand the issue, there were a few years in the late 1990's where a lot of low-quality "clone" capacitors were produced and used, so it is generally generally smart to replace capacitors of that vintage. These caps, on the other hand, seem to be from the era when stuff was genuinely made to last. The wiring also looks a lot more robust than the relatively fragile, paper-thin etching used on modern PCB's, so I doubt the cap would cause irreparable damage if it were to fail catastrophically or short out. This device is also designed to work at high voltages, so a short to ground wouldn't cause nearly the same catastrophic failures as if it were releasing 110/220v into a 5v digital circuit. I see commenters more knowledgeable than myself discussing manufacturing differences and thermal properties, but I will refrain from discussing those aspects in detail for lack of knowledge on my part. If any of my assessment is wrong, I ask the experts to correct me, but you don't have to start a war over it; I am genuinely interested in learning more about electronics in general.
Great video - it really needs to be said. Since returning to electronics, this time as hobby, I've collected scores of vintage test equipment (largely HP & Keithley) and with a few exceptions most of the electrolytic caps are solid even after several decades. Since most of this used equipment cost thousands if not tens of thousands new the top of the line manufactures could afford specifying the best. Also the military was big customer as well further prompting higher quality and reliability standards. If I bought a consumer grade product of similar age I probably would be more suspect of their electrolytic's since they where not driven by the same expectations or objectives.
As a fan of everything that's vintage I go for the philosophy: "Keep as many original part as you can, replace only broken or dangerous part if repair is not possible". And, although I don't know very much of electronics, I think you're right. I always enjoy your videos, because you make what you're doing easy to understand also for who is not an expert.
@Lassi Kinnunen Hogwash. The entire reason people collect old equipment is to preserve the historical context. The components used in the device are PART of the historical context. I only ever replace the ones that fail. Heck, by that logic, replace all capacitors, resistors, circuit boards, speakers, wires, CRT, case, and just call it a new device?
I don't know from replacing old caps or no... I just like what you gentlemen do! The first simple program I ever wrote (in high school), I uploaded to a mainframe to run and get output by keying it into an old teletype machine. It was one of the most amazing things I ever saw up to that point in my life! Great content!
You, sir, are living the dream! BTW, I also (still) own a model 15, which back in the day I had hooked it to my TRS-80 as a printer. A little z80 ascii to baudot assembly code never hurt anyone. Also noticed that the rolls of paper that used to be used for drying your hands fit in the machine as print paper although clearly not optimal. The cheap stuff wasn't very absorbant for it's intended purpose but worked for ink in the teletype. Oh, I used a model train transformer as the power supply to the selenoid, controlled by a transistor driven by ttl.
Thumbs up! When I repaired my HP 8671A and HP8566B and only replaced the failed capacitor. I got similar comments on why I didn't replace all the electrolytic caps. My philosophy has always been that unless there is a good reason (e.g. high ESR and low capacitance), leave them alone! New ones aren't always better!
Thank you so much for making this video. I repair amplifiers. Very rarely do I ever have to replace capacitors. The main issue I’ve seen is leaking caps from the 1990s. All capacitors from my vintage test gear and power supplies from the 1950s-1980s are still within spec. People need to lay off the snake oil and actually do some research and read. I especially love the one where caps in the rails or filter caps for power supply for car audio amps make a difference in the sound😂😂😂 If a capacitor is still within spec why risk damaging the board or pulling Vias because you listened to an ignorant person off the internet who doesn’t understand how to test capacitors. There is much much more to capacitors other than farad measurements.
Many folks don't know how to think. Some older electro.s do dry out, but you test them and confirm they are in rigor mortis; not wholesale replacement.
For me I ended up in studio/broadcast engineering and related electronics where the default of about any age from old tube radios I like to dabble with like Mr. Carlson (who seem to be against reforming) to modern solid state stuff seems to be recap and try to align is the first steps to bringing it around working logically of course going from the power supply out fixing as we go. It seems they can often be reformed and will work but if it does fail it's usually pretty spectacular and can take out other things not easily replaced on vintage gear even if it fails open you now have dirty supply voltage. I just recently redid the power supply in a 70s? mixer console for this reason.... when you can buy a smaller, better spec new one and have some peace of mind over "is that cap acting up and causing ghosts in this very expensive hard to get parts for thing has to be running 24/7 on the air? Is it going to blow up and fry it all? " Is it going to go out of spec and get me a nastygram from the FCC? For a part you can usually find new for not much and a little bit of shipping time if you live in the middle of nowhere like me. After having a big one go critical in a confined shop where everyone especially when the military trained engineers who taught me what I know including recap all the things hit the deck out of reflex has got me in the don't second guess just replace it camp..... SO... I am not arguing because I have learned every engineer has their own ways of things that they swear by usually a little different from others and won't budge on some things so this is only a story about how I formed mine and how it would go when we would butt heads. We could even be saying the same things just them in older terms/methods and me saying newer ones would end up in an hot argument to the point even the non technically minded at the station would be like "guys you are arguing both sides of your own arguments with each other again just pick one." It's really the only time we argued. Out of 4 of us in the area when I started I am the only one left not retired or passed on now. I am told being in my 40s makes me one of the younger out of the few left in my trade and have to go with what they taught me or I pick up on my own I never took courses on any of this or even went to high school.I learn by doing with my hands from them and that's how they do. What to do with old caps seems to be a spicy subject for pretty much everyone in our circles no matter which side you fall on.
To start, let me say don't let the haters get you down. Your channel, your methods, my viewing pleasure. To be honest, I was surprised the old electrolytic capacitor survived as well as they did. And I'm most certainly guilty of replacing a lot of them. I do however test at least a few of the capacitors of every device I go through to make a proper evaluation. In my opinion how bad they will be depends a lot on make, model, material quality and usage/load (did I get every variable there :)) But there are good generalizations. Paper wax for example is almost always bad, but even early metal foil types are mostly fine, unless their cases cracked or had bad sealing against moisture or such. Liquid filled electrolytic's that dripped and corroded your chassis are bad of course, but the more solid ones might be fine. The example vintage capacitor in this video was military spec, solid and had low frequency (and maybe quite constant) load in contrast to the bad ones you found in the alto power supply. So it survived well. I'm curious though what their spec MTBF was. My guess the reason you see so few bad ones: Your repairing mostly high end expensive gear. Maybe that stuff was actually worth its price :) cannot wait to see the next restore on the teletype
I appreciate your video. I have noticed that many vintage computer restorers reform the caps rather than replace them, unless they show signs that reforming was unsuccessful. I have always wondered about that. I have wanted to learn more about the failure mode of capacitors in general. I do a little bit of work on vintage radios and have seen many electrolytics go bad causing the familiar hum. I have also had them get hot at voltages lower than they normally run at. That was cause for concern. Perhaps if they had been properly reformed that wouldn't have been an issue. Not sure though. I have also taken some apart and have discovered badly corroded internals leading me to believe that vintage caps are always suspect. But, there was always those vintage computers that seem to work just fine with their original electrolytic caps. The only conclusion I can make is that those old caps are of superior quality, especially ones used in commercial equipment. It also appears that replacing the caps should be on a case by case basis. Given the low cost of caps for vintage radios and similar equipment, I may still replace them just to be sure but I can certainly see why you don't want to just replace them because they are old. Thanks for a very interesting and enlightening video.
When capacitors are leaking, bulging, leaky, open etc, then they need replaced does not matter the age. I have replaced relatively new capacitors because they were bad, and have left old ones in the circuit because they were good. This is just part of troubleshooting. The reason that some recommend replacing old ones is because they are poor quality to begin with, as in consumer quality.
I've seen electrolytic cap failed short once. It was already repaired and assembled device which was turned on for final tests and I was lucky to watch whole process. Can't remember it precisely but I believe it was something around 200uF and 350V vintage cap from early - mid 70's. It was a filter cap on 250V rail. The only consequences was humming transformer for a short period of time and burned 10W resistor. Device worked fine after I changed resistor and cap.
Shorted tantalums are the best, instant firework display right in the comfort of your lab! :D Also, an apt AvE-ism: "if it ain't broke, fix it 'till it is!"
I have caps in my 80’s guitar amp that still check good, so they stay. Why change them if they test good? Seems alot of the new ones are worse than the old stuff. More surface area = less heat and stress.
There will always be haters and/or people trying to appear smart.. I would recommend to try to ignore it, but on the other hand I do appreciate this extra video with more details on the details!
I work at a fusion energy research facility in the UK. I am literally surrounded by millions of capacitors of all different types and ages. We had one go pop last year.
I've seen intact capacitors in 1960s gear, but especially large, high capacity ones often fail by way of drying out (=below spec capacity, high ESR). It's my understanding that this is a self-accelerating process: Electrolyte escapes -> ESR rises, capacity sinks -> Lower capacity means larger charge angle and deltaU -> Larger and broader current spikes -> more internal heat -> more electrolyte escapes. One particular issue is that it is difficult to source capacitors with equivalent ripple currents. Premium / Long-life non-miniature capacitors kinda match up, but it is doubtful whether they will last as long as the original, much larger capacitors, since the new capacitors are often stressed fairly closely to their limits, while the older ones had much more breathing room.
Some of them actually last longer than the new ones! The people telling you to replace them don't know anything about what they're talking about half the time anyways. What people also fail to realize is that vacuum tube equipment is far less sensitive to these things than digital electronics are - A television will usually work perfectly fine with some slightly off capacitors or dying resistors in circuit! Watch shango066's ressurection videos for proof of concept there. If it can sit outside in the California desert for 40 years and be brought back to life by cleaning a switch and changing two capacitors, then there's obviously some serious leeway in the component values!
Fully agree with you Marc. I'm a restorer myself and pro technologist...Carpet-Bombing with replacement caps (otherwise good resident caps) does more collateral damage that good. As a matter of facts, If it happens that see an add claiming that a unit have been "Fully Recapped/Restored"...I pass my turn as the reliability and ESD damage seriously compromise the longivity of it. Thanks.
I totally agree with the point of this video. It all goes to "if not broken, don't fix it". Some "experts" just blindly recap every piece of gear they touch, just because it is so, or just because they saw a video. Basically, instead of doing an analysis on what needs fixing, they just conclude "oh, it must be the caps, because they are old". And all they end up doing, in some cases, is replacing excellent vintage caps from reputable brands, by modern capxon or onehanglow crap. And, they also end up soldering them with blobs of solder, because the morer the betterer.
Many of the 1960's components were made extremely well, so it stands to reason that they would last a long time. I know we rarely change capacitors in vintage Marshall amps, since it effects the value of the amp and the old ones sound good.
I am with you that good electrolytics might last but I replace them a bit earlier. They have to perform well in their temperature range and up to their specified voltage. I usually don't want to bother testing a half decent cap (that shows alright at room temperature) at 70C and 400V so I'll just replace them. The 2-3$ are not worth the hassle. But I don't do historic restauration so the looks and authenticity is not important to me. And, I don't want them leaking. Like, literally leaking electrolyte. PCB mounts are way worse than the standing cans seen in this video
In the interest of safety it is always wise, though maybe not absolutely necessary to change old electrolytic caps for new ones. Those old bulletproof Mallorys may be the one exception, though for how long? All vintage caps should be tested before being powered up after a long hiatus, and if they perform as the video shows then you can probably leave them in for a time, but they will fail eventually, even the Mallorys. As an example, I own a 1939 Hammarlund HQ-120x shortwave receiver that I inherited from my great grandfather. At his death in 1954 it was put in a closet where it sat for 25 years. In 1979, as a young electronics technician just starting his career, I unpacked it, put up an antenna and fired it up. It worked great!... for three days... then the main filter cap blew nearly taking out an irreplaceable 30 Henry choke with it. I don't recall the cap's manufacturer, it may have been a Mallory, it may not have been, but the moral of the story for me then and now is that it is safer and more cost effective in the long run to replace old electrolytics rather than leave them in and hope they will work for years to come.
Thanks for this! I hate seing categorically replacing all electrolytics in everything they work on and thinking that they are doing the equipment a favor. Electrolytics in older equipment are mostly all goood in my experience. If it lasted 30 or 70 years why change it? It's a reliable cap, and no one guarantees that the replacement one is going to last longer than the old one if left in.
I agree. If an old electrolytic can reform and test with good capacitance and ESR why replace? I run alot of 1950's equipment here and like you said they rarely short go open and it's quite obvious when that occurs. However, the paper / wax / bumblebees are quite the different story. Don't fix what is not broken! Alot of ppl just bulk replace capacitors not knowing that the originals are still likely going to still exceed the lifespan of newer , contemporary components. Great video, thanks for posting it!
I've read somewhere that a lot of those electrolytic caps made back in the 70's were reckoned to have a maximum working-life of about 20 years before they could be expected to fail.
That Edison socket fuse might be the same standard I saw in European house wiring of that vintage. If so the height of the bottom contact in the holder sets the maximum fuse rating, as higher rated fuses are shorter from top of threading to contact point. A same size successor standard (upgradable) had a ceramic ring around the bottom point with higher rated fuses having a larger tip diameter, and the threading being on a reusable porcelain holder that accepted the new bottle-shaped fuses.
That is an interesting tit-bit f information about the earlier fuse standard. I have seen them from a distance on some industrial equipment some decades back but they were rare already 40 years ago. You can buy the small ceramic current selecting rings that go in the bottom of a bottle fuse holder at the hardware shop in Finland, if you change the purpose of a circuit and want a smaller fuse or replace with heavier wire for a larger current you might adjust the fuse holder in an old panel . All new build and larger renovation is done with modern circuit breakers but I have lived in a house with the bottle fuses and seen many others.
Well done, thanks for sharing. sure, if someone are servicing a switched PSU that are in daily use and replace all caps for 20$ as a precaution , it's a different scenario. Glad I stumbled over your channel.
I repair vintage audio gear for a living. And as you might imagine, I get requests to recap gear all the time. I tell people that I will test the old caps, and if (a) they pass a visual inspection, and (b) they pass my leakage, capacity, and esr tests, then there's really no need to replace them unless they simply insist.
I'm glad I saw this. My nad 3130 amp sounds great yet ther is so much on how I should change the caps. I know elite caps aren't the best in the world but it's been looked after and dusted and hoovered regular and I keep thinking I'm happy with the sound so what would be the point? Bloody forums
@@Derundurel just checked you're model out. Couple if years older and still going strong which is reassuring. I'm guessing you have the green elite caps which a couple of beefy Black nippon chemi-con ones? Which to be fair can't be the worst out there, considering the later JH ones got a fair ear bashing
Caps in consumer products tend to be the cheapest that could be sourced and why those tend to go bad with age. The "capacitor plague" didn't start till sometime in the 2000's when it became pretty difficult to not get defective electrolytic caps from most manufactures.
We used to have those fuses that fit into lamp sockets in the houses of the 1950s. Some folks would replace the fuse with a penny, which in house wiring was a very bad idea, as it could lead to house fires.
Hello Marc, I enjoy your videos, although the reason for this is quite sad. I’m pretty sure, that you guys know, what you are doing. The problem is, that your main audience is not aware of the robustness of vintage industrial equipment, which can’t be compared to modern consumer electronics. Please don’t care about the - sorry - stupid comments! Keep on going!
When deciding whether to replace a vintage capacitor, I mainly look at visual appearance, capacitance, capacitance frequency response, leakage and ESR. Last is the amount of work to get the job done nicely and the risk of not doing it. In many vintage repair job I did was to gut the original capacitor and fill the cleaned empty can with modern parts so they look the same after repair. I often found the inside of vintage cap almost bone dry, even with good capacitance and leakage numbers. Capacitor like this are associate with increased ESR and they are not going to last very long. Over the years capacitor technology improved a lot with thorough understanding of failure mechanism and reliability. We see unreliable results sometimes because we overlooked subtle circuit parameters or bought low quality/incorrect parts. Not all capacitors are created equal, today some parts are deliberately designed with shorter life span because of cost/size/temperature/ripple current .....reasons. I use better spec parts to make sure I do not need to repeat the same work again in the near future.
Capacitor technology improved, but due to cost cutting, some new electrolytic capacitors are just too unreliable, even if they are from reputable brands. It is not an accident that a new capacitor of a given capacity, voltage rating and temperature is way smaller than a vintage capacitor with the same specs. That is due to cost cutting, and not due to improvement. The new capacitors have much less headroom in terms of voltage.
I've been an electronics hobbyist for decades now (I'm 53), and I've repaired many vintage equipment over those years. But I've always checked electrolytic capacitors for ESR and capacitance, and only twice did I encounter failing ones (aluminum types that were cheap unknown brand in a linear PSU, and some low capacitance ones in a Roland/Boss guitar effect pedal). I've got many 30 and 40 years old circuits with high quality ones from known brands (Matsushita, Philips, Brel, Mallory just to name a few). So this "you MUST replace electrolytic capacitors if they're old" phobia, regardless of their condition, is also annoying me a little, too. Thanks for trying to clear this myth. I've had many more Tantalum caps go bad on me (some actually shorted out), than electrolytic ones.
I had a 1964 Super Reverb that had never had much use, and the original high voltage supply capacitors reformed and tested fine, so I left them in the amp. Within a year of being back in regular use, the amp started to hum and then started to blow fuses. The lesson is that it's very difficult to take a snapshot of a component and then extrapolate from that how it will perform into the future unless you have extensive experience with that particular piece of equipment and the capacitor model and brand. Even then, your prediction may turn out to be wrong.
When people say they need to be replaced i think they mean the wax paper ones, those are almost always leaky, by leaky i dont mean like leaking a liquid or other substance from the capacitor but leaking dc voltage across them which they are ABSOLUTELY NOT SUPPOSED TO DO. They are supposed to only pass ac and block dc, if its clearly a waxed paper or bumblebee capacitor, IT HAS GOT TO GO, paper in oil are likely fine but waxed paper caps have got to go or be restuffed with new capacitors in the old casing.
To recap or not? There's to many "if" to take in to consideration. My case, I tend to follow the manufacturer instructions, by average this parts has a max 2 years shelf life and a life up to 10 years, I also consider de deterioration of the chemicals inside the component, contamination, etc. Since they are considered "wet" type capacitors, they also tend to dry out over time. Another thing that I take in to consideration is that this tests might indicate one thing, under load the characteristics might change. There are some that can be treated/reconditioned and others that can't, it's up to the technician to decide. Is the client willing to pay for the job and the technician bench time? Does it worth it? Etc, etc.
The reason why I replace a lot of electrolytic capacitors is convenience. If I have the device taken apart to fix a problem and find that some electrolytic (or film/paper) caps are bad, it is better for me to just replace them all right now instead of just replacing the bad ones, then some time later having to take the device apart again to replace some other failed capacitor. Once I took a motherboard with a bunch of bulged capacitors to get it fixed and specifically asked them to replace all electrolytics there. They only replaced the bad ones and less than a year later the other ones failed and I had to replace them. I also have a similar experience with some tape decks made in the USSR in the 80s - replace only the bad caps and you will be taking the device apart again to replace some other cap rather soon. So, if the device does not have problems with capacitors, I leave them be, but if problems start because of the caps, I would rather replace them in one go rather than having to fix the device multiple times.
I often find the smaller older axial caps to be a problem. Lower voltage, particularly. On modern aged gear, the bigger caps usually don't fail as often, and they are usually much more expensive too.
One answer to this is experience , I have found generally most but not all vintage caps to be in okay or good order, it seems to be only the 1990 and a few years ahead that fail in the worst manner. I have been involved in electronics for over 40 years and have seen it all and repaired all sorts of weird stuff.
@John Smith The 'death cap' is a film capacitor in a specific location (namely between the AC input and chassis, which is why it's deadly when it fails shorted). No sane repairer could conflate it with the other capacitors in the unit (certainly not with electrolytics).
I vaguely seem to recall that the dielectric of those old Mallory caps was oil based and not a water based electrolyte. They cannot go duff in that case...
Most of my instruments still have all the original caps. During Restoration I measure leakage, reform if needed, measure C an D. If all values are within specs stated by the datasheet, or if no data can be found, from my own lists it will be placed back in the instrument. If not good I replace it for a new high quality comparable cap. That is the same specs for WV, leakage, ripple current, D, temperature and hours of use. Often old good quality caps turn out to be better as good new caps. This whole witch hunt is caused by people repairing consumer stuff without much knowledge and armed with their holy in circuit ESR indicators and the new trend, a LV leakage tester... I use several LCR meters and bridges from HP, IET, ESI and GR and 2 HV leakage testers (my own design) My best bridge is a GR-1520 ( with a famous history being the EX-Genrad standards lab bridge)
Here is one I like. It's a little slow but accurate and thorough: ruclips.net/video/ifWfvaIbWZY/видео.html . Reforming simply consists in progressively applying voltage to the capacitor while monitoring the leakage current. Which, if all goes well, reforms the thin oxide on the electrode to its original condition and gets rid of the leakage. Not only valid for old capacitors, but new ones that are kept on the shelf for a few years without being used also benefit from reforming. However just putting them in use in a regular circuit without further precaution is usually enough to do that, so nobody thinks about it.
@@sheep1ewe The new replacements will fail sooner than these originals unless you buy the highest grade, custom made, custom specced Nichicon or something.
@@jan.tichavsky Yes, there are still a few good brand on the market, but it apends on the particular type (as You probably know), for example some types of old cheramics has no real experience date, they can still be good after many, many years, but sheap E-lytes usually last around 10 to 30 years still conciderable reliable. But in this case those bigger ones also represent a part of the historical value of the device it self as long as they work good enough for this purpose, as i see it, i think that was the point of testing and saving them. In equipment with no real historical value i see no problem replacing everything, but in very old apparates the component it self also hawe a historical value if it can be preserved without risking damage other, more valuable parts of affect the function badly on a device that are restored to be functional. I found a video here made by an old man who saved the papershell with the original markings of old wax capacitors, but he putted modern ones inside the original shells and re-waxed the surface, it looked realy awsome at my opinion when it was finished. (Sorry for my English, it's not my native language...)
Could you make a video on the capacitors diagnostic? Like is a component tester (that shows capacitance and ESR) good enough or do you need to power it to its working voltage and measure the leaking current or what?
when it comes to caps I always test before replacement. For example when it came to restoring my 1965 IBM 1800 i tested all the caps and all were fine, some were boardeline but its still on its 1960s filters and its working perfectly although i have a memory circuit error meaning i can only access 28 of the 32kw of core.
In my experience with tube equipment, about 50% of electrolytics need replacement. Many Philips 1950s caps are totally fine. I have some radios which have the original electrolytics (not the original paper caps!) and have been in use for thousands of hours without failure. I have just as many radios that have new electrolytics, but those did not respond to a reforming voltage, were physically leaky or had too low capacitance. Always measure leakage at working voltage. No significant leakage after reforming for a good amount of time? Capacity okay? Leave it in!
You could drop that power supply out a second story window and the only damage would be to the pavement. My dad was a radioman shipboard in the Navy during world war II and he used to collect these old military power supplies and other equipment and one of the reasons that these 70 yr. old caps are still good is that the military only used Mil Spec parts. They demanded reliability in the components they bought and then took those they received and pushed them to way over their rated specs in house and if a cap could survive that torture test they were placed in service. Cap failure is tough thing to predict. You can use a tester like the Heathkit to weed out failed caps or ones on the way toward failure but you can't predict when a good one will fail. There is no crystal ball on that tester that will give you that info. So automatically replacing all the caps on old electronics is a judgement call.
Thank you for the much better and more detailed testing of the electrolytics. The results from the heathkit tester where not that conclusive. I did think they looked a little new for 1930's. Being from 1953 the paper inside the capacitor would have been impregnated with PCB oil, those caps do last much longer (PCB oil kept being used in big electrolytics until the late 60's) 1930's caps are almost always dried out. As for the 80's caps in the SE/30 Surface Mount caps of the 80's where at about the same point as standard caps in the 20's. In the 90's there was a flood of fake capacitors and a fake electrolyte formula floating around, Panasonic and Elna caps where effected the most.
Most of the failed vintage capacitors I have seen were melted wax capacitors. For me, its been a 50/50 shot if the metal sealed electrolytic capacitors are OK. I like to test them first, though, when I get my hands on something vintage just to be sure they are OK :).
Question since you're an expert.. If I own high end audio equipment from Accuphase, dCS, and McIntosh and I want it to last as long as possible... is it better to leave it on all the time? or turn it off when not in use ... in terms of keeping the capacitors and other components health inside.. and last longest?
I recently picked up a 1934 Telefunken radio. It looked like it had some work done back in the late 1940s, but the cathode bypass electrolytic capacitor was still original. I was amazed that it read in spec, both on capacity and on ESR. Also the leakage at the rated voltage was acceptable, at least for a device that was probably last used in mid 1960s. However, paper capacitors are all shot, some measured as low as 2k Ohms, this is almost a short on a 280V DC rail, and would quickly take out the very expensive RGN1064 rectifier.
Merci Curiosmarc pour cette video pedagogique, je n'avais jamais compris comment tester la fuite d'une capa sans testeur rlc. thanks Petite question de debutant, j'imagine que par precaution, si l'on est pas sur du condo, au moment de la mst, on regle la regulation courant assez bas pour eviter d'eventuels court-circuit fatal genre electroboum et electrolyte partout? en fonction de la capa quelle est la fourchette admissible de courant de fuite ? Merci
I had a late 1980's TV explode an Electrolytic in my face in the early 1990's while I was looking for a completely unrelated fault. Yes the boiling electrolyte made it explode. Old fashioned capacitors didn't have this issue. If they are good then they are more than good enough to keep as is. This was way before the capacitor plague.
Most of my capacitor failures tend to be 1uf 50V types and there about's. The bigger ones seem to be more reliable. My biggest problems come from modern surface mount caps that leak all over the PCB and ofter cause unrepairable damage to the board - I mean how on earth can you clear away electrolyte that has permeated into the middle of a PCB. Fix a 1950's power supply - quite possible, Fix an 8mm camcorder with leaky SMD's - forget it!
I think a lot of these people advocating changing certain caps out don't really understand what causes a capacitor to develop internal leakage, they assume that the cap naturally does it itself, but for those who don't understand, a capacitor is made-up of two conductors that are separated by an insulator, or dielectric, normally the insulator or dielectric presents a very large resistance to D.C., the insulator/dielectric has a certain voltage-rating it can withstand before it breaks-down and starts to conduct, most good insulators have a D.C. resistance of at least a few giga-ohms (well at least an ideal insulator should anyway) a giga-ohm is one thousand mega-ohms, normally you would design an electronic circuit so that the expected voltages in the circuit do not exceed the maximum working voltage of the capacitor, or, as I personally like to do, you would use capacitors with a working-voltage greater than what you would expect to see when the circuit is operating normally, so, you wouldn't use a 25V electrolytic cap in a circuit where you'd be expecting something like 40V max, the 25V cap would go "BANG" and let-out the "magic-smoke" as well as a particularly pungent smell....Tantalums are notorious for doing that, and believe me it's not pleasant, I know from first-hand experience.
Very good. Altho my beef is with some more modern caps. Got hold of a nad c352 amp. It performed really bad so I checked out the caps. Despite the amp being around 15 years old, many of the caps were visibly failing. The cheapo caps I replaced with mainly Panasonic and it's amazing now. Just noticed you had a nad cap gone bad also. The two main caps are still original and are starting to bulge! They are 120k uf and will cost a fortune to replace. In vintage gear I've come across several failed caps, and replacing them did indeed fix the faults. But a lot of the time caps seem ok and I'm sure they were made better back in the day 🙂
the only time it makes sense to recap is in 1980+ Designs where the people started penny pinching in Homecomputers , causing some parts to be stressed more and those capacitors tempt to fail (looking at you Amiga 500) , otherwise its a waste of time and money
Marc isn't refusing to change them. He changes them if needed, but only if needed. In old tube gear, the reason leakage is dangerous concerns circuits where the capacitor is a DC block and leakage would introduce bias on the grid of the next tube in a chain. The bias changes the operating point and can overstress the tube by causing excess conduction - due to the DC voltage on the grid. In this case, leakage is bad and has to be very low to be tolerable. If a circuit would be damaged by exceeding some current, leakage might be a problem but the design of the circuit sets the bar for when leakage becomes harmful. A power supply should be designed so that even a dead short across the output will be handled - perhaps by a fuse (and maybe a crowbar device) - thus the filter caps aren't a risk. Their failure impact would be inadequate filtering, excess ripple, rather than excess current. It depends on what the capacitor is doing and the circuit in which it is used, in other words. In the real world, every capacitor also has leakage, series resistance, and inductance. The specifications for new capacitors include that information, allowing a designer to pick a very low leakage if that is critical for the circuit.
My favorite quote from an old IBM mainframe computer service manual:
"Never attempt to repair or replace equipment which is working to the customer's satisfaction."
Or in colloquial terms: If it ain't "broke" don't try to "fix" it.
This reminds me of the robot that created the celebrated light painting in Asimov's collection of robot stories. The maker performed an unauthorised service on the robot and the wanted malfunction was removed.
unfortunately for me, I can't learn that lesson.
Why "never" though? What's the problem in replacing?
@@nbudzinski. Thank you for the explanation!
It's almost like a well trained team of vintage electronics restorers know what they're doing. Fancy that. Keep up the good work, Marc.
I wonder if people insisting on wholesale replacement have ever stopped to really think. What is behind the fact that they replace a 10 cubic inch capacitor with a “modern” 1 cubic inch version? In comparison with resistors we can conclude that a smaller package does not dissipate as much heat as the large one does at the same temperatures. Or, if the same heat generated has to be dissipated from the small package, its temperature is higher. And higher temperatures are what is a main detriment of the component life time. Just check any document that refers to the Activation Energy.
Why are the new capacitors so much smaller than the really old ones? Some time after WW2 the manufacturers started increasing the aluminum foil surface area by etching (roughening) the surface and reducing all the materials, increasing the per unit stresses. That of course included the amount of electrolyte. If you open a true long life electrolytic capacitor and compare the amount of liquid it contained to a “standard” capacitor, you see a big difference.
Checking the data sheet of most any modern capacitor you find a reference life time of 10,000 hours at whatever reference temperature. How long is that time in more common terms? It is a little over 1 year... If the oldies would have had a 10,000 hour life specification, none of them would be any good after 65 years. Simply, they were made with smooth foil and plenty of liquid for self healing. And then the larger case kept the temperature down.
I found out that high quality capacitors have same size as old ones I was looking for 8200 and there's Mundorf a German company makes hi end version which is same size as one from the 70s 80s, while cheaper ones have gotten smaller.
They will tell you the ESR is lower so the cap doesn't heat as much. They never stop to question if they ESR being lower will impact the overall circuit as it was originally designed. The audio guys really get me, they restore amps because they want that old sound back and then put in all these low ESR caps and often modern replacement transistors. Does it work, sure...but it doesn't sound original anymore
@@jeffm2787 Okay, but "sounding original" is the stupidest goal to try and achieve anyway.
@@AureliusR So then why not just go with a modern amp which is probably going to have better specs and be far far less work then fixing up old crap that doesn't sound original anymore? I'm of the mindset that people like the imperfection's which of course only come from the 'original' circuit. Tubes are a good example. Second order harmonics anyone....
"Why are the new capacitors so much smaller than the really old ones?" - So that devices that use them don't take up a quarter to half of your living room, that's why. With that kind of technology a satellite receiver (if possible to be built at all) would take half of your bedroom and would probably be bigger than the TV sets from around the same time. You can't have smaller sizes, bigger capacity and higher longevity at the same time, it's a compromise.
Clearly the solution is to recap our old Macs with 1930s capacitors :D
SAerror1 thats what I was thinking.. lol
Main reason for the need to replace caps in more modern equipment is because cap plague was largely caused by [incomplete] stolen electrolyte formulas that made them fail.
Those of us that grew up using caps that were made before this theft took place are aware of the fact that they still perform very well.
TLDR: Get over it!
@@MadScientist267 Wow. I would love to see that proper formula.
@@AmaroqStarwind Basically the "replace every cap" thought process is based on inferior caps. There are many cases where the caps in older equipment simply perform as well as they ever did, and replacement simply isn't necessary.
The problem comes where higher speed switching supplies and such took over and the inadequate ESR and such became an actual issue. In things designed for use with the slower frequencies (50 and 60 Hz) as well as any use where there's no detrimental operation because of minor drift in tolerance, it's simply not necessary.
I'm not saying caps don't age and eventually need replacement, just saying that they don't need it "just because they're old". It matters what they do, and what they're designed for.
@@MadScientist267 It also depends on the type of cap. Wax caps have the same sin than electrolytics with bootleg formulas
Keep it up Marc! Do whatever the hell you decide is best. It's YOUR equipment after all. If it breaks it's yours, if it lasts a lifetime it's still yours. Only you have a say in what you do with it.
Thanks. Exactly. We are evil absolute rulers. We can keep the old caps in there even if no one likes it. Mwhahahahaha. I feel better already.
Love your content as always Marc, and the sense of humor! Muahhhahahaha
There are a lot of idiots out there that don't know what they're doing or looking at.
Marc isn't one of them.
as time goes on i find that recapping is more and more necessary, 1950s caps are fine, 1960s caps are generally fine, 1970s-1990s, i almost always recap
better made in the 50's ?
Better made? Or survivorship bias. That's another possibility I suppose. @@memecoinmafia2732
@@memecoinmafia2732 The thing is with 50s and 60s capacitors is that it was rare to make compact capacitors since it wasnt really possible at that time. 70s-present there have been more and more pushes to make things smaller. With the improvements things got smaller and only slightly worse when it comes to durability
@@AlpineTheHusky Do not forget insane dimensions of old devices. With very free air circulation. Now everything is smd and every manufacturer produce devices as small as possible. 60c is new "normal", as time goes devices gradually increases incase temperature to 80c. Due to degrading components, dust and bad case location without airflow.
What a lot of cargo cultists don't understand is that any electrolytic that has been unused for a long period of time is likely to need reforming. This includes new old stock. Typically you will find that capacitors in need of reforming will exhibit higher than normal leakage and also higher than normal capacitance owing to thinning of the oxide layer. While reforming, it's wise to gradually increase the voltage up to the rated working voltage while keeping the current at or below 1 mA in order to minimise internal heating. It's also wise to monitor for increasing, rather than decreasing current. As noted by Marc, increasing leakage current is not a good sign. Note that a 100V capacitor that has spent its life at, say, 10V will likely still need reforming if reused in an application that will see it at 100V.
oldblokeh - Please provide a link to your “science.” I’ve been doing this work for nearly 50 years. Your technical, material science ideas must be in some peer reviewed publications with actual data that this old Ph.D. adviser can understand.
My God. All those grad students I need to call and tell them we were all wrong about charge tunneling, polymorphic surface layer structure changes, and oh Lord, simple old atmospheric penetration!
Please help me.
How about an MSc thesis: digital.library.unt.edu/ark:/67531/metadc3104/m2/1/high_res_d/thesis.pdf
That sounds wrong to me. I have successfully reformed old electrolytics for years (since the '70s) with a good success rate. oldblokeh experience is similar to mine. I'm not really sure about your credentials, but I do actually have a bachelor's in EE. We *never* had *any* material about reforming electrolytics, or any other restoration work. That is something that hobbyists and restoration people do. Vey much not covered in college, especially even for more advanced degrees. Troll or BS artist, I'm not sure which...
@@boonedockjourneyman7979 How come you never came back and shared your thoughts about the paper posted here? Or are you still busy chasing down former grad students and informing them? LMAO
@@oldblokeh how do you reform a cap?
Never had to replace 80's Sprague caps in pinball machines.
Mallory caps are meant to last until mankind's extinction also.
Don't know how electrolytic caps were made back in the day but theses things are orders of magnitude more reliable than modern ones, not to speak about Chinese ready-to-dump craps.
Keeping old but working caps also looks better in vintage machines.
Do you mean "spraguemex" capacitors? made in mexico? I have a 60's stereo amp with those caps, are they good?
Or Siemens electrolytic caps, I have some huge ones from drum printers that still work.
those caps deserve a raise and a promotion. They're dedicated.
But it’s a Mallory capacitor... they’re bulletproof!
Seriously though, I replace capacitors in 30 year old Macs because they are cheap.
Given the historical significance of this teletype terminal, I would only replace those items that have been tested and found to be faulty or dangerous.
“If it ain’t broke …”
MatchstalkMan but why not use the same logic with macs... I do.... I only replace what needs to go
Plus we had the infamous Capacitor Plague which many people remember because computers built with those improperly built parts would fail, and essentially the entire production runs from those manufacturers need to be replaced. That sustains some of the belief in wholesale capacitor replacement, broadening suspicion from the era of crap parts, 1999 to 2007, to every capacitor ever made.
Those Macs use SMD caps, which in general are the most likely to fail. The really big electrolytics are usually fine.
I only replace the dead ones... no use heating up a board with bad traces on... my methods always work the best :)
If I was going to replace the caps in this thing, I'd probably open up the Mallory caps and turn them into a shell using a smaller (physically) modern cap inside.
Hi! I LOVE your channel! In this video, you mentioned a capacitor that exploded, and nearly caught-fire inside an NAD amplifier. I was a bench tech for years in a shop that sold those. NAD issued a service bulletin, including a fix, for that problem. They'll probably pay you for your time and cost of parts. :) If memory serves, that capacitor has reversed voltage across it at power-up for some reason, and will fail again if the fix isn't done.
Very interesting story! Failure analysis is a fascinating science, there is always a cause...
Hi Marc, I totally agree with what you're saying in this video, personally the only time I would ever seriously consider replacing a capacitor is if it shows a significant-enough amount of leakage of maybe a milliamp or something like that, but not something like 560 micro-amps as you're seeing with those 1953 vintage caps which were obviously very well-made.
Thank you for this confirmation, Because I keep hearing caps last around 20 years but I have equipment that's running for 40 + years just fine. But I have a tv that blew two caps after one year. So I don't want to replacing old caps that are working for newer caps what will not last two years.
I enjoy computing history, and love understanding how we got here, so a few years back I started that journey.
I started with the machines which made a large impact on my educational experience, so the first machine I acquired as a restore project was a VT100.
Most of the components checked out fine. There were a few tricky caps that were bulging from the bottom in the power supply, tricky Dicks!
In the end I replaced them all anyway.
In fact the entire power supply board is pretty much rebuilt 90% of the components.
I have to go through and replace more components on the video board resistors/etc. Which I am planning on doing this year sometime.
My reason is the same reason I replace the items on my car.
I don't wait for a tire to fail, or the battery to die, or the belts to snap before I replace them when I know, they are 3-4 years old and have a lot of wear and tear on them.
For fairly vintage equipment, I don't want to risk damage when some of the components on my VT100 for example, I am not sure if I could get them replaced without scavenging.
Just my two cents.
The components are part of the historical context of when that particular device was made. I only replace failed components to preserve the originality. Replacing resistors as "routine maintenance" is truly crazy. By the 1970s, industry was using metal film and metal oxide resistors which are highly stable, and are the same type of resistors used today.
@@gregorymalchuk272 I actually use my museum pieces for fun and work, so I am more concerned over failure than your ordinary collector. Preventive maintenance is imperative.
Last year they were fine, but they just reached 65 years old, which over here is retirement age. That means no more full-time work for them!
That's a disturbing and insightful comment. I'm not near retirement age yet but sometimes I listen to and watch the ever rising tide of rote bullshit I start to think to myself maybe I need to start thinking about putting my feet up voluntarily.
That was my first thought because I was "made" in 1953 and will be 65 in 4 weeks time!
@@vincei4252 Vince, forget about putting your feet up, just bring yourself up on a variac, and reform! You'll be right up to par in no time! 👍👍👍
@@Mikael5732 Funny you should say that, I've put off buying a variac for years, thanks to Marc I ordered one last week!
My only objection to the earlier video was the interpretation fo the results from the Heathkit capacitor tester. Due to the range selection, you were testing at only 75% of the rated voltage, and the magic eye only opened to a sliver of a gap. I’m no expert, but I have two of the C-3s and two IT-28s. When testing a good cap at it’s full rated voltage (where possible) I typically see the window open nearly instantly to it’s full width. I would have judged that cap as marginal or “maybe good enough” at best. The test you did in this video shows that the leakage is improving with time at its operating voltage. If it took over 60 years to reach that level of leakage, I suppose it will last awhile longer.
As a bit of a novice, it does make me think when experts such as yourself and your team have opinions that are diametrically opposed to others, Mr. Carlson’s Lab, for example. Thinking is a good thing, of course. Before I saw these videos, I was leaning towards the “just replace them all” camp, but now I’m not so sure. When I saw Mr. Carlson’s explanation, I was wondering why he was concerned that a fused input transformer would be damaged by a filter cap drawing excessive current. Maybe we can’t always rely on the fussing being properly designed, especially in consumer products. Looking at your transformer, it looks like you’d be safe. I have seen (I can’t remember where) demonstrations of coupling caps causing tubes to run hot, but that’s not applicable in a power supply.
Great video series, by the way. I actually saw much later versions of the teletype in my USMC 5938 training in the early 1980s. Our equipment was a weird mix of things from the 1950s and state of the art digital automatic weather stations and satellite receivers. Thanks for your, and your teams, efforts. I am generally amazed. It’s great there are people putting in the effort to preserve this kind of history.
That's a good way to put it. We thought it was a "marginal good enough" cap, which would be fine in a filtering application. And probably not for a coupling cap, as Carl recently pointed out to me. I like following Mr. Carlson (should we call it Professor Carlson?). In general he does older stuff than us, with much higher voltages and consumer paper caps and all, so I am not sure that his experience directly translates to our more recent and professional stuff.
CuriousMarc this is the kind of conversation i sit on the side at nodding knowingly while understanding maybe 15%.
"Maybe we can’t always rely on the fussing being properly designed, especially in consumer products"
Fusing exists to protect wires, not the transformer. You can easily destroy a winding without exceeding the mains current rating.
@@userPrehistoricman
That is not true in most cases. Cap-changers are the type of people who don't seem capable of doing proper diagnosis and repair if an individual component fails.
@@gregorymalchuk272 Thankfully i wasnt one of those cases, my Macintosh LC had electrolytic all over the board coming from ever capacitor. thankfully no traces were damaged. but that is the issue with people just replacing capacitors even if they are good, a lot of people would say to recap my dell dimension 4100 but im not going to because all of the capacitors are fine and people say the capacitors from 1999 to around 2007 are always bad, my dell dimension 9100 and 4100 are both fine just the 9100 has 1 bulging cap but what do you expect it was running for probably 3 years straight
These capacitors have lasted 65 years and you have proved they are still working well, it would be crazy to replace them with new capacitors with an unknown lifespan.
Very well put!!
The rule I use is, if it's encased in wax it should go, though if I check it and it has no voltage across it I will probably not replace it, but wax coated paper/electrolytic caps are nearly always bad, if not they will definitely go bad eventually. The reason is because the wax let's moisture in and will degrade the electrolytic, the paper also goes acidic.
I just converted a tube radio into an AM transmitter, since somebody already stripped half of it (lytics included) I didn't have any filter caps to use only a 50 year old elwa branded filter, I installed it and the transmitter is working fine to this day. I only change old filters like that if they are damaged on the outside too much, are dry or have noticeable faults. The capacitor itself reforms when in use so those are actually gonna get better with time.
That opening shot, the DC power looks super clean, I wouldn't mess with anything else in the think if it was running that beautifully.
I'll start with the most important point: It's your equipment, do what you want with it.
As I understand the issue, there were a few years in the late 1990's where a lot of low-quality "clone" capacitors were produced and used, so it is generally generally smart to replace capacitors of that vintage. These caps, on the other hand, seem to be from the era when stuff was genuinely made to last. The wiring also looks a lot more robust than the relatively fragile, paper-thin etching used on modern PCB's, so I doubt the cap would cause irreparable damage if it were to fail catastrophically or short out. This device is also designed to work at high voltages, so a short to ground wouldn't cause nearly the same catastrophic failures as if it were releasing 110/220v into a 5v digital circuit.
I see commenters more knowledgeable than myself discussing manufacturing differences and thermal properties, but I will refrain from discussing those aspects in detail for lack of knowledge on my part. If any of my assessment is wrong, I ask the experts to correct me, but you don't have to start a war over it; I am genuinely interested in learning more about electronics in general.
Great video - it really needs to be said. Since returning to electronics, this time as hobby, I've collected scores of vintage test equipment (largely HP & Keithley) and with a few exceptions most of the electrolytic caps are solid even after several decades. Since most of this used equipment cost thousands if not tens of thousands new the top of the line manufactures could afford specifying the best. Also the military was big customer as well further prompting higher quality and reliability standards. If I bought a consumer grade product of similar age I probably would be more suspect of their electrolytic's since they where not driven by the same expectations or objectives.
Thanks @Rk Str. You hit it on the nail on the head exactly.
As a fan of everything that's vintage I go for the philosophy: "Keep as many original part as you can, replace only broken or dangerous part if repair is not possible". And, although I don't know very much of electronics, I think you're right. I always enjoy your videos, because you make what you're doing easy to understand also for who is not an expert.
that is the hallmark of someone who has a dominating grasp of his field of expertise... the ability to explain to the layman.
@Lassi Kinnunen
Hogwash. The entire reason people collect old equipment is to preserve the historical context. The components used in the device are PART of the historical context. I only ever replace the ones that fail. Heck, by that logic, replace all capacitors, resistors, circuit boards, speakers, wires, CRT, case, and just call it a new device?
I don't know from replacing old caps or no... I just like what you gentlemen do! The first simple program I ever wrote (in high school), I uploaded to a mainframe to run and get output by keying it into an old teletype machine. It was one of the most amazing things I ever saw up to that point in my life! Great content!
You, sir, are living the dream! BTW, I also (still) own a model 15, which back in the day I had hooked it to my TRS-80 as a printer. A little z80 ascii to baudot assembly code never hurt anyone. Also noticed that the rolls of paper that used to be used for drying your hands fit in the machine as print paper although clearly not optimal. The cheap stuff wasn't very absorbant for it's intended purpose but worked for ink in the teletype. Oh, I used a model train transformer as the power supply to the selenoid, controlled by a transistor driven by ttl.
Thumbs up! When I repaired my HP 8671A and HP8566B and only replaced the failed capacitor. I got similar comments on why I didn't replace all the electrolytic caps. My philosophy has always been that unless there is a good reason (e.g. high ESR and low capacitance), leave them alone! New ones aren't always better!
plus, the risk of causing damage to the PCB traces when doing the rework.
Thank you so much for making this video. I repair amplifiers. Very rarely do I ever have to replace capacitors. The main issue I’ve seen is leaking caps from the 1990s. All capacitors from my vintage test gear and power supplies from the 1950s-1980s are still within spec. People need to lay off the snake oil and actually do some research and read. I especially love the one where caps in the rails or filter caps for power supply for car audio amps make a difference in the sound😂😂😂
If a capacitor is still within spec why risk damaging the board or pulling Vias because you listened to an ignorant person off the internet who doesn’t understand how to test capacitors. There is much much more to capacitors other than farad measurements.
Many folks don't know how to think. Some older electro.s do dry out, but you test them and confirm they are in rigor mortis; not wholesale replacement.
For me I ended up in studio/broadcast engineering and related electronics where the default of about any age from old tube radios I like to dabble with like Mr. Carlson (who seem to be against reforming) to modern solid state stuff seems to be recap and try to align is the first steps to bringing it around working logically of course going from the power supply out fixing as we go. It seems they can often be reformed and will work but if it does fail it's usually pretty spectacular and can take out other things not easily replaced on vintage gear even if it fails open you now have dirty supply voltage. I just recently redid the power supply in a 70s? mixer console for this reason.... when you can buy a smaller, better spec new one and have some peace of mind over "is that cap acting up and causing ghosts in this very expensive hard to get parts for thing has to be running 24/7 on the air? Is it going to blow up and fry it all? " Is it going to go out of spec and get me a nastygram from the FCC? For a part you can usually find new for not much and a little bit of shipping time if you live in the middle of nowhere like me. After having a big one go critical in a confined shop where everyone especially when the military trained engineers who taught me what I know including recap all the things hit the deck out of reflex has got me in the don't second guess just replace it camp..... SO... I am not arguing because I have learned every engineer has their own ways of things that they swear by usually a little different from others and won't budge on some things so this is only a story about how I formed mine and how it would go when we would butt heads. We could even be saying the same things just them in older terms/methods and me saying newer ones would end up in an hot argument to the point even the non technically minded at the station would be like "guys you are arguing both sides of your own arguments with each other again just pick one." It's really the only time we argued. Out of 4 of us in the area when I started I am the only one left not retired or passed on now. I am told being in my 40s makes me one of the younger out of the few left in my trade and have to go with what they taught me or I pick up on my own I never took courses on any of this or even went to high school.I learn by doing with my hands from them and that's how they do. What to do with old caps seems to be a spicy subject for pretty much everyone in our circles no matter which side you fall on.
To start, let me say don't let the haters get you down. Your channel, your methods, my viewing pleasure.
To be honest, I was surprised the old electrolytic capacitor survived as well as they did. And I'm most certainly guilty of replacing a lot of them. I do however test at least a few of the capacitors of every device I go through to make a proper evaluation.
In my opinion how bad they will be depends a lot on make, model, material quality and usage/load (did I get every variable there :)) But there are good generalizations. Paper wax for example is almost always bad, but even early metal foil types are mostly fine, unless their cases cracked or had bad sealing against moisture or such. Liquid filled electrolytic's that dripped and corroded your chassis are bad of course, but the more solid ones might be fine. The example vintage capacitor in this video was military spec, solid and had low frequency (and maybe quite constant) load in contrast to the bad ones you found in the alto power supply. So it survived well. I'm curious though what their spec MTBF was.
My guess the reason you see so few bad ones: Your repairing mostly high end expensive gear. Maybe that stuff was actually worth its price :)
cannot wait to see the next restore on the teletype
I appreciate your video. I have noticed that many vintage computer restorers reform the caps rather than replace them, unless they show signs that reforming was unsuccessful. I have always wondered about that. I have wanted to learn more about the failure mode of capacitors in general. I do a little bit of work on vintage radios and have seen many electrolytics go bad causing the familiar hum. I have also had them get hot at voltages lower than they normally run at. That was cause for concern. Perhaps if they had been properly reformed that wouldn't have been an issue. Not sure though. I have also taken some apart and have discovered badly corroded internals leading me to believe that vintage caps are always suspect. But, there was always those vintage computers that seem to work just fine with their original electrolytic caps. The only conclusion I can make is that those old caps are of superior quality, especially ones used in commercial equipment. It also appears that replacing the caps should be on a case by case basis. Given the low cost of caps for vintage radios and similar equipment, I may still replace them just to be sure but I can certainly see why you don't want to just replace them because they are old. Thanks for a very interesting and enlightening video.
Wiring harnesses that use waxed lacing cord are just works of art.
When capacitors are leaking, bulging, leaky, open etc, then they need replaced does not matter the age. I have replaced relatively new capacitors because they were bad, and have left old ones in the circuit because they were good. This is just part of troubleshooting. The reason that some recommend replacing old ones is because they are poor quality to begin with, as in consumer quality.
I've seen electrolytic cap failed short once. It was already repaired and assembled device which was turned on for final tests and I was lucky to watch whole process. Can't remember it precisely but I believe it was something around 200uF and 350V vintage cap from early - mid 70's. It was a filter cap on 250V rail. The only consequences was humming transformer for a short period of time and burned 10W resistor. Device worked fine after I changed resistor and cap.
Shorted tantalums are the best, instant firework display right in the comfort of your lab! :D
Also, an apt AvE-ism: "if it ain't broke, fix it 'till it is!"
I have caps in my 80’s guitar amp that still check good, so they stay. Why change them if they test good? Seems alot of the new ones are worse than the old stuff. More surface area = less heat and stress.
There will always be haters and/or people trying to appear smart.. I would recommend to try to ignore it, but on the other hand I do appreciate this extra video with more details on the details!
I work at a fusion energy research facility in the UK. I am literally surrounded by millions of capacitors of all different types and ages. We had one go pop last year.
I've seen intact capacitors in 1960s gear, but especially large, high capacity ones often fail by way of drying out (=below spec capacity, high ESR). It's my understanding that this is a self-accelerating process: Electrolyte escapes -> ESR rises, capacity sinks -> Lower capacity means larger charge angle and deltaU -> Larger and broader current spikes -> more internal heat -> more electrolyte escapes.
One particular issue is that it is difficult to source capacitors with equivalent ripple currents. Premium / Long-life non-miniature capacitors kinda match up, but it is doubtful whether they will last as long as the original, much larger capacitors, since the new capacitors are often stressed fairly closely to their limits, while the older ones had much more breathing room.
Some of them actually last longer than the new ones!
The people telling you to replace them don't know anything about what they're talking about half the time anyways.
What people also fail to realize is that vacuum tube equipment is far less sensitive to these things than digital electronics are - A television will usually work perfectly fine with some slightly off capacitors or dying resistors in circuit! Watch shango066's ressurection videos for proof of concept there. If it can sit outside in the California desert for 40 years and be brought back to life by cleaning a switch and changing two capacitors, then there's obviously some serious leeway in the component values!
Fully agree with you Marc. I'm a restorer myself and pro technologist...Carpet-Bombing with replacement caps (otherwise good resident caps) does more collateral damage that good. As a matter of facts, If it happens that see an add claiming that a unit have been "Fully Recapped/Restored"...I pass my turn as the reliability and ESD damage seriously compromise the longivity of it. Thanks.
I totally agree with the point of this video. It all goes to "if not broken, don't fix it". Some "experts" just blindly recap every piece of gear they touch, just because it is so, or just because they saw a video. Basically, instead of doing an analysis on what needs fixing, they just conclude "oh, it must be the caps, because they are old". And all they end up doing, in some cases, is replacing excellent vintage caps from reputable brands, by modern capxon or onehanglow crap. And, they also end up soldering them with blobs of solder, because the morer the betterer.
Many of the 1960's components were made extremely well, so it stands to reason that they would last a long time. I know we rarely change capacitors in vintage Marshall amps, since it effects the value of the amp and the old ones sound good.
I am with you that good electrolytics might last but I replace them a bit earlier. They have to perform well in their temperature range and up to their specified voltage. I usually don't want to bother testing a half decent cap (that shows alright at room temperature) at 70C and 400V so I'll just replace them. The 2-3$ are not worth the hassle. But I don't do historic restauration so the looks and authenticity is not important to me.
And, I don't want them leaking. Like, literally leaking electrolyte. PCB mounts are way worse than the standing cans seen in this video
Well put.
Apple put an "x" at the end of the model name to indicate a math processor added. "C" -> "CX", "F" ->"FX". So someone came up with the name "SE"...
This is very useful information, since I have piles and piles of old electronics laying around that I've been wanting to restore.
Thank you for a balanced review of old capacitors😊
In the interest of safety it is always wise, though maybe not absolutely necessary to change old electrolytic caps for new ones. Those old bulletproof Mallorys may be the one exception, though for how long? All vintage caps should be tested before being powered up after a long hiatus, and if they perform as the video shows then you can probably leave them in for a time, but they will fail eventually, even the Mallorys. As an example, I own a 1939 Hammarlund HQ-120x shortwave receiver that I inherited from my great grandfather. At his death in 1954 it was put in a closet where it sat for 25 years. In 1979, as a young electronics technician just starting his career, I unpacked it, put up an antenna and fired it up. It worked great!... for three days... then the main filter cap blew nearly taking out an irreplaceable 30 Henry choke with it. I don't recall the cap's manufacturer, it may have been a Mallory, it may not have been, but the moral of the story for me then and now is that it is safer and more cost effective in the long run to replace old electrolytics rather than leave them in and hope they will work for years to come.
Thanks for this! I hate seing categorically replacing all electrolytics in everything they work on and thinking that they are doing the equipment a favor. Electrolytics in older equipment are mostly all goood in my experience. If it lasted 30 or 70 years why change it? It's a reliable cap, and no one guarantees that the replacement one is going to last longer than the old one if left in.
Nice, you show us what today we have a lot of to discover, also the science.
Greetings from Rosario/Santa Fe/Argentina.
I agree. If an old electrolytic can reform and test with good capacitance and ESR why replace? I run alot of 1950's equipment here and like you said they rarely short go open and it's quite obvious when that occurs. However, the paper / wax / bumblebees are quite the different story. Don't fix what is not broken! Alot of ppl just bulk replace capacitors not knowing that the originals are still likely going to still exceed the lifespan of newer , contemporary components. Great video, thanks for posting it!
I've read somewhere that a lot of those electrolytic caps made back in the 70's were reckoned to have a maximum working-life of about 20 years before they could be expected to fail.
2:00 But.. "You can't trust everything you see on the Internet." (Leonardo DaVinci)
Sent from my smartphone (Leonardo Da Vinci)
That Edison socket fuse might be the same standard I saw in European house wiring of that vintage. If so the height of the bottom contact in the holder sets the maximum fuse rating, as higher rated fuses are shorter from top of threading to contact point. A same size successor standard (upgradable) had a ceramic ring around the bottom point with higher rated fuses having a larger tip diameter, and the threading being on a reusable porcelain holder that accepted the new bottle-shaped fuses.
That is an interesting tit-bit f information about the earlier fuse standard. I have seen them from a distance on some industrial equipment some decades back but they were rare already 40 years ago.
You can buy the small ceramic current selecting rings that go in the bottom of a bottle fuse holder at the hardware shop in Finland, if you change the purpose of a circuit and want a smaller fuse or replace with heavier wire for a larger current you might adjust the fuse holder in an old panel . All new build and larger renovation is done with modern circuit breakers but I have lived in a house with the bottle fuses and seen many others.
This video is very informative, cheers for putting it up. I always love watching your video :)
Well done, thanks for sharing. sure, if someone are servicing a switched PSU that are in daily use and replace all caps for 20$ as a precaution , it's a different scenario.
Glad I stumbled over your channel.
I repair vintage audio gear for a living. And as you might imagine, I get requests to recap gear all the time. I tell people that I will test the old caps, and if (a) they pass a visual inspection, and (b) they pass my leakage, capacity, and esr tests, then there's really no need to replace them unless they simply insist.
Very clarifying ! I really have never seen a Fuse like that ! Really awesome to see such rarities !
Anyway, Marc, I would love to say that I really enjoy watching your series of youtube videos, please keep them coming....cheers!!!!!!
My experience with decades old capacitors is that they are nearly always completely fine. If they are doing their job, I leave them be.
I'm glad I saw this. My nad 3130 amp sounds great yet ther is so much on how I should change the caps. I know elite caps aren't the best in the world but it's been looked after and dusted and hoovered regular and I keep thinking I'm happy with the sound so what would be the point? Bloody forums
@@CM-dw3gh I have a NAD 7130 tuner / amp that still works perfectly. The only thing I have done is add a squirt of contact cleaner to the controls.
@@Derundurel just checked you're model out. Couple if years older and still going strong which is reassuring. I'm guessing you have the green elite caps which a couple of beefy Black nippon chemi-con ones? Which to be fair can't be the worst out there, considering the later JH ones got a fair ear bashing
that's a fairly good capacitor for it's age, I just love that PSU, it's even beautiful working, never saw one like this, a work of art
Caps in consumer products tend to be the cheapest that could be sourced and why those tend to go bad with age. The "capacitor plague" didn't start till sometime in the 2000's when it became pretty difficult to not get defective electrolytic caps from most manufactures.
I used to use those vintage light bulb fuses in a 1970 trailer mobile home I lived in.
We used to have those fuses that fit into lamp sockets in the houses of the 1950s. Some folks would replace the fuse with a penny, which in house wiring was a very bad idea, as it could lead to house fires.
In an electrical circuit scourged by an old electrolytic, the transformer will blow to protect the fuse. :P
Nope, i don't see Apple logo on it... :P
Not if the fuse is sized correctly.
Hello Marc, I enjoy your videos, although the reason for this is quite sad. I’m pretty sure, that you guys know, what you are doing. The problem is, that your main audience is not aware of the robustness of vintage industrial equipment, which can’t be compared to modern consumer electronics. Please don’t care about the - sorry - stupid comments! Keep on going!
When deciding whether to replace a vintage capacitor, I mainly look at visual appearance, capacitance, capacitance frequency response, leakage and ESR. Last is the amount of work to get the job done nicely and the risk of not doing it. In many vintage repair job I did was to gut the original capacitor and fill the cleaned empty can with modern parts so they look the same after repair. I often found the inside of vintage cap almost bone dry, even with good capacitance and leakage numbers. Capacitor like this are associate with increased ESR and they are not going to last very long. Over the years capacitor technology improved a lot with thorough understanding of failure mechanism and reliability. We see unreliable results sometimes because we overlooked subtle circuit parameters or bought low quality/incorrect parts. Not all capacitors are created equal, today some parts are deliberately designed with shorter life span because of cost/size/temperature/ripple current .....reasons. I use better spec parts to make sure I do not need to repeat the same work again in the near future.
Capacitor technology improved, but due to cost cutting, some new electrolytic capacitors are just too unreliable, even if they are from reputable brands. It is not an accident that a new capacitor of a given capacity, voltage rating and temperature is way smaller than a vintage capacitor with the same specs. That is due to cost cutting, and not due to improvement. The new capacitors have much less headroom in terms of voltage.
I've been an electronics hobbyist for decades now (I'm 53), and I've repaired many vintage equipment over those years. But I've always checked electrolytic capacitors for ESR and capacitance, and only twice did I encounter failing ones (aluminum types that were cheap unknown brand in a linear PSU, and some low capacitance ones in a Roland/Boss guitar effect pedal). I've got many 30 and 40 years old circuits with high quality ones from known brands (Matsushita, Philips, Brel, Mallory just to name a few). So this "you MUST replace electrolytic capacitors if they're old" phobia, regardless of their condition, is also annoying me a little, too. Thanks for trying to clear this myth. I've had many more Tantalum caps go bad on me (some actually shorted out), than electrolytic ones.
it's been the opposite experience with me, but I do purchase my equipment with issues.
I had a 1964 Super Reverb that had never had much use, and the original high voltage supply capacitors reformed and tested fine, so I left them in the amp. Within a year of being back in regular use, the amp started to hum and then started to blow fuses. The lesson is that it's very difficult to take a snapshot of a component and then extrapolate from that how it will perform into the future unless you have extensive experience with that particular piece of equipment and the capacitor model and brand. Even then, your prediction may turn out to be wrong.
When people say they need to be replaced i think they mean the wax paper ones, those are almost always leaky, by leaky i dont mean like leaking a liquid or other substance from the capacitor but leaking dc voltage across them which they are ABSOLUTELY NOT SUPPOSED TO DO. They are supposed to only pass ac and block dc, if its clearly a waxed paper or bumblebee capacitor, IT HAS GOT TO GO, paper in oil are likely fine but waxed paper caps have got to go or be restuffed with new capacitors in the old casing.
I concur. Waxed caps and black beauty caps are a liability in high voltage tube era circuits.
To recap or not?
There's to many "if" to take in to consideration.
My case, I tend to follow the manufacturer instructions, by average this parts has a max 2 years shelf life and a life up to 10 years, I also consider de deterioration of the chemicals inside the component, contamination, etc. Since they are considered "wet" type capacitors, they also tend to dry out over time. Another thing that I take in to consideration is that this tests might indicate one thing, under load the characteristics might change.
There are some that can be treated/reconditioned and others that can't, it's up to the technician to decide. Is the client willing to pay for the job and the technician bench time? Does it worth it? Etc, etc.
I have to recap an amplifier from a wurlitzer jukebox , it was from the 60's, most of the cap was litterally disintegrated!
The reason why I replace a lot of electrolytic capacitors is convenience. If I have the device taken apart to fix a problem and find that some electrolytic (or film/paper) caps are bad, it is better for me to just replace them all right now instead of just replacing the bad ones, then some time later having to take the device apart again to replace some other failed capacitor.
Once I took a motherboard with a bunch of bulged capacitors to get it fixed and specifically asked them to replace all electrolytics there. They only replaced the bad ones and less than a year later the other ones failed and I had to replace them.
I also have a similar experience with some tape decks made in the USSR in the 80s - replace only the bad caps and you will be taking the device apart again to replace some other cap rather soon.
So, if the device does not have problems with capacitors, I leave them be, but if problems start because of the caps, I would rather replace them in one go rather than having to fix the device multiple times.
I often find the smaller older axial caps to be a problem. Lower voltage, particularly. On modern aged gear, the bigger caps usually don't fail as often, and they are usually much more expensive too.
One answer to this is experience , I have found generally most but not all vintage caps to be in okay or good order, it seems to be only the 1990 and a few years ahead that fail in the worst manner. I have been involved in electronics for over 40 years and have seen it all and repaired all sorts of weird stuff.
I think you're good on the cap decision (BSEE opinion here). I did enjoy seeing the thyratrons in operation, I have never seen one in the wild before.
What about the ESR (Equivalent Series Resistance) of such a capacitor? (Year of construction 1950 and older)
You get that "replace every old cap" mentality in radio restorations, TV repairs and old guitar amps too.
@John Smith The 'death cap' is a film capacitor in a specific location (namely between the AC input and chassis, which is why it's deadly when it fails shorted).
No sane repairer could conflate it with the other capacitors in the unit (certainly not with electrolytics).
I vaguely seem to recall that the dielectric of those old Mallory caps was oil based and not a water based electrolyte. They cannot go duff in that case...
Merci Marc. Quelle beauté que ces anciennes platines :)
Most of my instruments still have all the original caps. During Restoration I measure leakage, reform if needed, measure C an D. If all values are within specs stated by the datasheet, or if no data can be found, from my own lists it will be placed back in the instrument. If not good I replace it for a new high quality comparable cap. That is the same specs for WV, leakage, ripple current, D, temperature and hours of use. Often old good quality caps turn out to be better as good new caps.
This whole witch hunt is caused by people repairing consumer stuff without much knowledge and armed with their holy in circuit ESR indicators and the new trend, a LV leakage tester... I use several LCR meters and bridges from HP, IET, ESI and GR and 2 HV leakage testers (my own design) My best bridge is a GR-1520 ( with a famous history being the EX-Genrad standards lab bridge)
Anyone know of a video that explains reforming capacitors. I have heard the term many times but don't understand it.
Here is one I like. It's a little slow but accurate and thorough: ruclips.net/video/ifWfvaIbWZY/видео.html . Reforming simply consists in progressively applying voltage to the capacitor while monitoring the leakage current. Which, if all goes well, reforms the thin oxide on the electrode to its original condition and gets rid of the leakage. Not only valid for old capacitors, but new ones that are kept on the shelf for a few years without being used also benefit from reforming. However just putting them in use in a regular circuit without further precaution is usually enough to do that, so nobody thinks about it.
if it's not broken, don't fix it. especially when they're not cheap 90's caps from taiwan
Certanly no, they WILL fail ower time i can assure from my experience...
@@sheep1ewe The new replacements will fail sooner than these originals unless you buy the highest grade, custom made, custom specced Nichicon or something.
@@jan.tichavsky Yes, there are still a few good brand on the market, but it apends on the particular type (as You probably know), for example some types of old cheramics has no real experience date, they can still be good after many, many years, but sheap E-lytes usually last around 10 to 30 years still conciderable reliable. But in this case those bigger ones also represent a part of the historical value of the device it self as long as they work good enough for this purpose, as i see it, i think that was the point of testing and saving them. In equipment with no real historical value i see no problem replacing everything, but in very old apparates the component it self also hawe a historical value if it can be preserved without risking damage other, more valuable parts of affect the function badly on a device that are restored to be functional.
I found a video here made by an old man who saved the papershell with the original markings of old wax capacitors, but he putted modern ones inside the original shells and re-waxed the surface, it looked realy awsome at my opinion when it was finished.
(Sorry for my English, it's not my native language...)
@@jan.tichavsky Absolutely correct Jan!!!!
Could you make a video on the capacitors diagnostic?
Like is a component tester (that shows capacitance and ESR) good enough or do you need to power it to its working voltage and measure the leaking current or what?
great video. Is there any reason (I guess) why a teletype needs such a "high performance" power supply?
Best regards from Germany...
when it comes to caps I always test before replacement. For example when it came to restoring my 1965 IBM 1800 i tested all the caps and all were fine, some were boardeline but its still on its 1960s filters and its working perfectly although i have a memory circuit error meaning i can only access 28 of the 32kw of core.
In my experience with tube equipment, about 50% of electrolytics need replacement.
Many Philips 1950s caps are totally fine. I have some radios which have the original electrolytics (not the original paper caps!) and have been in use for thousands of hours without failure.
I have just as many radios that have new electrolytics, but those did not respond to a reforming voltage, were physically leaky or had too low capacitance.
Always measure leakage at working voltage. No significant leakage after reforming for a good amount of time? Capacity okay? Leave it in!
You could drop that power supply out a second story window and the only damage would be to the pavement. My dad was a radioman shipboard in the Navy during world war II and he used to collect these old military power supplies and other equipment and one of the reasons that these 70 yr. old caps are still good is that the military only used Mil Spec parts. They demanded reliability in the components they bought and then took those they received and pushed them to way over their rated specs in house and if a cap could survive that torture test they were placed in service. Cap failure is tough thing to predict. You can use a tester like the Heathkit to weed out failed caps or ones on the way toward failure but you can't predict when a good one will fail. There is no crystal ball on that tester that will give you that info. So automatically replacing all the caps on old electronics is a judgement call.
Thank you for the much better and more detailed testing of the electrolytics. The results from the heathkit tester where not that conclusive. I did think they looked a little new for 1930's. Being from 1953 the paper inside the capacitor would have been impregnated with PCB oil, those caps do last much longer (PCB oil kept being used in big electrolytics until the late 60's) 1930's caps are almost always dried out. As for the 80's caps in the SE/30 Surface Mount caps of the 80's where at about the same point as standard caps in the 20's. In the 90's there was a flood of fake capacitors and a fake electrolyte formula floating around, Panasonic and Elna caps where effected the most.
Replacing bad caps is a good thing, but replacing those awesome caps would be a crime!
Most of the failed vintage capacitors I have seen were melted wax capacitors. For me, its been a 50/50 shot if the metal sealed electrolytic capacitors are OK. I like to test them first, though, when I get my hands on something vintage just to be sure they are OK :).
Question since you're an expert.. If I own high end audio equipment from Accuphase, dCS, and McIntosh and I want it to last as long as possible... is it better to leave it on all the time? or turn it off when not in use ... in terms of keeping the capacitors and other components health inside.. and last longest?
I recently picked up a 1934 Telefunken radio. It looked like it had some work done back in the late 1940s, but the cathode bypass electrolytic capacitor was still original. I was amazed that it read in spec, both on capacity and on ESR. Also the leakage at the rated voltage was acceptable, at least for a device that was probably last used in mid 1960s. However, paper capacitors are all shot, some measured as low as 2k Ohms, this is almost a short on a 280V DC rail, and would quickly take out the very expensive RGN1064 rectifier.
The Germans made great tubes and caps. I collect telefunken and Grundig radios and have never changed a cap.
Merci Curiosmarc pour cette video pedagogique, je n'avais jamais compris comment tester la fuite d'une capa sans testeur rlc. thanks
Petite question de debutant, j'imagine que par precaution, si l'on est pas sur du condo, au moment de la mst, on regle la regulation courant assez bas pour eviter d'eventuels court-circuit fatal genre electroboum et electrolyte partout?
en fonction de la capa quelle est la fourchette admissible de courant de fuite ?
Merci
I had a late 1980's TV explode an Electrolytic in my face in the early 1990's while I was looking for a completely unrelated fault. Yes the boiling electrolyte made it explode. Old fashioned capacitors didn't have this issue. If they are good then they are more than good enough to keep as is. This was way before the capacitor plague.
Most of my capacitor failures tend to be 1uf 50V types and there about's. The bigger ones seem to be more reliable.
My biggest problems come from modern surface mount caps that leak all over the PCB and ofter cause unrepairable damage to the board - I mean how on earth can you clear away electrolyte that has permeated into the middle of a PCB.
Fix a 1950's power supply - quite possible, Fix an 8mm camcorder with leaky SMD's - forget it!
I think a lot of these people advocating changing certain caps out don't really understand what causes a capacitor to develop internal leakage, they assume that the cap naturally does it itself, but for those who don't understand, a capacitor is made-up of two conductors that are separated by an insulator, or dielectric, normally the insulator or dielectric presents a very large resistance to D.C., the insulator/dielectric has a certain voltage-rating it can withstand before it breaks-down and starts to conduct, most good insulators have a D.C. resistance of at least a few giga-ohms (well at least an ideal insulator should anyway) a giga-ohm is one thousand mega-ohms, normally you would design an electronic circuit so that the expected voltages in the circuit do not exceed the maximum working voltage of the capacitor, or, as I personally like to do, you would use capacitors with a working-voltage greater than what you would expect to see when the circuit is operating normally, so, you wouldn't use a 25V electrolytic cap in a circuit where you'd be expecting something like 40V max, the 25V cap would go "BANG" and let-out the "magic-smoke" as well as a particularly pungent smell....Tantalums are notorious for doing that, and believe me it's not pleasant, I know from first-hand experience.
Very good.
Altho my beef is with some more modern caps. Got hold of a nad c352 amp. It performed really bad so I checked out the caps. Despite the amp being around 15 years old, many of the caps were visibly failing.
The cheapo caps I replaced with mainly Panasonic and it's amazing now.
Just noticed you had a nad cap gone bad also.
The two main caps are still original and are starting to bulge! They are 120k uf and will cost a fortune to replace.
In vintage gear I've come across several failed caps, and replacing them did indeed fix the faults. But a lot of the time caps seem ok and I'm sure they were made better back in the day 🙂
Yes, I also seem to have more problems with modern caps, and ran into the same NAD issue: ruclips.net/video/kTYlGG9CdtA/видео.html !
A new fly on the wall, I'll stay clear of the electrobooms. I deeply respect your manner.
the only time it makes sense to recap is in 1980+ Designs where the people started penny pinching in Homecomputers , causing some parts to be stressed more and those capacitors tempt to fail (looking at you Amiga 500) , otherwise its a waste of time and money
Marc isn't refusing to change them. He changes them if needed, but only if needed.
In old tube gear, the reason leakage is dangerous concerns circuits where the capacitor is a DC block and leakage would introduce bias on the grid of the next tube in a chain. The bias changes the operating point and can overstress the tube by causing excess conduction - due to the DC voltage on the grid. In this case, leakage is bad and has to be very low to be tolerable.
If a circuit would be damaged by exceeding some current, leakage might be a problem but the design of the circuit sets the bar for when leakage becomes harmful. A power supply should be designed so that even a dead short across the output will be handled - perhaps by a fuse (and maybe a crowbar device) - thus the filter caps aren't a risk. Their failure impact would be inadequate filtering, excess ripple, rather than excess current.
It depends on what the capacitor is doing and the circuit in which it is used, in other words.
In the real world, every capacitor also has leakage, series resistance, and inductance. The specifications for new capacitors include that information, allowing a designer to pick a very low leakage if that is critical for the circuit.