In German the word for rectifier is Gleichrichter We called these "gleich riecht er" which translates to "he will smell soon" Especially useful as a shout out to someone cranking up some equipment beyond its limits. And everyone was aware of the consequences of one of these releasing the magic smoke But I prefer rectumfire nowadays if I see one of these nasty things
My first contact with a (single) package selenium rectifier was as a school boy. If I recall it was rated 250 V, 50 or 100 mA, intended for center tapped transformer. Later on, at my first long term work, we used International Rectifier selenium stacks to clamp overvoltage spikes. Those parts we assembled ourselves from IR plates, 1 x 1” or 3 x 3” sizes. That was the time when I learned that selenium can produce a terrible stink. Finally and even later, I was told thatMiller Electric built their own selenium rectifiers for some of their welders still when the first, barely high enough rated current silicon components became available. I understood that Miller experienced voltage spike failures on the silicon diodes that they had tested. The selenium parts were “more robust”, until they released their stinking smoke. The robustness was a side effect of the leakage as well as the probably 6” square plates of the selenium assemblies. By the way, at one point I also found that the phone company installed small selenium spike clamping diodes to their phones, when some users complained about the excessive scratch noises (or in some cases intentional teasing by young pranksters). I then tested whether two anti parallel silicon diodes could do the same and found that silicon had too high a forward voltage to compete with selenium in this application.
First time I have seen the curve of a selenium rectifier. The reverse curve was most surprising. The capacitance is expected with such large parallel plates but the crappy breakdown threshold is the surprise. These get quite warm in normal operation at their rated voltage so not only are there forward losses as expected, but also reverse losses that are noticeable.
Its great that you watch this channel! I watch you all the time! Two of my favorite creators in one place! 🤓 And YES the reverse curve was very interesting! NOT what I expected at all
Well Ron in our younger years we had a lot to deal with these. Per plate they are rated for only 20V in reverse. That's why you need at least 10 for a rectifier in a tube radio. Surprised that you were surprised 😆
Thank you so much for taking one apart! I was yelling at the screen, just then you did! And watching that curve tracer with the Selenium Rectifier was more entertaining than watch HBO.
Woohoo! Some old Sarkes Tarzian selenium rectumfriers! You know why they call them rectumfriers? Well, if you run too much current through them, the hydrogen sulfide will off-gas, and the surrounding area will smell like a......rectum! In the old Zenith AM/FM radios from the 50s, they used a 22 ohm in series with the selenium; these can be replaced with a 1N4001 and a 47 ohm. Basically, you adjust the silicon diode series resistance to keep the B+ the same as it was for selenium.
As far as I remember those selenium diode stacks were built for about 20 V per stage and a forward current according to the size of the cooling fin. Such a size could be for around 1 A or little less. I remember two bridge rectifiers I bought to play with 55 years ago which were for 1 A ans 2 A and 20 V. The ones in tube radios had quite a number of stages to reach about 200V. They also had much smaller plates for lower current.
Selenium diodes were just barely better than having no rectifier at all. Worked okay to provide low current high voltage supplies in valve equipment but are totally unsuitable for signal detection or anything that requires fast switching and low capacitance.
Interesting. I was trying to figure out why the curve tracer started to slide away from 0V (and looped a little) as you increased the maximum forward voltage. It makes more sense now when you looked at the reverse characteristics and the huge loop... explaining it to be due to capacitance. So even though the drive voltage reaches 0V, the "capacitor" doesn't have time to discharge to 0V before the drive voltage increases again. Makes sense, but then the I-V curve would differ dramatically based on the frequency, yeah? The loop also reminds me of the hysteresis curves seen in magnetic core memory... but then as I understand, that's not a capacitive effect, but a magnetic effect. It seems a major part of understanding I-V curves is understanding the physics behind the part!
I have an old 6 / 12v car battery charger that's full of selenium rectifiers (basically just those and a transformer). I'm sure it's 40 or more years old, but it;s the best charger I have for resurrecting older or completely discharged batteries!
Yes, it's true. The explanation is that there's some reverse current, so some AC on top of the DC. The small amount of AC "breaks through" sulfation on the lead plates and helps a little with liberating the gas bubbles at the plate/electrolyte surface. There were devices sold for anti-sulfation or battery rejuvenation that emulated the output by injecting some AC onto the output of a DC charger to recreate the effect once silicon rectified chargers became the norm.
Me too - the brand name is James Grose and my Granddad got it by collecting cigarette cards. it's got an ammeter and some sort of thermal cutout... and that's it. brutal yet effective I guess.
@@joinedupjon That's it! The bimetal strip pushes a button out if things get too hot. There's no true ammeter on mine - just the charging current - an actual ammeter with negative readings would have been handy. There's no brand name left on mine, as it was rescued from a damp barn about 6 years ago, when it has stopped working, after servicing tractors and other farm vehicles for decades. Most of the paint was gone from rust, and the front panel controls were just about functioning and legible. The only fault it had was one that the bolts - not solder - holding the charging wires had rusted so badly there was no connection. Have it a good clean with fresh bolts and it has worked perfectly since then.
Thank you for this! I noticed the capacitance aspect of selenium rectifiers when replacing a selenium bridge rectifier on an old German radio. It seems that after installing silicone diodes I was getting a lot of oscillations until I added one small ceramic capacitor across each diode of the bridge. I later found a post somewhere in German that recommended the same thing...
Ones I dismantled were made of aluminium. They had some bismuth plating on one side, then some dark selenium on it, and then this tin-cadmium alloy on top. Layers on top had slightly smaller surface area than ones below, they didn't go all the way to cover them, so there was a dark "frame" around the tin-cadmium layer. The entire plate surface is a diode, and a capacitor. In vacuum tube devices I've seen there were multiple of those plates stacked for more voltage. Some had holes and screws, others were inside a plastic box with some springy inserts on both ends of the stack, and didn't have holes in them.
6:26 you could even wire all three terminals to that component tester and it should tell you that it is a dual diode. that way you could find out whether they are common anode or cathode if you were unsure.
I can smell this video. Remember when some really large half bridges blew up in a forklift battery charger and the smell was God awful 🤢🤮😂 Ended up replacing them with some heavy duty alternator diodes.❤
The reason stacks of these are needed for tube radio HT circuits is that a selenium can only accept a reverse voltage of around 18 volts, so you need several in series to cope with the high reverse voltage. When conducting (forward voltage), each junction may only drop around 0.3 volts, but as there's a stack of them, several volts will be dropped across it. So if a silicon diode was just slapped in instead, the HT voltage to the radio would probably be around 15v too high. So a beefy (maybe 10W or so) resistor needs to be added, to drop that sort of voltage. It might be around 200 to 300 Ohms. But the current is only milliamps of course.
As others have mentioned, selenium is photosensitive, so perhaps those thick layers of paint are to prevent them acting as tiny little solar powered generators. Big Clive also has a nice video on these.
As a kid I remember shorting one out and the smell was horrendous (selenium hydride, etc) and extremely toxic too (cadmium ain't good for you either). Interesting video, I took one apart and didn't realize the actual diode was a thin layer on the plates. I would wear gloves if taking these apart.
The old transformers for my 1960ish Triang electric trains have this style selenium rectifiers in them. They handle 15V 1A max so I've been thinking of replacing them with a bridge rectifier or maybe just some diodes. Thanks for sharing!
In old radios and very small TV sets we replaced a Selenium rectifier with a silicon rectifier and a 10. Ω at 10 Watts resistor. (called a "sand Ohm" resistor) Ron W4BIN
A while back I helped a friend retrofit a modern motor controller into a 1950s electric forklift. The selenium parts that came out were the size of a loaf of bread.
Nasty smell when they overheat. I believe they are somewhat toxic also when they outgas. Definitely a type of rectifier that hardly is useful anymore. Whenever I find these in old radios or old TVs I replace them with silicon diodes with a series resistor.
Selenium rectifiers were in use for a long time in small scale model locomotives. Just the blank disc with no further protective coating. Especially Fleischmann seemed to have a huge stock of these. I'm not sure if there is one or the other model remaining where these still are used.
I've taken these out of electronics that were in use since the 50s and 60s and 70s, they are very smelly even when they don't go bad, and I've seen them in all sorts of shapes and sizes, I wish I hadn't thrown them all out a year ago in a move, I liked to keep them in a 10 gallon bucket next to my work bench since they interested me with their shapes and sizes and rarity
I have seen them used mainly in low voltage dc supplies. (Battery chargers, etc.) Standard balanced transformer outputs to the anodes, and get dc from the transformer center tap and the common cathode.
I run across rectifiers like that in tube amps, but not for the B+ which is still tube, but rather for the bias supply and DC filament for phono preamp tubes. Sherwood used this design a lot.
Fisher tube high fi used them in the bias circuit as well, or perhaps it was Scott ? The rectifiers were a shiny flat pack and made in Germany by seamans.
trying to understand the various shapes of component curves, when there is a circle like I-V on a device that is not considered a capacitor,is this more of a electrical defect.or a physical characteristic of the part being measured? thanks John
A forward voltage of around 0,3V indicates that this might more likely be a Germanium diode.. Selenium would be in the 0,8-1V range per platter, as far as I recall.
This is the problem with replacing selenium rectifiers.. You certainly can use a diode + resistor combo, but if the original rectifier had any significant power rating, you end up needing a large, hi-power resistor to replace it. This is why the selenium parts usually have such large cooling fins - they were terribly inefficient, producing large voltage drops and making a great deal more heat than a modern silicon diode by comparison.
I have a RCA voltohmyst that had a Selenium in it. I took it out and replaced it with a 1n4007. I forgot what value resistor I used but it's a 2watt metal oxide.
I cannot tell where your probes are connected; however, metal semiconductor junctions are Schottkey diodes. To get ohmic contacts it is necessary to overdope the semiconductor under the metallization.
very interesting video - thanks! Lead sulphide (galena) was used in crystal radios, a low-signal application, and the Vf was 0.37V. I would be interested to know what the forward characteristics of a copper-copper oxide rectifier is, if you have looked at them?
I have an old, very large, HF linear amplifier that had 2 sets of selenium TVS diodes across the primary of the plate transformer. I didn't even dare power that thing up with those in place. Good thing too, one of them was shorted. They stink really bad when they go out... and they WILL go out. Producing hydrogen selenide gas. Toxic stuff. Had no idea on suitable modern replacements as I can't find specs on these ones.
The power this should be able to handle is about couple of W, so with 5 Ohm and 0.3V initial drop we will be talking about roughly 0.5Arms rating, ballpark. In real application virtually the only limit is the resulting operating temperature resulting from the power dissipation and the cooling.
Technology is ever changing and in regards to these old rectifiers I say throw them out. Replacing one found in an old tube radio from the 1950s , lets say, is so easy just by using a silicon diode which are cheaper, smaller and way more reliable. Put a resistor (of pertinent value for the tubes) in series with the silicon one and be done with it. It is a no brainer for this young man's vintage radio restoration efforts. Just like a 50 year old car tire ( never used), the tire and the selenium rectifier are both are bad now. Let go of the past.
@@IMSAIGuy Thanks! Looks interesting, lots of fun optoelectronics. Do you have some go to pages to find interesting stuff? Im in EU and ironically the strict recycling legislation makes it super hard to find second hand testing equipment, cool components etc
1) IIUC, lots of folks think these are dangerous and are a replace-on-sight component. 2) If you wanted to build a Voltage Controlled Oscillator from a pair of back-to-back diodes, would a weak knee be advantageous, because the linear area of operation is extended over a greater slope? I am just a hobbyist and I am talking way out of my league, here... Thanks.
Something that only us old men (I'm 75) can, but would rather not remember. The SMELL of a shorted hi voltage electrolytic filter cap, and the SMELL of a failing/ shorted selenium rectifier. In the 70's-80's I worked at a big auto parts store which also had a separate welding store. I used to repair the older unregulated battery charges which had only a selenium, nothing else--with a pair of silicone stud rectifiers out of one of the smaller Miller welders. Made a small heat sink out of a scrap of aluminum sheet, and off we go.
I'm a bit younger than you ( not so much ), and i remember times in which me and my friends tried some tests on electronics ... i got an old television to dismontle, it had a big selenium rectifier inside. We would obtain DC current from the mains, my friend was convinced to connect the selenium rectifier in a ( for me ) erratic manner, so i said ... "ok, if you are sure, let's try ... " ... Soon smoke arised from the rectifier, and at once on orrible smell was around. Open the window was just a minor solution, we ( three friends ) had to escape from my room and also close the door ... I can't forget THAT smell, even today ... ( i also believe was tossic ... )
The point of the full wave rectifier is to get both swings of the ac voltage, from both power legs. You can get DC from AC with only one rectifier, but it will be very choppy.
@@tableseven8133, true, but you still get full wave from the CT xformer and two diodes. That was a cost trade when diodes and tubes were more expensive than copper windings.
@@tableseven8133 I have my amplifier build that way. Transformer is 12-0-12 and 550va. When i use one schotky-diode for each end. i can get 100hz pulsing dc that is dropped unloaded voltages only one diode forvard voltage drop per wave. if i have used typical full bridge rectifier, there are always two diodes series, when wave is present and voltages and current are generated for the load. Class A operated amplifier with pi filter and that transformer diode combination give me useful output voltage about 14.2volt +-0.3. Before filter, voltage is about 15v, and biased constant current is about 14A so big load all the time, when powered on. But it is so good the sound quality. And if you want to know, speaker output terminal ripple for idle with this configuration is about 4mv pp, which is not so much at least my setup for not hyper sensitive speakers. Still, about nominal 10W for 2 ohm and stable to support speaker dip to low as 0.5 ohm gives a good umph for the many songs. Here is a link for that full wave two diode rectified circuit. madpcb.com/wp-content/uploads/2020/11/Center-tapped-Full-Wave-Rectifier.jpg
1:30 "What's in side this thing?" A heinous, infernal stink. Had one go on a 600A welding machine years ago. 600A three phase from the mag-amp controlled transformer. Big stacks. Oh My God. It cleared the shop out for the day. I could smell it for weeks, at work or not. It was a couple months before visitors came in and didn't mention the stink. In a welding shop.
I still remember it with the occasional flashback. Welder still rand pretty smooth. These have such a soft characteristic, loss of 1/6 didn't hurt it much. Fantastic arc gouging source, and great for running ceramic heater blankets. As the only real engineer, I was supposed to fix it. I was NOT going to go in to that case. Ran it for a couple days at low load to get the stink out, then felt good tack welding the case shut. Probably still in use. @@mfbfreak
A nice look into selenium rectifiers - I wouldn't discombobulate one without gloves though. Still better than letting it burn, emitting toxic hydrogen selenide.
Every tube radio that I have ever seen that uses a selenium rectifier used it as a half wave rectifier.... It's pretty common to replace them with a standard 1n400x diode and a resistor. To get the right resistance it's kinda guess and check by comparing the B+ voltage to the schematic. A 3-5 watt resistor is usually adequate for an AC mains radio, but a 10-20 watt resistor may be needed on an AC/Battery radio. The antique radio restoration community is pretty split about replacing selenium rectifiers. Some swear they are fine to leave if working, some swear they are a failure waiting to happen, and some worry about the cadmium and want them gone working or not. Most that I have seen have crept up in resistance to where they need replaced because the B+ voltage is too low.
Interesting; the selenium rectifiers in tube radios I've stumbled across were all FULL BRIDGE RECTIFIERs and they were mostly encased in a metal package, making it easy to replace the innards with four 1N4007 diodes.
@@KeritechElectronics Depends on the radio, if it was a drop in for a rectifier tube a half bridge, like here, was common, as the transformer was already there, and wound for a half bridge with common cathode and 2 anodes. For HV the stack would have to be 20 plates, though they often stacked them close together, so they fitted roughly in the volume of the old tube, to get a 400V reverse withstand, to allow B+ and peak reverse voltage to not break them down. Only after the transformer was redesigned, old stock run out, did they switch to the full bridge, as it used the same number of plates, but cut the cost of copper down. Full wave always has a jumper wire, as the stack is arranged to have a common positive point, and 2 series diodes to that point. Then connect the outer plates, as those are the negative, to make the bridge. Higher the current larger the plates. Replaced a selenium rectifier in my battery charger, after 50 years those diodes were rather tired, and I did the change, using a 1970's era silicon block almost half the size of the old ones, that was rated for 10A instead of the 5A peak of the old ones. Did also add in a charge controller with a SCR, and a festoon lamp to illuminate the dial, plus reverse voltage connection protection. Now a 70 year old GE charger, still working well as needed.
They were' Extensively used on British motorbike 12volt systems as a packaged bridge rectifier on single phase alternator, the soft characteristics are well suited to the zener regulated battery charging systems.
Yes, even in the 70's, as im currently restoring my dad's Honda 750. They are said to be indestructible, but inefficient. Since Honda's have little knack of producing 'just enough' electric power, im kinda thinking about swapping it for a silicon rectifier. As I already did with my 60's Honda. On the other hand, keeping it original is kinda my goal, maybe if I notice that it's going to lack power ill swap it...
@@sciencetestsubject CdSe is also used I think (wikipedia has a link to a Silonex datasheet listing it). I've only ever seen CdS ones personally though.
The LDR used for the tremolo circuit in vintage blackface Fender guitar amps has a cadmium cell in it. They still use it in some of the US market amps but not for the amps designed for export because the EU won't allow the cadmium content.
I have never seen a selenium rectifier used as signal device. They universaly used as power supply rectifiers. They were lossy and heated up a bit. When the filter cap shorted out the selenium rectifier would go into thermal death. They smelled horriblr when that happened. And you touched the fins, you get a nasty burn that take a long time to heal.. Don't ask me how I know this. For a higher voltage rectifier they just added more layers. In the military we had a full wave bridge rectifier in a single, it had several terminals sticking out and jumpers tying them together into a bridge configuration. I enjoy your videos and working on vintage main frame equipment.
Old tube radios tend to use a pair of diodes with a center tapped transformer, not a full bridge rectifier, so your double diode would be perfect. There were double diode tubes as well. Tube filaments do eat a lot of power, but the plate supply is not so power hungry.
Your should try to release the magical smell. Exceeding the maximum current melts and burns the thin selenium layer giving up a smelll ranging from horseradish to fart. On a second note well used them as solar cell. For this the cover electrode usually tin has to be melted and wiped away to allow light to enter. Wasn't working well but after a few attempts over a gas flame with a rag in the other hand you could quickly wipe the molten tin layer off. By leaving on the contact spring washer your retained the contact and viola got half a volt in bright sunlight.
Oh my God. You have found a better replacement for the 1N34A diode!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Well done. You made a good example of sounding like someone who doesn't really know what he's talking about. I laughed when you needed a calculator to work out ,2 divided by 40. Did you go to school?😥😢
The long stacks were there for high voltage capability, to rectify 2-300V for the tube anodes. Not a lot of current needed there. Filaments (where high amperages were required) were driven with AC. They came in dual packs like that because they wanted to avoid the double voltage drop of a 4-diode rectifier, and used a center-tapped transformer with two diodes to do full-wave rectification.
I remember stripping them out of old TV's, stripping the paint off of them and using them for solar cells. I haven't seen or thought about them for decades.
Yes got some nos ones as well. You really want a funny one look for copper oxide rectifiers, which you can actually make out of copper sheet and a flame.
Honda motorcycles used a stack of these under their seats next to the battery. I don't think they would work as a substitute for a Germanium diode in a crystal radio.
Selenium rectifiers as you said would be high current devices is not true though, They are typically low current rated in the milliamps range. Tube amplifiers usually operate at high voltage low current and the average sheet of Selenium rectifier can handle about 25volts in average, they use many sheets in series to increase the voltage but the current stays the same in a series connection making them become these huge long bulky rectifier abominations. you can also find them in compact forms in amplifiers from the 50s. Selenium rectifiers have many downsides and most commonly is that they have a limited life span, they don´t last indefinitely and they absolutely hate getting warm and when they burn you get the most unpleasant and terrifying evil smell of burning rotten eggs in a hazardous poison smoke inferno that takes ages to get out from your house and nose, they also drift in voltage drop over time and act unstable to say the least. most people that ever worked with Selenium rectifiers will agree that getting rid of them in stuff you actively use is the best idea in the end and that silicon diodes are the superior choice. But they also are pretty cool in the way they work
Some of the AC cranes I worked on years back had a couple of banks of Selenium rectifiers fitted to control the torque of the motor and boy when they blew they stank big time also when you changed them the small stayed on your hands for ages
these are great for stationary car battery chargers. The small reverse current of the selenium rectifier acts like a small load on every halfwave cycle, regenerating the lead-acid battery of the car.This is a proven fact.
One of the major uses for selenium rectifiers still manufactured today is for battery chargers. As you say, they have pretty much ideal characteristics for simple yet extremely effective lead-acid battery charging.
I found a few big chunks of Silicon on the railroad tracks. Before I knew what it was I was probing it with a multimeter and discovered the resistance would change depending on where I put the probes. Ive been wanting to make a crystal radio from some of it.
Those are metallurgical silicon. They use it by the hopper car full for deoxidizing steel. It is usually pretty conductive and you can use small chunks as the detector crystal in a crystal radio. On a more humorous note I brought a bunch to a hamfest after the movie Avatar was released and a bunch of people asked if it was unobtanium 😅😅😅
As a teenager, I used it to build power supplies. Selenium rectifiers were built into all tube radios and televisions. And if the power supply was overloaded, you would immediately recognize it because of the smell of rotten eggs 😵💫😅 73 de DL1LAJ, Andreas
Kid, have you smelled a burning Se rectifier yet? My first one was in 1959 or 60. I didn't listen to your whole gig: Hopefully you got into the electrical characteristics etc. I hope you make money at this, and I tip my hat to glasslinger.
A friend of mine had to repaint the wall of the house and replace all the carpets when a big selenium rectifier of an old power supply, met a shorted electrolytic filter capacitor. The smoke from burned selenium has such bad odour, 1000x worse than sulphur. Truly evilish...
![Tee Grizzley - Blow for Blow (feat. J. Cole) [Official Video]](http://i.ytimg.com/vi/wNLN5xsx5dc/mqdefault.jpg)
"Some magic I don't know" adequately describes how semiconductors work.
In German the word for rectifier is Gleichrichter
We called these "gleich riecht er" which translates to "he will smell soon"
Especially useful as a shout out to someone cranking up some equipment beyond its limits. And everyone was aware of the consequences of one of these releasing the magic smoke
But I prefer rectumfire nowadays if I see one of these nasty things
Nice German pun, I love that :D
Full bridge rectum fire! (if you watched Electroboom's latest short)
My first contact with a (single) package selenium rectifier was as a school boy. If I recall it was rated 250 V, 50 or 100 mA, intended for center tapped transformer. Later on, at my first long term work, we used International Rectifier selenium stacks to clamp overvoltage spikes. Those parts we assembled ourselves from IR plates, 1 x 1” or 3 x 3” sizes. That was the time when I learned that selenium can produce a terrible stink. Finally and even later, I was told thatMiller Electric built their own selenium rectifiers for some of their welders still when the first, barely high enough rated current silicon components became available. I understood that Miller experienced voltage spike failures on the silicon diodes that they had tested. The selenium parts were “more robust”, until they released their stinking smoke. The robustness was a side effect of the leakage as well as the probably 6” square plates of the selenium assemblies. By the way, at one point I also found that the phone company installed small selenium spike clamping diodes to their phones, when some users complained about the excessive scratch noises (or in some cases intentional teasing by young pranksters). I then tested whether two anti parallel silicon diodes could do the same and found that silicon had too high a forward voltage to compete with selenium in this application.
First time I have seen the curve of a selenium rectifier. The reverse curve was most surprising. The capacitance is expected with such large parallel plates but the crappy breakdown threshold is the surprise. These get quite warm in normal operation at their rated voltage so not only are there forward losses as expected, but also reverse losses that are noticeable.
Its great that you watch this channel! I watch you all the time! Two of my favorite creators in one place! 🤓 And YES the reverse curve was very interesting! NOT what I expected at all
More vintage rectifier fun ruclips.net/video/DvYxR51YYus/видео.htmlsi=P6t_KJ9tzxvB0oCM
just subscribed !! 😀
Well Ron in our younger years we had a lot to deal with these. Per plate they are rated for only 20V in reverse. That's why you need at least 10 for a rectifier in a tube radio. Surprised that you were surprised 😆
@@LutzSchafer To Ron a rectifier is glass and steel, with a glowing filament inside, or beefy silicon. Ignoring the smell years is a good thing.
Thank you so much for taking one apart! I was yelling at the screen, just then you did! And watching that curve tracer with the Selenium Rectifier was more entertaining than watch HBO.
Woohoo! Some old Sarkes Tarzian selenium rectumfriers! You know why they call them rectumfriers? Well, if you run too much current through them, the hydrogen sulfide will off-gas, and the surrounding area will smell like a......rectum!
In the old Zenith AM/FM radios from the 50s, they used a 22 ohm in series with the selenium; these can be replaced with a 1N4001 and a 47 ohm. Basically, you adjust the silicon diode series resistance to keep the B+ the same as it was for selenium.
Or, in German, the "Selen-gleich-riecht-er" 😊
As far as I remember those selenium diode stacks were built for about 20 V per stage and a forward current according to the size of the cooling fin. Such a size could be for around 1 A or little less. I remember two bridge rectifiers I bought to play with 55 years ago which were for 1 A ans 2 A and 20 V.
The ones in tube radios had quite a number of stages to reach about 200V. They also had much smaller plates for lower current.
Selenium diodes were just barely better than having no rectifier at all. Worked okay to provide low current high voltage supplies in valve equipment but are totally unsuitable for signal detection or anything that requires fast switching and low capacitance.
Interesting. I was trying to figure out why the curve tracer started to slide away from 0V (and looped a little) as you increased the maximum forward voltage. It makes more sense now when you looked at the reverse characteristics and the huge loop... explaining it to be due to capacitance. So even though the drive voltage reaches 0V, the "capacitor" doesn't have time to discharge to 0V before the drive voltage increases again. Makes sense, but then the I-V curve would differ dramatically based on the frequency, yeah?
The loop also reminds me of the hysteresis curves seen in magnetic core memory... but then as I understand, that's not a capacitive effect, but a magnetic effect. It seems a major part of understanding I-V curves is understanding the physics behind the part!
I have an old 6 / 12v car battery charger that's full of selenium rectifiers (basically just those and a transformer). I'm sure it's 40 or more years old, but it;s the best charger I have for resurrecting older or completely discharged batteries!
Yes, it's true. The explanation is that there's some reverse current, so some AC on top of the DC. The small amount of AC "breaks through" sulfation on the lead plates and helps a little with liberating the gas bubbles at the plate/electrolyte surface.
There were devices sold for anti-sulfation or battery rejuvenation that emulated the output by injecting some AC onto the output of a DC charger to recreate the effect once silicon rectified chargers became the norm.
Me too - the brand name is James Grose and my Granddad got it by collecting cigarette cards.
it's got an ammeter and some sort of thermal cutout... and that's it. brutal yet effective I guess.
@@joinedupjon That's it! The bimetal strip pushes a button out if things get too hot. There's no true ammeter on mine - just the charging current - an actual ammeter with negative readings would have been handy. There's no brand name left on mine, as it was rescued from a damp barn about 6 years ago, when it has stopped working, after servicing tractors and other farm vehicles for decades. Most of the paint was gone from rust, and the front panel controls were just about functioning and legible. The only fault it had was one that the bolts - not solder - holding the charging wires had rusted so badly there was no connection. Have it a good clean with fresh bolts and it has worked perfectly since then.
Thank you for this! I noticed the capacitance aspect of selenium rectifiers when replacing a selenium bridge rectifier on an old German radio. It seems that after installing silicone diodes I was getting a lot of oscillations until I added one small ceramic capacitor across each diode of the bridge. I later found a post somewhere in German that recommended the same thing...
I have some of those in old, Model Train Power Supplies.
Yes, I got one too with model train. It was early 1970's. It was good powersuply to experiment until it failed...
Ones I dismantled were made of aluminium. They had some bismuth plating on one side, then some dark selenium on it, and then this tin-cadmium alloy on top. Layers on top had slightly smaller surface area than ones below, they didn't go all the way to cover them, so there was a dark "frame" around the tin-cadmium layer. The entire plate surface is a diode, and a capacitor. In vacuum tube devices I've seen there were multiple of those plates stacked for more voltage. Some had holes and screws, others were inside a plastic box with some springy inserts on both ends of the stack, and didn't have holes in them.
6:26 you could even wire all three terminals to that component tester and it should tell you that it is a dual diode. that way you could find out whether they are common anode or cathode if you were unsure.
I can smell this video. Remember when some really large half bridges blew up in a forklift battery charger and the smell was God awful 🤢🤮😂 Ended up replacing them with some heavy duty alternator diodes.❤
See if you can find some copper oxide rectifiers and put them on the curve tracer. Also get an old galena crystal with cat's whisker for the tracer.
One of my Hickok tube testers has a copper rectifier in the meter circuit.
The reason stacks of these are needed for tube radio HT circuits is that a selenium can only accept a reverse voltage of around 18 volts, so you need several in series to cope with the high reverse voltage. When conducting (forward voltage), each junction may only drop around 0.3 volts, but as there's a stack of them, several volts will be dropped across it. So if a silicon diode was just slapped in instead, the HT voltage to the radio would probably be around 15v too high. So a beefy (maybe 10W or so) resistor needs to be added, to drop that sort of voltage. It might be around 200 to 300 Ohms. But the current is only milliamps of course.
As others have mentioned, selenium is photosensitive, so perhaps those thick layers of paint are to prevent them acting as tiny little solar powered generators.
Big Clive also has a nice video on these.
As a kid I remember shorting one out and the smell was horrendous (selenium hydride, etc) and extremely toxic too (cadmium ain't good for you either).
Interesting video, I took one apart and didn't realize the actual diode was a thin layer on the plates. I would wear gloves if taking these apart.
The old transformers for my 1960ish Triang electric trains have this style selenium rectifiers in them. They handle 15V 1A max so I've been thinking of replacing them with a bridge rectifier or maybe just some diodes. Thanks for sharing!
In old radios and very small TV sets we replaced a Selenium rectifier with a silicon rectifier and a 10. Ω at 10 Watts resistor. (called a "sand Ohm" resistor) Ron W4BIN
hope you washed your hands well after that.
A while back I helped a friend retrofit a modern motor controller into a 1950s electric forklift. The selenium parts that came out were the size of a loaf of bread.
Nasty smell when they overheat. I believe they are somewhat toxic also when they outgas. Definitely a type of rectifier that hardly is useful anymore. Whenever I find these in old radios or old TVs I replace them with silicon diodes with a series resistor.
Isn't it great when things are not as you expect.
Selenium rectifiers were in use for a long time in small scale model locomotives. Just the blank disc with no further protective coating. Especially Fleischmann seemed to have a huge stock of these. I'm not sure if there is one or the other model remaining where these still are used.
I've taken these out of electronics that were in use since the 50s and 60s and 70s, they are very smelly even when they don't go bad, and I've seen them in all sorts of shapes and sizes, I wish I hadn't thrown them all out a year ago in a move, I liked to keep them in a 10 gallon bucket next to my work bench since they interested me with their shapes and sizes and rarity
I have seen them used mainly in low voltage dc supplies. (Battery chargers, etc.) Standard balanced transformer outputs to the anodes, and get dc from the transformer center tap and the common cathode.
I worked on a couple Zenth Trans Oceanics; both had mostly dead half wave seleniums. Replaced both with 1N4007 and series 2 watt 50 ohm resistor.
I run across rectifiers like that in tube amps, but not for the B+ which is still tube, but rather for the bias supply and DC filament for phono preamp tubes. Sherwood used this design a lot.
Fisher tube high fi used them in the bias circuit as well, or perhaps it was Scott ? The rectifiers were a shiny flat pack and made in Germany by seamans.
trying to understand the various shapes of component curves, when there is a circle like I-V on a device that is not considered a capacitor,is this more of a electrical defect.or a physical characteristic of the part being measured? thanks
John
Did you know you will get a small amount of voltage across the center of the selenium rectifier and a bright light.
A forward voltage of around 0,3V indicates that this might more likely be a Germanium diode.. Selenium would be in the 0,8-1V range per platter, as far as I recall.
my old troy built garden tiller has one mounted on the side of a 7hp tecumseh its a charger for the starters battery
This is the problem with replacing selenium rectifiers.. You certainly can use a diode + resistor combo, but if the original rectifier had any significant power rating, you end up needing a large, hi-power resistor to replace it.
This is why the selenium parts usually have such large cooling fins - they were terribly inefficient, producing large voltage drops and making a great deal more heat than a modern silicon diode by comparison.
I have a RCA voltohmyst that had a Selenium in it. I took it out and replaced it with a 1n4007. I forgot what value resistor I used but it's a 2watt metal oxide.
I cannot tell where your probes are connected; however, metal semiconductor junctions are Schottkey diodes. To get ohmic contacts it is necessary to overdope the semiconductor under the metallization.
very interesting video - thanks! Lead sulphide (galena) was used in crystal radios, a low-signal application, and the Vf was 0.37V. I would be interested to know what the forward characteristics of a copper-copper oxide rectifier is, if you have looked at them?
I have an old, very large, HF linear amplifier that had 2 sets of selenium TVS diodes across the primary of the plate transformer. I didn't even dare power that thing up with those in place. Good thing too, one of them was shorted. They stink really bad when they go out... and they WILL go out. Producing hydrogen selenide gas. Toxic stuff. Had no idea on suitable modern replacements as I can't find specs on these ones.
Always interesting, thank you.
I am thinking that selenium is toxic and should be avoided? Am I correct?
yes, avoid these things
FOOOOOL BRIIIIDGE -RECTIFIERS- DIODES
The power this should be able to handle is about couple of W, so with 5 Ohm and 0.3V initial drop we will be talking about roughly 0.5Arms rating, ballpark. In real application virtually the only limit is the resulting operating temperature resulting from the power dissipation and the cooling.
Technology is ever changing and in regards to these old rectifiers I say throw them out. Replacing one found in an old
tube radio from the 1950s , lets say, is so easy just by using a silicon diode which are cheaper, smaller and way more
reliable. Put a resistor (of pertinent value for the tubes) in series with the silicon one and be done with it. It is a no brainer
for this young man's vintage radio restoration efforts. Just like a 50 year old car tire ( never used), the tire and the
selenium rectifier are both are bad now. Let go of the past.
Ah, i see! Selenium Rectifier is not Rectifiers, is asymetric charger/discharger for the capacitor
Nothing smells like a burning selenium rectifier. 👍♥️
"Selenium Rectum Fryers"
Any modern use for these? Have a pile from old radios and stuff.
Old timers know them by the terrible smell when they let go.
19:56 do you measure the normal diode or only the plate?
I could smell your thumbnail...
What is the surplus site? Would love to know where do you all get your old timey electronics, all I do is browse ebay
surplus shed
@@IMSAIGuy Thanks! Looks interesting, lots of fun optoelectronics. Do you have some go to pages to find interesting stuff? Im in EU and ironically the strict recycling legislation makes it super hard to find second hand testing equipment, cool components etc
@@guilldea most of my stuff is either local or eBay
@@IMSAIGuy thank you for your answers, guess it's just a matter of being more patient, love your videos by the way! :)
And they are actually photosensitive. They will put out milliwatts in full sun.
1) IIUC, lots of folks think these are dangerous and are a replace-on-sight component. 2) If you wanted to build a Voltage Controlled Oscillator from a pair of back-to-back diodes, would a weak knee be advantageous, because the linear area of operation is extended over a greater slope? I am just a hobbyist and I am talking way out of my league, here... Thanks.
Deadly smoke when they go BTW. They also age on the shelf so I wouldn't trust the measurements you make on old ones.
Smellifier
In the day use seleniun rectifiers also poor solar cell remover the paint .
make a FET out of selenium
u have nice tools
What then? No scratch and sniff test?
What's really inside them is fart gas. Ask me how I know.
I think the PIV shows a bad diode.
These would still require descent PIV to be useful.
I require regular PIV to be useful...
;-)
Maybe they are painted because cadmium is very toxic
as is selenium
Stinkers!
Something that only us old men (I'm 75) can, but would rather not remember. The SMELL of a shorted hi voltage electrolytic filter cap, and the SMELL of a failing/ shorted selenium rectifier. In the 70's-80's I worked at a big auto parts store which also had a separate welding store. I used to repair the older unregulated battery charges which had only a selenium, nothing else--with a pair of silicone stud rectifiers out of one of the smaller Miller welders. Made a small heat sink out of a scrap of aluminum sheet, and off we go.
I'm a bit younger than you ( not so much ), and i remember times in which me and my friends tried some tests on electronics ... i got an old television to dismontle, it had a big selenium rectifier inside. We would obtain DC current from the mains, my friend was convinced to connect the selenium rectifier in a ( for me ) erratic manner, so i said ... "ok, if you are sure, let's try ... " ... Soon smoke arised from the rectifier, and at once on orrible smell was around. Open the window was just a minor solution, we ( three friends ) had to escape from my room and also close the door ... I can't forget THAT smell, even today ...
( i also believe was tossic ... )
You don't need 4 diodes for a full wave rectifier if you use a center-tapped transformer. That was a common practice.
Or use two equal transformers, one leg both of them to make common ground and rest is known if you know ;)
The point of the full wave rectifier is to get both swings of the ac voltage, from both power legs. You can get DC from AC with only one rectifier, but it will be very choppy.
@@tableseven8133, true, but you still get full wave from the CT xformer and two diodes. That was a cost trade when diodes and tubes were more expensive than copper windings.
@@tableseven8133 I have my amplifier build that way. Transformer is 12-0-12 and 550va. When i use one schotky-diode for each end. i can get 100hz pulsing dc that is dropped unloaded voltages only one diode forvard voltage drop per wave. if i have used typical full bridge rectifier, there are always two diodes series, when wave is present and voltages and current are generated for the load.
Class A operated amplifier with pi filter and that transformer diode combination give me useful output voltage about 14.2volt +-0.3.
Before filter, voltage is about 15v, and biased constant current is about 14A so big load all the time, when powered on.
But it is so good the sound quality.
And if you want to know, speaker output terminal ripple for idle with this configuration is about 4mv pp, which is not so much at least my setup for not hyper sensitive speakers.
Still, about nominal 10W for 2 ohm and stable to support speaker dip to low as 0.5 ohm gives a good umph for the many songs.
Here is a link for that full wave two diode rectified circuit.
madpcb.com/wp-content/uploads/2020/11/Center-tapped-Full-Wave-Rectifier.jpg
A full wave rectifier uses two (2) diodes. The full wave _bridge_ uses 4.
1:30 "What's in side this thing?" A heinous, infernal stink. Had one go on a 600A welding machine years ago. 600A three phase from the mag-amp controlled transformer. Big stacks. Oh My God. It cleared the shop out for the day. I could smell it for weeks, at work or not. It was a couple months before visitors came in and didn't mention the stink. In a welding shop.
I've had a tiny 250v 75mA one go. I can't imagine how amazingly bad a 600a one must have smelled.
I still remember it with the occasional flashback. Welder still rand pretty smooth. These have such a soft characteristic, loss of 1/6 didn't hurt it much. Fantastic arc gouging source, and great for running ceramic heater blankets. As the only real engineer, I was supposed to fix it. I was NOT going to go in to that case. Ran it for a couple days at low load to get the stink out, then felt good tack welding the case shut. Probably still in use. @@mfbfreak
A nice look into selenium rectifiers - I wouldn't discombobulate one without gloves though. Still better than letting it burn, emitting toxic hydrogen selenide.
When these diodes overheat they smell like "rotten-eggs" and I was advised by my "elmer" to never breath the smell/smoke back in the 1970's.
I've heard so much about the stink.... I "need" to smell it.
@@edgeeffect Trust me you do not, it lingers for a really long time.
@@SeanBZA see, if this was a chemistry video, not electronics, you'd just say "there's always got to be one". ;)
@@edgeeffect I hope you're just sarcastic :)
Every tube radio that I have ever seen that uses a selenium rectifier used it as a half wave rectifier....
It's pretty common to replace them with a standard 1n400x diode and a resistor. To get the right resistance it's kinda guess and check by comparing the B+ voltage to the schematic. A 3-5 watt resistor is usually adequate for an AC mains radio, but a 10-20 watt resistor may be needed on an AC/Battery radio.
The antique radio restoration community is pretty split about replacing selenium rectifiers. Some swear they are fine to leave if working, some swear they are a failure waiting to happen, and some worry about the cadmium and want them gone working or not.
Most that I have seen have crept up in resistance to where they need replaced because the B+ voltage is too low.
Interesting; the selenium rectifiers in tube radios I've stumbled across were all FULL BRIDGE RECTIFIERs and they were mostly encased in a metal package, making it easy to replace the innards with four 1N4007 diodes.
@@KeritechElectronics Odd, what radios?
@@nobodynoone2500 Polish and German ones from late 1950s/60s if they didn't have tube rectifiers. I recall Tatry 3281 having a selenium FBR.
@@KeritechElectronics Depends on the radio, if it was a drop in for a rectifier tube a half bridge, like here, was common, as the transformer was already there, and wound for a half bridge with common cathode and 2 anodes. For HV the stack would have to be 20 plates, though they often stacked them close together, so they fitted roughly in the volume of the old tube, to get a 400V reverse withstand, to allow B+ and peak reverse voltage to not break them down. Only after the transformer was redesigned, old stock run out, did they switch to the full bridge, as it used the same number of plates, but cut the cost of copper down. Full wave always has a jumper wire, as the stack is arranged to have a common positive point, and 2 series diodes to that point. Then connect the outer plates, as those are the negative, to make the bridge.
Higher the current larger the plates. Replaced a selenium rectifier in my battery charger, after 50 years those diodes were rather tired, and I did the change, using a 1970's era silicon block almost half the size of the old ones, that was rated for 10A instead of the 5A peak of the old ones. Did also add in a charge controller with a SCR, and a festoon lamp to illuminate the dial, plus reverse voltage connection protection. Now a 70 year old GE charger, still working well as needed.
I have seen German made Siemens selenium rectifiers used for the negative voltage bias supply in old Fisher high fi equipment.
They were' Extensively used on British motorbike 12volt systems as a packaged bridge rectifier on single phase alternator, the soft characteristics are well suited to the zener regulated battery charging systems.
They were also used on 1960's motorcycles. Normally dull red in color.
Yes, even in the 70's, as im currently restoring my dad's Honda 750.
They are said to be indestructible, but inefficient. Since Honda's have little knack of producing 'just enough' electric power, im kinda thinking about swapping it for a silicon rectifier. As I already did with my 60's Honda.
On the other hand, keeping it original is kinda my goal, maybe if I notice that it's going to lack power ill swap it...
When I hear Cadmium Selenide I think light dependent resistor
Close cadmium sulfide CdS
@@sciencetestsubject CdSe is also used I think (wikipedia has a link to a Silonex datasheet listing it). I've only ever seen CdS ones personally though.
The LDR used for the tremolo circuit in vintage blackface Fender guitar amps has a cadmium cell in it. They still use it in some of the US market amps but not for the amps designed for export because the EU won't allow the cadmium content.
I have never seen a selenium rectifier used as signal device. They universaly used as power supply rectifiers. They were lossy and heated up a bit. When the filter cap shorted out the selenium rectifier would go into thermal death. They smelled horriblr when that happened. And you touched the fins, you get a nasty burn that take a long time to heal.. Don't ask me how I know this. For a higher voltage rectifier they just added more layers. In the military we had a full wave bridge rectifier in a single, it had several terminals sticking out and jumpers tying them together into a bridge configuration.
I enjoy your videos and working on vintage main frame equipment.
Old tube radios tend to use a pair of diodes with a center tapped transformer, not a full bridge rectifier, so your double diode would be perfect. There were double diode tubes as well. Tube filaments do eat a lot of power, but the plate supply is not so power hungry.
They say you can't smell pictures or video's but this is simply not true!
Your should try to release the magical smell. Exceeding the maximum current melts and burns the thin selenium layer giving up a smelll ranging from horseradish to fart.
On a second note well used them as solar cell. For this the cover electrode usually tin has to be melted and wiped away to allow light to enter. Wasn't working well but after a few attempts over a gas flame with a rag in the other hand you could quickly wipe the molten tin layer off. By leaving on the contact spring washer your retained the contact and viola got half a volt in bright sunlight.
I didn't know zinc selenide was a semiconductor, I use it daily for a infrared spectrometer as a window.
I use a Zinc Selenide plano convex lens in front of my Thermal Camera so that it will focus on PCB electronics from very close up.
Oh my God. You have found a better replacement for the 1N34A diode!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Well done. You made a good example of sounding like someone who doesn't really know what he's talking about. I laughed when you needed a calculator to work out ,2 divided by 40. Did you go to school?😥😢
My 1970 Honda SL70 had one of these under the seat. It never failed, so i didnt get to experience the smell of one burning.
The long stacks were there for high voltage capability, to rectify 2-300V for the tube anodes. Not a lot of current needed there.
Filaments (where high amperages were required) were driven with AC.
They came in dual packs like that because they wanted to avoid the double voltage drop of a 4-diode rectifier, and used a center-tapped transformer with two diodes to do full-wave rectification.
Right ...
Yep. I did replace a selenium rectifier in a old tube radio with a 1N4007 diode and my B+ was much higher. In went the resistor.
Golf carts and my Dad's Dynakit audio amp. Good old days!
I remember stripping them out of old TV's, stripping the paint off of them and using them for solar cells.
I haven't seen or thought about them for decades.
14:36 the paint is for isolation also.
Yes got some nos ones as well. You really want a funny one look for copper oxide rectifiers, which you can actually make out of copper sheet and a flame.
At the TV shop I worked at as a kid my boss said, "If you see any of those damn things, the first thing you do is replace them with real diodes."
Honda motorcycles used a stack of these under their seats next to the battery.
I don't think they would work as a substitute for a Germanium diode in a crystal radio.
Selenium rectifiers as you said would be high current devices is not true though, They are typically low current rated in the milliamps range. Tube amplifiers usually operate at high voltage low current and the average sheet of Selenium rectifier can handle about 25volts in average, they use many sheets in series to increase the voltage but the current stays the same in a series connection making them become these huge long bulky rectifier abominations. you can also find them in compact forms in amplifiers from the 50s. Selenium rectifiers have many downsides and most commonly is that they have a limited life span, they don´t last indefinitely and they absolutely hate getting warm and when they burn you get the most unpleasant and terrifying evil smell of burning rotten eggs in a hazardous poison smoke inferno that takes ages to get out from your house and nose, they also drift in voltage drop over time and act unstable to say the least. most people that ever worked with Selenium rectifiers will agree that getting rid of them in stuff you actively use is the best idea in the end and that silicon diodes are the superior choice. But they also are pretty cool in the way they work
Some of the AC cranes I worked on years back had a couple of banks of Selenium rectifiers fitted to control the torque of the motor and boy when they blew they stank big time also when you changed them the small stayed on your hands for ages
That curve tracer looks interesting! Do you have a video about it?
ruclips.net/video/BabCunP9ib8/видео.htmlsi=94_8T5tzVTygSV46
these are great for stationary car battery chargers. The small reverse current of the selenium rectifier acts like a small load on every halfwave cycle, regenerating the lead-acid battery of the car.This is a proven fact.
One of the major uses for selenium rectifiers still manufactured today is for battery chargers. As you say, they have pretty much ideal characteristics for simple yet extremely effective lead-acid battery charging.
When In learned about Selenium Rectifiers , I was told that they are very poisonous by my mentor.
When they failed the red smoke from them would give you a bad headache.
I can't think of a reason someone would want to 'emulate' these poor I-V chactaristic
if you just drop in a diode, the rest of the circuit will over voltage and over current. Fixing old stuff is not always straightforward.
I found a few big chunks of Silicon on the railroad tracks. Before I knew what it was I was probing it with a multimeter and discovered the resistance would change depending on where I put the probes. Ive been wanting to make a crystal radio from some of it.
Those are metallurgical silicon. They use it by the hopper car full for deoxidizing steel. It is usually pretty conductive and you can use small chunks as the detector crystal in a crystal radio. On a more humorous note I brought a bunch to a hamfest after the movie Avatar was released and a bunch of people asked if it was unobtanium 😅😅😅
@@christopherleubner6633 I guess on a less funny note I actually used to mine UNO 😛
I've run into a few of those, and if they burn out, they really stink. Schottky & a resistor, all the way!
You can find great big ones in welding machines and battery chargers.
In some countries it is known as cuproks.....including Indonesia...very oldddddddddd
Manuel on www.youtube.com/@electronicsoldandnew is often replacing selenium bridge rectifiers in European vintage radios.
As a teenager, I used it to build power supplies. Selenium rectifiers were built into all tube radios and televisions. And if the power supply was overloaded, you would immediately recognize it because of the smell of rotten eggs 😵💫😅 73 de DL1LAJ, Andreas
Kid, have you smelled a burning Se rectifier yet? My first one was in 1959 or 60. I didn't listen to your whole gig: Hopefully you got into the electrical characteristics etc. I hope you make money at this, and I tip my hat to glasslinger.
A friend of mine had to repaint the wall of the house and replace all the carpets when a big selenium rectifier of an old power supply, met a shorted electrolytic filter capacitor.
The smoke from burned selenium has such bad odour, 1000x worse than sulphur. Truly evilish...