This the best and most thorough explanation of how fluorescent lighting systems work I've seen on RUclips so far. I couldn't find a single error - well done! Also, thanks for putting the mercury fear-mongers in place. :P
Exactly, I recall in my freshman year High School Earth Science playing with Mercury as a curious metal that had many wonderful uses. That was back in 1966 or so I guess so not as much was understood about the "danger" of Mercury. Seems we came away unscathed even though we were mere inches from that dangerous fluid metal.
I remember doing stuff in A-level chemistry, in the late '80s, with mercury. I also remember that my gran had a few hand-held maze games that used mercury instead of a ball bearing. The danger from mercury is if you're exposed to it regularly, and not occasionally like most people, or if it gets into your body (contaminated food/water). Even then it'll take a while for anything to happen.
People fail to realize our kidneys and liver are pretty efficient organs. Always giggled when people get up in arms about minor exposure to mercury yet ignore gasoline and smog...
+bigclivedotcom you guys should do follow ups on each other's video's/topics.i would love to see Ben's version of playing with lighting technology like this. and Clive's version of some of the mind bending stuff he does. "applied science; narrated by a guy with a beard."
+dannyjaar Hot cathode lamps go down to around T4 (half inch) diameter, but they go down to very small diameters in the cold cathode style with one electrode wire at each end. They can make them around 1mm diameter for specialist applications.
+THE VERY BEST OF RUclips It's insane...they are planning on outlawing every single type of incandescent light source, down to the tiniest pilot light.
Excellent explanation. I think a LOT people don't understand the difference between rapid, instant and preheat start. Instant start was one of the worst things to hit the industry and I have no idea why they were even allowed. I have been ripping out instant start ballasts when I can and replace them with programmed start or rapid start. Best ones are where they have instant start ballasts on motion sensors. Needless to say those lamps are tortured quite badly. Still can't beat a preheat lamp with a simple choke ballast and starter for efficiency and reliability.
In old fittings we had at work, there was a metal gear tray, but the shade was plastic and was the cover for the gear tray, this is going back to bayonet holders and terry clips. So the earth was bad near the tube, but the original tubes had a metal tape along one side that connected to the metal bayonet holders. Replacing the tubes we found some bayonet to 2 pin adaptors, then used some aluminium foil (the type that is used on windows for burgular alarms) to make the earth strip. the quick fix for a shorted starter was to twist it back and forth to get the light working, then remove it, and at some point in the day wwe changed the tube and starter because the tube would have deteriorated enough that it wouldn't last much longer. btw, they wanted a very visible flashing light at work to indicate that the freezer room door was open, i used a 40w starter in series with a 40w lamp, worked great for years :D
Thanks for a fascinating video which makes me feel a bit more comfortable about what I did in the 70s when we in UK, suffered the now historic Power Strikes leading to the 3 day week. I was not fiddling with tiny tubes but full 4 foot or 5 foot tubes. I was further hampered by the fact that I didn't really know WHAT I was doing. This CAN prove to be a bit of a disadvantage. I built a two transistor inverter (2N3055) much in your skeleton circuit with cross collector-base feedback. 12.5 vdc I guessed that there was need for a heater voltage at each end and plenty of EMF from end to end. In those days, old TVs could supply ferrite cores from cannibalised, line output/EHT transformers. One transformer supplied 2 cores for this job.. I got the LT winding turns from a circuit for an HT inverter for a ham tx. The over-wind was a "that looks enough" exercise, although it wasn't... and trial and error played a major part. The same applied to the heater winding but I eventually lit 4 and 6 foot tubes. Some interesting things emerged from this. If the system set off on a fully charged car battery, no apparent reduction in light output occurred until the batt volts dropped to 8 or 7 volts or lower. But don't switch it off or it won't start again until recharged! The waveform interfered with MW/LW radio just as much as it did when the ferrite was inside a mono-tv. Couldn't make them fast enough during the 3 Day week.. Never made another when it stopped! I worked out that it was running not at 50 hz but 6/7 Khz.. I was later informed that running at higher freqs produced more light for a lower DC. consumption. Working, not scientific tests showed this to be so. Who informed me of this and other aspects? Raymond Baxter on Tomorrow's World with an item in his show relating to this break through that scientists had made for energy saving....two and a half years later. Thanks for the mention Raymond!
Hi Clive, I’ve had to deal with all the problems you mentioned over many years & when i was an apprentice i loved working on fluorescent signs & fittings but none of the sparks that i worked under understood how they worked or could explain to me what went on. I used to dismantle the old fittings to find out how they worked - later on when i worked on many TV Rental shops i would supply & install hundreds of feet of back light fluorescent tubes to the shelving that ran around the shop - at the time it was very futuristic. I still like fluorescent problems & can deal with them all when called. Edward.
Great video. You're very good at explaining this stuff, so that even guys like me without any more electronics knowledge than what you learnt in high school can understand what you're basically saying. And it's not always easy to be an expert and explain things clearly to a layman.
Great video with lots of information. I love gas discharge lamps and am sorry to see them being phased out in favor of LED, even when in some cases there are few benefits.
Also worth pointing out that the metallic mercury vapour in flourescent tubes is a great deal less harmful than organic compounds of mercury, which is the really nasty stuff that requires very little exposure to cause serious health effects. Most people don't understand the difference.
People don't explain fault situations like this often... fucking great stuff man... I appreciate it, I'll have to watch it again and take notes. Edit: I need more of this stuff... this is awesome.
Thanks Clive for clearing up a matter that has had me wondering for many years. I've been using various methods of 12 volt lighting for most of my life. There was one particular type of 12 volt fluoro lights that I bought, where the tubes only lasted a couple of weeks to a month with daily(or nightly) use. I now understand why. It was a double fluoro with 2 x 8 watt tubes driven in series. The circuit itself didn't look much different visually to that of a single 12 volt fluoro. Obviously, it must have driven them quite hard. lol. It was a bloody bright light too, it's what disappointed me when the tubes died so quickly.
I really learned so much! Thanks so much, Clive. I knew I knew nothing about florescent tubes, and never got around to learning. I'm chilling at home waiting for the missus to come back from visiting a friend and your new videos made me afternoon great!
Than you Clive for explaining this "rectifying effect" I always wondered why tubes that are at the end of their useful life sometimes go into 50Hz operation!
Nice video. Merry Christmas! In my fledgling Electronics Geek days I took a 12v relay and wired it so when the coil pulled the contact in, it disconnected the coil and when it returned, it pulled in, essentially a mechanical oscillator. Fed the connections across the contacts into a step up transformer and the output of the transformer into a 4' florescent tube with the filaments wired in series. Amazingly, it lit the tube. I know there are so many reasons this isn't a good idea, but I look at it as a learning experience. Many years later, you explained how they are supposed to work and why.
Great video! I'm glad you mentioned the need for a grounded reflector with some bulbs. I discovered this one morning at work (broadcast engineer) when the background lights in our TV studio weren't working. I took the trouble call, grabbed my DMM and a screwdriver and went to see what was wrong. There was about twenty 32 watt lamps and none were working. The voltage into the ballasts was OK. Thinking that the bulbs weren't seated properly, I grabbed a bulb to remove it and it lit up! Each bulb lit up as I touched it. It turns out that an audio engineer was troubleshooting some hum issues had disconnected the ground wire to the set.
@bigclivedotcom I am an electrician from Sweden and I do appreciate the topic and I do understand what you are talking about. I think the Starter you mentioned during the first start-up is the 'Reactor' as proper term. Non-electricians call it Glimmtändare though. The second circuit, I have not really seen. Maybe I am too young to really see them since Energy efficiency in light and the ergonomics of it is fairly big subject in Sweden both for us electricians and those that project(Plans bigger jobs). So it is generally either two different topics of subjects regarding starters now days. It is the Reactor and it is the HD-reactor. The latter being the electrical version. On the topic of light bulbs, both business and electricians in general tend to recommend you burn the tubes in for roughly 1k hours for the best light. Obviously not a topic you bring to a private customer that won't appreciate paying such electrician bills but I thought it was interesting to enlighten you. The reason, I am not really sure but I assume it has something to do with the gases within the tube along with the potential Mercury vapour to settle into its environment and get used to the equipment. But this is hardly the focus for Electrical. The advantage lies in the shifting of the Sinus curve. The faster the shifting, the less load on the bulb but also better for individuals in general whom work under this light. You said one could see shimmering/flickering. Technically, you can still see this on a good bulb as well if you have good eyes and science has proven that 60hz is a reason many people in office environments etcetra suffer headache and lack of ability to think since the brain unconsciously notices the shifting light that seems constant to you and therefore uses a bit of brain process for that part. How ever, once you exceed a couple of thousands, I think it was 5-6k hz the brain simply stops thinking about it. Many people seem to work better under such high speed frequencies. Oh, a warm light is also generally a colour to not use in office environments as well. Cold light produces less depression like effects. But out of company aspect when creating these electrical circuits, you generally want 10k since it is easier to calculate and construct money wise. Don't ask me why but perhaps because it is an even number and they can limit the electrical parts? Ah well. Now, it comes to a topic that is generally a discussion between electricians in sweden and customers. Both private and business alike, it is when it comes to the HD-reactors themselves. On certain brands and models, the HD-reactor dies quickly and on some it doesn't, generally the reason is price and where they bought it from. While everything else is fairly simply and straight forward, it might be a cheap China brand which is known to be a bad product in Sweden. We have highest electrical standards in EU to my knowledge. I wonder, what you think is the reason why some HD-reactors burn out quickly while others don't? IS it potential damage in shipping or handling? Or is it simply product deficiencies? Also, for LED, you should preferably don't use them in bulb form on a location where you work under many hours. They generally do not offer good long term effect on your eyes. Although you might not process it, your brain will process things unconsciously even for LED. I suggest simply buying an overpowered Light fixture and simply dim it down. Lengthens the life of the bulb and improves general long term effects. Especially good for detail work where LED is simply inferior still. LUX being the reason to my knowledge. LED has too much, direct light in comparison to normal fixtures is my theoretical reason also something to do with colour but that is beyond my ability to comprehend. Either way, long rant but hope it was an interesting read. EDIT; LED is useful in colder environments though, it is fairly easier to start up in lower temperatures. Regards.
older vid i know but commenting anyways. this basically perfectly explained what was wrong with my bathroom light where I had no clue how to fix it. liking these videos more and more, so much learning :D
Great topic, man I know a lot about this but it is always good to get e refresher on it. When I was a much younger lad (1963) I had a television transformer that had, I believe a 750v winding. I took a 40 watt tube and powered it on, albeit very briefly, with said Xformer. Talk about a flash tube... It blew the mains fuse and all I could see in the dark basement was the retinal burn of that lamp. The lamp was now ALL black and made plinking sounds when shook. Ballasts are very unique devices and you are correct that the topic gets going on to know about them all. An electrical engineering text I had, spoke to the development of specialized high frequency gen sets for office buildings fluorescent lighting that would have yielded more lumens per watt.(over 50 or 60 Hz) I do not believe this was a commercial success. Fast forward 5 decades and the compact fluorescent is indeed a marvel of efficiency BTW Merry Christmas Clive
+Craig Nehring The old TV transformer may have been an auto transformer with the higher voltage referenced directly to the mains. That way, the high voltage would strike the tube and it would then try to crowbar the mains.
+Craig Nehring I am glad you mentioned that, I'd forgotten they tried to run at HF. Weren't they 400Hz? (that way transformers would have been readily available from the aerospace industry)
Pretty sure that transformer secondary was isolated and just dumped the full output across the lamp, then again it could be as you say. It was all over in literally a flash. Back in my early experimental days I would try and run things backwards, for example using an electric model train transformer I would output that low voltage AC into the various winding's to see what voltages I could get. One thing for sure there was a substantial current flow through that lamp. Cheers
***** Aircraft do run at 400Hz for good reason the magnetic cores can be smaller and lighter, not sure what freq they had hoped for on the genset in office buildings but the CFL I believe is running at 10kHz. Most aircraft also use 28 VDC for the same savings, wire capacity is thinner and motor are smaller. Aerospace hardware is very spendy: certification, regulations et al. I used to do component overhaul on some Lockheed and Airbus craft.
Confused...? 21 years since I have been involved with electrons, professionally, but surely the lower the volts the higher the current requiring more copper. ? Higher the Hz higher the impedance, good, less copper. ?
Some of this was over my head, but still very informative. I've seen all of the problems described, and now I have a better understanding of what causes them and how to address the problem. Thanks!
Thank you for the awesome video :) The videos you make are always packed with so much valuable info. I have watched so many and actually learned something that i still remember and use so often in everyday life. So Thank you, you are doing a valuable service and should be commended on it!
This is a very thorough and explanatory video. I'm not sure what the UK standards are, but here in North America our fluorescent lighting systems utilize a special transformer called a "ballast." There are two types of ballasts: magnetic and electronic. Magnetic ballasts are essentially a choke connected in series with a starter, although newer magnetic ballasts often have built-in starters. The only downside to magnetic ballasts is that they generate A LOT of heat, especially when left on for more than 6 hours a day. For this reason, many magnetic ballasts are equipped with thermal protection so if they should overheat, they will shut off and restrike once the fixture has cooled down. However, this isn't always the case, and older magnetic ballasts were sometimes blamed for starting fires. There also have been instances where the thermal protection has failed, causing the unit to become dangerously hot. In 2013, the government "phazzed out" the manufacturing of magnetic ballasts. Electronic ballasts are much safer and more efficient in the long run. In fact, many modern electronic ballasts are instant-start and capable of driving 2-pin tubes. I have some in use and I've noticed that the tubes are "sputtering" much quicker than with the old magnetic ballasts. There is another alternative, a programmed-start electronic ballast, that starts the lamps in the same manner as the previous magnetic ballasts. These are often a bit pricier and may be somewhat difficult to come across, although they do not "abuse" the lamps like the instant-start electronic ballasts. By the way, mercury is NOT dangerous! The government overemphasizes safety when it comes to handling fluorescent lamps. I have shattered many fluorescent tubes, including a shattered CFL in my den. Not to mention I also played with mercury when I was only 6 years old...
Thank you. That was an excellent introdction. The books normally just explain what you should do but do not expose the consequence of not doing it the recommended way.
The one thing I've never understood was the weird faults you see on buses in Edinburgh. When the system starts to fail the tubes will appear to have flowing light and dark areas moving from one end to the other, like a good old rope light sequencing. What's weirder is that when I spot this there's usually three or more tubes doing this in sync. I always thought it was an effect from a high frequency power supply failing. Thanks for the description of what appears to be happening in their lob-sided power supply. Happy 25th December.
+PIXscotland We have a tube in the basement that does the same. It has darker bands (didn't really measure, but maybe approx. 3-5cm apart) that slowly move in one direction (again, didn't measure, but it's also in the order of 3-5cm per second). The tube does not have an electronic driver and the banding vanishes after 5 minutes or so. If I had to guess I'd say it has got something to do with the gas heating up, expanding and contracting with the drive frequency near the cathodes of the tube. This could lead to some kind of standing (or slightly moving) pressure wave until all of the gas in the tube has heated up uniformly.
+PIXscotland If they are being driven from an electronic supply you may be seeing a standing wave effect called jelly-beaning in the neon industry. There used to be electronic transformers that could be accurately tuned to cause jelly-beaning deliberately.
Real good video. Thoroughly explains why my 12V floru drivers i built 20yrs ago is still going strong, yes the filaments at both ends also driven with low voltage, the extra windings on the TX as you show. Additionally used another current sensor Tx for f/b. Used a 555 to drive a mosfet.
Hey BC, ever play with "No Cathode" fluorescent tubes? They last centuries. Or, about 25yrs in continuous operation. Eventually the phosphor starts to dim out. With no electrodes inside to trap gas ions or sputter a black stain, the main failure mode is removed. The commercial type is the Induction Bulb, about $30 on eBay. Or, make your own 1890s Nikola Tesla-style bulb by wrapping adhesive foil rings around the tube ends, placed on the glass a bit forward from the filament location, positioned so the plasma doesn't touch the unused metal parts inside. Attach a CCFL driver to the two foil electrodes (use those little 2-transistor drivers found in cold cathode LCD backlights.) A small conventional tube will light up full brightness. No electrodes, at least not inside. Watch out though, since the drive is about a kilovolt, and any sparks can be a burn hazard! When CFL spiral-bulbs were first announced in 1974 Popular Science, they said they were going to be powered capacitively by high-volt 27MHz, lack electrodes, and last for over 100 years. Then, nothing. No peep about them for years. When they eventually appeared, they weren't like those glass-capsule electrodeless "RF lamps" being promised. Instead they were just conventional fluorescents, but bent into a spiral. And, still employing the interior metal electrodes which get destroyed after a few years of operation. True "RF lighting" does exist, and it lasts far longer than HID, longer than LED. Not sold in stores everywhere! Buy it now, NOT! We can get the fluorescent forever-bulbs from offshore suppliers via eBay. Still, those have an induction loop coil and expensive ballast driver circuit, and a torus-shaped glass globe. Not the simpler capacitive coupling done by metal rings outside the glass. Oh, and about "mercury freakout?" This used to be a real issue back in the 40s/50s when lamp internal coating included BERYLLIUM OXIDE, and the white dust from a shattered lamp was carcinogenic toxic inhalation danger. Not the mercury. People in the early-1950s factories were dropping like flies from beryllium poisoning, corporate coverup, big scandal. And everyone was supposed to leave the room if a tube shattered, until the BeO powder could be washed up. But then toxic fluorescent powder was replaced by other materials. Yet everyone for decades later still goes crazy when a fluorescent tube gets broken. Very silly. But we can see where it all comes from, as the older "dire warnings" get passed down the generations, even though the lethal hazard was removed fifty years back.
+wbeaty I have played with the electrodeless lamps. In fact I burst a full-price expensive one when they were new while trying to open it. :( There was a company selling DIY neon signs for a while that used electrodeless letters that clipped onto a high frequency exciter backplane. I also recall that my first EPROM eraser had a totally sealed tube which got its power capacitively coupled from two ordinary fuse clips. Then there are those intriguing linear lamps that are basically based on xenon tubes, and have one electrode at one end and a clear conductive coating. At the Edinburgh Tattoo we switched to using induction lamps for the house lighting. Fast warm-up time and low enough power to run on a battery powered backup system so we have full output in the event of power failure.
+wbeaty the mercury tubes do severe damage on submarines, long ago when I was in navy they had florescent tubes on subs and it killed one entire ship one lamp breaking. nice to know about the beryllium, thank you for that
+FoodOnCrack Nope. Tesla's lamps used external metal plates, capacitive-driven with no need for a glass torus. Demo: hold any discharge tube near a Tesla Coil, even one lacking electrodes, and it lights up. The version used in Tesla's laboratory ceilings were phosphor-coated capsule-pipes, with external metal near the ends, driven by HV at KHz. No-cathode, rather than cold-cathode. Last's nearly forever, eliminates the need for billions of dollars of "re-lamping" and hazardous waste stream. Mod your conventional fluorescent tubes by attaching foil electrodes outside. But Tesla used all his outside funding to build his Colorado lab, when he'd promised investors that he was starting a fluorescent lighting manufacturing plant. Coulda been like Elon Musk of 1895! Wiped out Edison corp's incandescent lamp sales, become a billionaire, and started any future projects he wanted. Oops.
Absolutely brilliant. Great vid. I have those very same tubes in my camper. Replaced them only 2 weeks ago. Worked fine until this morning: now the tubes are dim, slight purple colouring at the ends. I imagine the ballast units are defective. LED strips here I come...
I love the way you describe all the failed design to strike the tube. As it turns out the newer state of the art driver is called a "Programmed Start"... it makes the tube last 20Yrs with unlimited On/Off cycles - Lots of ballast manufacturers make them, easy to buy on eBay. Basically these ballast glow the filaments before striking them On with HV, then gradually power down the filaments as the tube keeps runing - No more black ends, endless life: Perfect!!!
Excellent video! No idea why I didn't see this in December! I had a few sets of these T5 (maybe they were T4?) in varying lengths as under-cupboard lights in the kitchen; some daisy-chained, all 240V... The tubes "blew" with alarming regularity (at least one every 6 months!), and unlike the larger T8, the fittings didn't have a separate ballast / starter. Almost every time I visited the *same* local shop to replace the tubes, they ended up being a different manufacturer and slightly different dimensions (actually within a year or two even THE SAME manufacturer, had different dimensions too!) with each tube being a few mm shorter or longer... So not only were tubes replaced, but the fittings were too... I gave up the last time - and instead purchased some LED strips and 12v transformers, which I have fitted. They give far better light, show no signs of becoming unstuck, are run cooler and use less power. The T8s in the garage I also replaced, but with "replacement LED tubes" - which also replace the "ballast/starter" with a LED compatible driver - so are "drop-in" replacements.... Oh - the joy of "instant light" in the garage. :)
Hey Big Clive. I wanted to point out something since you mentioned it. I just left the semi/truck driving industry, spent a year driving trucks of the likes of Internationals/Freightliners/Kenworths. I am happy to report that the manufacturers are moving towards using LEDs in more of their rigs. Cheers, mate.
CLive, i used to repair the 24v lighting ballasts used to light up tubes in coaches and buses, it was part of being the electronic repairer of an auto factory. The transformer with heater supplys was the exact way the ballasts worked, they operated at 20khz and a push pull two transistor symetrical oscillator with feedback windings and identical components was the way it worked. They worked really well :-D Your discription is spot on, no supprise there :-D
Once again, another amazing video to learn loads of information from. I can't watch one without learning something to help me with my projects. I have to say, this is likely my favorite channel on youtube.
Wow that video was packed tight with information and very well structured. You certainly know your fluorescent tubes, Clive. Lot's of interesting physics going on there.
Interesting. I didn't know these were warm cathode tubes. I have used various fluorescent tubes like this for moth trapping since about 2009 and never really looked up how they actually worked. I mostly use the high output t5 12" 15W visible blacklight tubes that produce a visible blue light. Since the end of 2014, I moved to using t5 12" 15W synergetic tubes, and these produce an almost lime green light. These produce two peak wavelengths of light at 368nm and 540nm, which is broader than the blue tubes which peak at 350nm, and as such they work far better at attracting moths than the blue tubes. I run them on the battery powered car inspection lamps that use the same 8W white tube you have, just switch out the tubes for mine. I power them on 12v 24 amp motorcycle batteries and I can get around 19 hours before the battery needs charged. However, after a long enough use, the plastic tube that protects the florescent tube, does turn a dark dirty yellow colour. Maybe because of the UV light the tubes I use, produces compared to the white tubes that come with the inspection lamp?
We purchased a house in 1964, which has a storeroom with a fair stock of spare fluorescent tubes. Over the years we used them to replace the bulbs in our kitchen, as they failed. These old tubes lasted for many years before they finally failed. Any of the newer tubes generally last less than a year. Planned obsolescence? Hard to argue against the evidence.
Flourescent tube that dies in a year is insane thing. If it was running 24/7 it can run more than 5 years i dont get it Must be instant start ballast and awful ton of starts?
Thanks BigCliveDotCom. This brought back memories from 1986-1990 when I designed emergency lighting systems for every tube from 8w T5 to 100w 6 footers. I am not that clever, so we did it by reverse engineering the transformers and pcb's then melded different circuits for our needs and CAD them on a BBC Model B computer. We undercut the Chinese and Germans on many products. I also did mercury lamps, Disco lighting and strobes. Never had many fails though and they went into many new UK buildings! Remember that Jon Kay?
That clearly explains the problem I had with battery operated fluorescent tubes in the past as they kept blowing the tubes. The 12V models were even sensitive to being switched on & off. E.g. if switched off and back on within a fraction of a second, it generally caused its electronics to self destruct - a trail of smoke and that burnt electronics smell.
One thing that I learned about learning is that... The more I learn about stuff, the less advanced I realize the human race actually is. Things are too alike. I mean, it seems complex, but once you get it down or hear enough about it, you are like, duh. This other thing actually works just like that too. And mostly, the thing with me about stuff is, I generally have no clue because it never came up in my life. But the second it comes up, I can figure it out. Almost like I knew about it for years already. Some things do not stick with me though. Like, is Psychology. I didn't find it important to remember who said what. All I cared about was why it was said. So when it came time to recall the dude, I was like, why? But yeah, names and flashy words don't stick with me. But, when it is time to learn and develop new things, I am generally the first one with an idea.
Fantastic! Thank you for some answers to my lurking questions... when I see odd florescent behaviour like darkening or purplish ends... humming and lopsided operation... now I know why. You made the answers fun and easy.. Happy Christmas Big Clive
At 4:35 I was waiting for the crash as the tube rolled off. Some older tubes had a phosphor containing beryllium which would be hazardous if broken but these tubes were discontinued years ago.
Back in the early to mid 1980's, my grandfather used to turn the heat off in his garage, next time he would come in and turn on the lights (USA fluorescent light bulbs) they would only come on dim of if used his cigarettes lighter to heat one or both ends, these were usually, 4 to 8 ft, why did this help, and does this still work to-day on straight tubes and curled household fluorescent light bulbs. Thank you Dennis
+Dennis Friedman In very low temperatures the mercury will condense out completely and the tube will just glow with the carrier gas. Heating the tube will help get the mercury back into vapour form. You get thermal sleeves for tubes used in refrigerated cabinets.
I love those preheat fixtures with starters, I actually have one in my basement with a rapid-start fixture, the preheat fluorescent in my basement is a sears fixture
I've always wondered why some tubes have a moving pattern in them when lit, (the light is on but not quite as bright as normal but with a snaking pattern that shouldn't be there). Nice one Clive.
+Graham R Dyer This often happens with new tubes randomly throughout an installation, but usually stops after a while. (But not always.) I actually like the squiggling effect. I've never come across a decisive reason for it though, other than "impurities in the tube" which may be residue from the electrodes. The neon industry tried to recreate the effect deliberately for many years without lasting success. You can do it with short xenon filled tubes and high frequency drivers though.
I remember having one of those battery-operated flourescent lights at one time, it took 4 C or D cells, and never did work really well. At this point I'm not sure what happened to it. I used to have a couple of 4-foot fixtures hanging over my workbench. These worked fine, but there were times when I noticed a certain "fuzziness" on my scope. Looking into it a bit, I found that the fixtures apparently had electronic ballasts in them which were radiating at around 38 KHz or so, not exactly the same frequency, but pretty close. After a move I have an LED bulb hanging over my bench, and those fixtures are in the garage, where they will probably get put up at some point. In recent months I have scrapped out a fair amount of gear, including some flatbed scanners and some flatscreens. The scanners had small flourescent tubes in them, and in some cases I also snagged a driver board to run these. I'm not sure what they want for power, but am guessing that 12V is pretty likely. What's done in flatscreens can get pretty interesting. They have inverters of a sort, and I've salvaged some of those transformers to play with. Apparently the logic of the way those work is that if there's any kind of an issue with either the lamps or the inverters, the set will shut things down, rather than letting it be obvious as to where the fault lies. I haven't figured out a way to troubleshoot that yet, and probably won't pursue it much further as the newer stuff is all using LEDs anyway. I do recall seeing a gadget or two on ebay that was supposed to let you test various bits of this stuff, but never pursued it, and at this point probably won't...
I use as my desk lamp a 12VDC 36W fluro. it uses an AC inverter(electronic starter+ ballast)(That is the name given the electronics by the manufacturer). This thing uses standard 240V AC fluro lamps and doesn't blacken the ends. Anyway the things that are of interest to me are the fluro fittings that are dimmable.
Thanks a lot for this. Please make a detailed video on how today's uv ballast works. Like Philips, osram or other brands. Btw This is the best video on uv light i have seen on RUclips.
I had one of the more weird electrical failures today. In my kitchen I have 2 compact florescent lights that hang over my counter top. They are connected to a dimmer and a aux switch (the cf used were rated to be on a dimmer circuit). I think installed them about ten years ago. The light bulbs were off and I went to turn them on from the aux switch. I could see a orange glow from the starter on both lamps but they did not start, after a second I heard one lamp pop and I turn off the aux switch. I went to the dimmer switch and saw it was on its lowest setting. Sometimes the lights don't turn on at that setting at first but eventually or if I raise and back down they turn on. So to see if I just burnt out one I raised to the max setting and turned it on. One second later the other bulb failed. White smoke and electronic smell. They both made a mess with white powder as I changed them out. Replaced one with a cree branded led from philips and it is working fine. Have to go buy another bulb but I find it strange that they operated for approx 10 years with no issues and been started in that condition many times and both failed in the same night.
Very interesting Clive ty once asked an electrician friend about this , but you know what the average electician is like ;-) I got the vague no idea answer, although , to be fair he was an apprentice at the time, but yours is the best explanation I hVE SEEN :-)
I bought a cheapo 12v fluoro fitting (one of them sub-£10 things with a prismatic diffuser) to use in my shed in addition to my LED tape lighting last year, it barely lit up and was out-shone by my LEDs, so I gutted it and fitted the inside of the thing with leftover LED tape and now it works great, not really fixing the fitting I know, but, it works for me... :P
Thanx about that part about symmetric and non symmetric waveform... i never thought of something coming out of secondary of transformer could be non-simmetric...just dumb i know ;-) it explaines so much weird things i observed i a quick and dirty PSUs i built for fluo. bulbs in a squatting complex without el. power where we run everything on deep cycle lead acid (getting slowly replaced by li ion/poly from laptop packs) + solar and/or gasoline generators... ok we switched allmost all lights to dumpsterdived LEDs(from those powersaver bulb recycling containers we have at supermarkets/shopping malls, where all the good stuff is, like phones, tablets, powerbanks... As they are big/deep bins you have to lean in with flashlight in your mouth/on forehead, like aladin's cave full of broken glass shards, magic smoke smell and mercury vapours hovering near bottom, where all the good/dense stuff is ;-), but anyway, it explains those WTF how is that possible at all moments...
Very interesting video. Thanks I have just started to replace the fluorescent bulbs in my 4ft fixtures with LEDs. The first ones I replaced use the ballast. The next ones I replaced did not need the ballast, which was my desire since the ballast was not running both the bulbs. How about a video explaining these two different styles of LED replacements. I do love the replacement LEDs. Instant start and the one which does not need the ballast has no hum. Also much brighter since all the light is shining in one direction
Used to get a lot of the Tubes on Southwest trains when half of the bulb rectifying and only half of the bulb lighting up. Also london underground used to have tons.
You forgot the tiny amount of Neon as well, droips the striking voltage. The mercury dose is under 4mg, or basically what is in a can of tuna. When the mercury all gets absorbed in the phosphor coat you get those tubes that are lit slightly but the middle is pink from the neon fill. Also happens on new tubes that have not heated up enough to boil the mercury fill out of the little glass dosing capsule, they start off new as pink then get to full brightness over 10 minutes. With the older starters they also had a tiny amount of Amerecium added, so the neon would ionise, otherwise they would not start in a totally dark room until you had a stray cosmic ray pass through the starter to ionise the gas, then it would light.
Wow thanks for the great video, our RV has many of these 12 volt florescent lights throughout the 35 feet of home. We spend our winters living in her, and love the travel and adventure. We rarely use the florescent, just at bedtime, as they are handy to turn on to see the way, and have local switches so you can turn them off as you pass. If we did more boon-docking (camping away from shore power) I would consider replacing them with LEDS but we are usually hooked to 50 AMP 120 Volt shore power so don't have to worry to much about draining our two 12 volt house batteries. Nice to be able to understand what is going on with them, I have been maintaining such lights off and on most of my adult life, so I am familiar with the starters, but always wondered what was going on with them. Now I know, thanks to you.
Duuuude! Very well explained! Sputtering, rectifying effect, high frequency operation, couldn't think of something missing. OK, maybe resonant operation and stuff like striations... ;)
Thanks for the video. Have you ever seen starters with a bulb attached to them? We had a couple of fittings in our house in the 'sixties. They were 4' and took 2'' tubes. At one end was a housing that took the starter. When switched on, the bulb would give a constant light, no flickering, whilst the starter made it's attempts to get a strike. By light output, I guess the bulb was about 40w. Although the bulb was constantly on, if the filament in it blew, the rest of the starter wouldn't work. So it must have had a purpose. The casing was like the old metal starters, with four pins. They became near impossible to source and so the fittings were changed out. I wish I'd have kept one now.
Thanks! This was genuinely very interesting! :) I have a bunch of those handheld fluorescent lights, bought them on a outlet department in a local hardware store once. They use 2 AA batteries and many of them came with the tubes already considerably darkened at the ends, so it's probably safe to assume they use the bad way of driving the tubes.
I really like your channel, but I would love it if you started doing more of these explanatory videos! Keep up the good work and lets get you 100k subscribers!
When I was 7 to 9 yrs old, we have a round florescent overhead light fixture that would only sometimes turn on. What I had found was if I flicked the switch a few times, it would eventually light (key word there: EVENTUALLY - not always as sometimes I gave up). As a child in 2nd grade of school, my mom was not eager to get a ladder (not sure we had one) to have me climb up and look at it. To this day, I would tell you that the starter (that small round can with the neon tube) needed to be replaced, but in 50 years, I've never discussed it with anyone who might know. I know you could see the neon bulb glowing (but don't remember if it was blinking or just on solid). Was my 9 yr old brain correctly guessing what needed to be done?
Don't know if it's already been said, but now I know why some of the really old tubes had that black "stripe" down them! I'm assuming it's a conductive strip between the two metal caps on the end that would help with the strike :) - thanks Clive!
The shown color temperature is intresting. Normally in the U.S, the color spectrcturm starts from 2700k-6000k. In other countries it could be a little different. But very interesting video.
Nice, could have mentioned or touched upon the beginnings, namely why we have tube lighting in the first place and that without Tesla, we would still be mostly in the dark ages.
I think you said the main purpose of the inductive ballast was current limiting for the filaments. I also generates a high voltage pulse when the starter open that helps get the gas ionized. If you replace it with a equal value resistance the filaments would heat but the lamp would not start.
technically, you do want to leave the area and ventilate if a tube breaks, mostly to allow the vapor to dissipate as even in small quantities mercury is harmful. however, you wouldn't need to call in hazmat or panic, only calmly leave the area for a few minutes
This the best and most thorough explanation of how fluorescent lighting systems work I've seen on RUclips so far. I couldn't find a single error - well done! Also, thanks for putting the mercury fear-mongers in place. :P
Yeah it anoys me when people talk crap about Mercury light sources
Exactly, I recall in my freshman year High School Earth Science playing with Mercury as a curious metal that had many wonderful uses. That was back in 1966 or so I guess so not as much was understood about the "danger" of Mercury. Seems we came away unscathed even though we were mere inches from that dangerous fluid metal.
I remember doing stuff in A-level chemistry, in the late '80s, with mercury. I also remember that my gran had a few hand-held maze games that used mercury instead of a ball bearing.
The danger from mercury is if you're exposed to it regularly, and not occasionally like most people, or if it gets into your body (contaminated food/water). Even then it'll take a while for anything to happen.
People fail to realize our kidneys and liver are pretty efficient organs. Always giggled when people get up in arms about minor exposure to mercury yet ignore gasoline and smog...
I've heard low molecular weight, organic mercury compounds are the stuff to avoid.
Good stuff!
+Applied Science That's a super-compliment from you Ben. I admire your technical skills greatly.
+bigclivedotcom you guys should do follow ups on each other's video's/topics.i would love to see Ben's version of playing with lighting technology like this. and Clive's version of some of the mind bending stuff he does. "applied science; narrated by a guy with a beard."
+bigclivedotcom what is the thinnest fluorecent tube made ever
+dannyjaar Hot cathode lamps go down to around T4 (half inch) diameter, but they go down to very small diameters in the cold cathode style with one electrode wire at each end. They can make them around 1mm diameter for specialist applications.
+THE VERY BEST OF RUclips
It's insane...they are planning on outlawing every single type of incandescent light source, down to the tiniest pilot light.
Excellent explanation. I think a LOT people don't understand the difference between rapid, instant and preheat start. Instant start was one of the worst things to hit the industry and I have no idea why they were even allowed. I have been ripping out instant start ballasts when I can and replace them with programmed start or rapid start. Best ones are where they have instant start ballasts on motion sensors. Needless to say those lamps are tortured quite badly.
Still can't beat a preheat lamp with a simple choke ballast and starter for efficiency and reliability.
In old fittings we had at work, there was a metal gear tray, but the shade was plastic and was the cover for the gear tray, this is going back to bayonet holders and terry clips.
So the earth was bad near the tube, but the original tubes had a metal tape along one side that connected to the metal bayonet holders.
Replacing the tubes we found some bayonet to 2 pin adaptors, then used some aluminium foil (the type that is used on windows for burgular alarms) to make the earth strip.
the quick fix for a shorted starter was to twist it back and forth to get the light working, then remove it, and at some point in the day wwe changed the tube and starter because the tube would have deteriorated enough that it wouldn't last much longer.
btw, they wanted a very visible flashing light at work to indicate that the freezer room door was open, i used a 40w starter in series with a 40w lamp, worked great for years :D
Thanks for a fascinating video which makes me feel a bit more comfortable about what I did in the 70s when we in UK, suffered the now historic Power Strikes leading to the 3 day week.
I was not fiddling with tiny tubes but full 4 foot or 5 foot tubes.
I was further hampered by the fact that I didn't really know WHAT I was doing.
This CAN prove to be a bit of a disadvantage.
I built a two transistor inverter (2N3055) much in your skeleton circuit with cross collector-base feedback. 12.5 vdc
I guessed that there was need for a heater voltage at each end and plenty of EMF from end to end.
In those days, old TVs could supply ferrite cores from cannibalised, line output/EHT transformers. One transformer supplied 2 cores for this job..
I got the LT winding turns from a circuit for an HT inverter for a ham tx.
The over-wind was a "that looks enough" exercise, although it wasn't... and trial and error
played a major part. The same applied to the heater winding but I eventually lit 4 and 6 foot tubes.
Some interesting things emerged from this.
If the system set off on a fully charged car battery, no apparent reduction in light output
occurred until the batt volts dropped to 8 or 7 volts or lower. But don't switch it off or it won't start again until recharged!
The waveform interfered with MW/LW radio just as much as it did when the ferrite
was inside a mono-tv.
Couldn't make them fast enough during the 3 Day week.. Never made another when it
stopped!
I worked out that it was running not at 50 hz but 6/7 Khz..
I was later informed that running at higher freqs produced more light for a lower DC. consumption. Working, not scientific tests showed this to be so.
Who informed me of this and other aspects?
Raymond Baxter on Tomorrow's World with an item in his show relating to this break through that scientists had made for energy saving....two and a half years later.
Thanks for the mention Raymond!
Appreciate you explaining failure modes.
Hi Clive, I’ve had to deal with all the problems you mentioned over many years & when i was an apprentice i loved working on fluorescent signs & fittings but none of the sparks that i worked under understood how they worked or could explain to me what went on.
I used to dismantle the old fittings to find out how they worked - later on when i worked on many TV Rental shops i would supply & install hundreds of feet of back light fluorescent tubes to the shelving that ran around the shop - at the time it was very futuristic.
I still like fluorescent problems & can deal with them all when called.
Edward.
Great video. You're very good at explaining this stuff, so that even guys like me without any more electronics knowledge than what you learnt in high school can understand what you're basically saying. And it's not always easy to be an expert and explain things clearly to a layman.
Great video with lots of information. I love gas discharge lamps and am sorry to see them being phased out in favor of LED, even when in some cases there are few benefits.
Also worth pointing out that the metallic mercury vapour in flourescent tubes is a great deal less harmful than organic compounds of mercury, which is the really nasty stuff that requires very little exposure to cause serious health effects. Most people don't understand the difference.
People don't explain fault situations like this often... fucking great stuff man... I appreciate it, I'll have to watch it again and take notes.
Edit: I need more of this stuff... this is awesome.
Thanks Clive for clearing up a matter that has had me wondering for many years. I've been using various methods of 12 volt lighting for most of my life. There was one particular type of 12 volt fluoro lights that I bought, where the tubes only lasted a couple of weeks to a month with daily(or nightly) use. I now understand why. It was a double fluoro with 2 x 8 watt tubes driven in series. The circuit itself didn't look much different visually to that of a single 12 volt fluoro. Obviously, it must have driven them quite hard. lol. It was a bloody bright light too, it's what disappointed me when the tubes died so quickly.
I really learned so much! Thanks so much, Clive. I knew I knew nothing about florescent tubes, and never got around to learning. I'm chilling at home waiting for the missus to come back from visiting a friend and your new videos made me afternoon great!
Than you Clive for explaining this "rectifying effect" I always wondered why tubes that are at the end of their useful life sometimes go into 50Hz operation!
Nice video. Merry Christmas!
In my fledgling Electronics Geek days I took a 12v relay and wired it so when the coil pulled the contact in, it disconnected the coil and when it returned, it pulled in, essentially a mechanical oscillator. Fed the connections across the contacts into a step up transformer and the output of the transformer into a 4' florescent tube with the filaments wired in series. Amazingly, it lit the tube. I know there are so many reasons this isn't a good idea, but I look at it as a learning experience. Many years later, you explained how they are supposed to work and why.
Rich Booth even i had this idea but didn't get confidence to do the work
Great video!
I'm glad you mentioned the need for a grounded reflector with some bulbs. I discovered this one morning at work (broadcast engineer) when the background lights in our TV studio weren't working. I took the trouble call, grabbed my DMM and a screwdriver and went to see what was wrong. There was about twenty 32 watt lamps and none were working. The voltage into the ballasts was OK. Thinking that the bulbs weren't seated properly, I grabbed a bulb to remove it and it lit up! Each bulb lit up as I touched it.
It turns out that an audio engineer was troubleshooting some hum issues had disconnected the ground wire to the set.
@bigclivedotcom
I am an electrician from Sweden and I do appreciate the topic and I do understand what you are talking about. I think the Starter you mentioned during the first start-up is the 'Reactor' as proper term. Non-electricians call it Glimmtändare though.
The second circuit, I have not really seen. Maybe I am too young to really see them since Energy efficiency in light and the ergonomics of it is fairly big subject in Sweden both for us electricians and those that project(Plans bigger jobs). So it is generally either two different topics of subjects regarding starters now days.
It is the Reactor and it is the HD-reactor. The latter being the electrical version. On the topic of light bulbs, both business and electricians in general tend to recommend you burn the tubes in for roughly 1k hours for the best light. Obviously not a topic you bring to a private customer that won't appreciate paying such electrician bills but I thought it was interesting to enlighten you.
The reason, I am not really sure but I assume it has something to do with the gases within the tube along with the potential Mercury vapour to settle into its environment and get used to the equipment.
But this is hardly the focus for Electrical. The advantage lies in the shifting of the Sinus curve. The faster the shifting, the less load on the bulb but also better for individuals in general whom work under this light. You said one could see shimmering/flickering. Technically, you can still see this on a good bulb as well if you have good eyes and science has proven that 60hz is a reason many people in office environments etcetra suffer headache and lack of ability to think since the brain unconsciously notices the shifting light that seems constant to you and therefore uses a bit of brain process for that part. How ever, once you exceed a couple of thousands, I think it was 5-6k hz the brain simply stops thinking about it. Many people seem to work better under such high speed frequencies. Oh, a warm light is also generally a colour to not use in office environments as well. Cold light produces less depression like effects.
But out of company aspect when creating these electrical circuits, you generally want 10k since it is easier to calculate and construct money wise. Don't ask me why but perhaps because it is an even number and they can limit the electrical parts? Ah well. Now, it comes to a topic that is generally a discussion between electricians in sweden and customers. Both private and business alike, it is when it comes to the HD-reactors themselves.
On certain brands and models, the HD-reactor dies quickly and on some it doesn't, generally the reason is price and where they bought it from. While everything else is fairly simply and straight forward, it might be a cheap China brand which is known to be a bad product in Sweden. We have highest electrical standards in EU to my knowledge. I wonder, what you think is the reason why some HD-reactors burn out quickly while others don't? IS it potential damage in shipping or handling? Or is it simply product deficiencies?
Also, for LED, you should preferably don't use them in bulb form on a location where you work under many hours. They generally do not offer good long term effect on your eyes. Although you might not process it, your brain will process things unconsciously even for LED. I suggest simply buying an overpowered Light fixture and simply dim it down. Lengthens the life of the bulb and improves general long term effects. Especially good for detail work where LED is simply inferior still. LUX being the reason to my knowledge. LED has too much, direct light in comparison to normal fixtures is my theoretical reason also something to do with colour but that is beyond my ability to comprehend.
Either way, long rant but hope it was an interesting read.
EDIT; LED is useful in colder environments though, it is fairly easier to start up in lower temperatures.
Regards.
The best easiest to understand video on Fluorescent lighting I have ever seen.
Wow Clive. You've come so far in four years!! Your presentations are so much more slick now. Well done!!
Thanks, we value the time and effort you spend for this channel
older vid i know but commenting anyways. this basically perfectly explained what was wrong with my bathroom light where I had no clue how to fix it. liking these videos more and more, so much learning :D
Great topic, man I know a lot about this but it is always good to get e refresher on it.
When I was a much younger lad (1963) I had a television transformer that had, I believe a 750v winding. I took a 40 watt tube and powered it on, albeit very briefly, with said Xformer. Talk about a flash tube... It blew the mains fuse and all I could see in the dark basement was the retinal burn of that lamp. The lamp was now ALL black and made plinking sounds when shook.
Ballasts are very unique devices and you are correct that the topic gets going on to know about them all.
An electrical engineering text I had, spoke to the development of specialized high frequency gen sets for office buildings fluorescent lighting that would have yielded more lumens per watt.(over 50 or 60 Hz) I do not believe this was a commercial success. Fast forward 5 decades and the compact fluorescent is indeed a marvel of efficiency
BTW
Merry Christmas Clive
+Craig Nehring The old TV transformer may have been an auto transformer with the higher voltage referenced directly to the mains. That way, the high voltage would strike the tube and it would then try to crowbar the mains.
+Craig Nehring I am glad you mentioned that, I'd forgotten they tried to run at HF. Weren't they 400Hz? (that way transformers would have been readily available from the aerospace industry)
Pretty sure that transformer secondary was isolated and just dumped the full output across the lamp, then again it could be as you say.
It was all over in literally a flash.
Back in my early experimental days I would try and run things backwards, for example using an electric model train transformer I would output that low voltage AC into the various winding's to see what voltages I could get.
One thing for sure there was a substantial current flow through that lamp.
Cheers
*****
Aircraft do run at 400Hz for good reason the magnetic cores can be smaller and lighter, not sure what freq they had hoped for on the genset in office buildings but the CFL I believe is running at 10kHz.
Most aircraft also use 28 VDC for the same savings, wire capacity is thinner and motor are smaller. Aerospace hardware is very spendy: certification, regulations et al.
I used to do component overhaul on some Lockheed and Airbus craft.
Confused...? 21 years since I have been involved with electrons, professionally, but surely the lower the volts the higher the current requiring more copper. ? Higher the Hz higher the impedance, good, less copper. ?
Some of this was over my head, but still very informative. I've seen all of the problems described, and now I have a better understanding of what causes them and how to address the problem. Thanks!
Thank you for the awesome video :) The videos you make are always packed with so much valuable info. I have watched so many and actually learned something that i still remember and use so often in everyday life. So Thank you, you are doing a valuable service and should be commended on it!
Even as a engineer in the lighting industry it's interesting to have this explained. Thanks.
This is a very thorough and explanatory video. I'm not sure what the UK standards are, but here in North America our fluorescent lighting systems utilize a special transformer called a "ballast." There are two types of ballasts: magnetic and electronic. Magnetic ballasts are essentially a choke connected in series with a starter, although newer magnetic ballasts often have built-in starters. The only downside to magnetic ballasts is that they generate A LOT of heat, especially when left on for more than 6 hours a day. For this reason, many magnetic ballasts are equipped with thermal protection so if they should overheat, they will shut off and restrike once the fixture has cooled down. However, this isn't always the case, and older magnetic ballasts were sometimes blamed for starting fires. There also have been instances where the thermal protection has failed, causing the unit to become dangerously hot. In 2013, the government "phazzed out" the manufacturing of magnetic ballasts.
Electronic ballasts are much safer and more efficient in the long run. In fact, many modern electronic ballasts are instant-start and capable of driving 2-pin tubes. I have some in use and I've noticed that the tubes are "sputtering" much quicker than with the old magnetic ballasts. There is another alternative, a programmed-start electronic ballast, that starts the lamps in the same manner as the previous magnetic ballasts. These are often a bit pricier and may be somewhat difficult to come across, although they do not "abuse" the lamps like the instant-start electronic ballasts.
By the way, mercury is NOT dangerous! The government overemphasizes safety when it comes to handling fluorescent lamps. I have shattered many fluorescent tubes, including a shattered CFL in my den. Not to mention I also played with mercury when I was only 6 years old...
Thank you. That was an excellent introdction. The books normally just explain what you should do but do not expose the consequence of not doing it the recommended way.
Thank you for taking the time to make this video - I thought I was the only fluorscent tube nerd on the planet.... :-)
The one thing I've never understood was the weird faults you see on buses in Edinburgh. When the system starts to fail the tubes will appear to have flowing light and dark areas moving from one end to the other, like a good old rope light sequencing. What's weirder is that when I spot this there's usually three or more tubes doing this in sync.
I always thought it was an effect from a high frequency power supply failing.
Thanks for the description of what appears to be happening in their lob-sided power supply.
Happy 25th December.
+PIXscotland
We have a tube in the basement that does the same. It has darker bands (didn't really measure, but maybe approx. 3-5cm apart) that slowly move in one direction (again, didn't measure, but it's also in the order of 3-5cm per second). The tube does not have an electronic driver and the banding vanishes after 5 minutes or so. If I had to guess I'd say it has got something to do with the gas heating up, expanding and contracting with the drive frequency near the cathodes of the tube. This could lead to some kind of standing (or slightly moving) pressure wave until all of the gas in the tube has heated up uniformly.
+PIXscotland If they are being driven from an electronic supply you may be seeing a standing wave effect called jelly-beaning in the neon industry. There used to be electronic transformers that could be accurately tuned to cause jelly-beaning deliberately.
All these years sat not interested in school work and staring at the florescent tubes makes sense now
Real good video. Thoroughly explains why my 12V floru drivers i built 20yrs ago is still going strong, yes the filaments at both ends also driven with low voltage, the extra windings on the TX as you show. Additionally used another current sensor Tx for f/b. Used a 555 to drive a mosfet.
Hey BC, ever play with "No Cathode" fluorescent tubes? They last centuries. Or, about 25yrs in continuous operation. Eventually the phosphor starts to dim out. With no electrodes inside to trap gas ions or sputter a black stain, the main failure mode is removed. The commercial type is the Induction Bulb, about $30 on eBay.
Or, make your own 1890s Nikola Tesla-style bulb by wrapping adhesive foil rings around the tube ends, placed on the glass a bit forward from the filament location, positioned so the plasma doesn't touch the unused metal parts inside. Attach a CCFL driver to the two foil electrodes (use those little 2-transistor drivers found in cold cathode LCD backlights.) A small conventional tube will light up full brightness. No electrodes, at least not inside. Watch out though, since the drive is about a kilovolt, and any sparks can be a burn hazard!
When CFL spiral-bulbs were first announced in 1974 Popular Science, they said they were going to be powered capacitively by high-volt 27MHz, lack electrodes, and last for over 100 years. Then, nothing. No peep about them for years. When they eventually appeared, they weren't like those glass-capsule electrodeless "RF lamps" being promised. Instead they were just conventional fluorescents, but bent into a spiral. And, still employing the interior metal electrodes which get destroyed after a few years of operation. True "RF lighting" does exist, and it lasts far longer than HID, longer than LED. Not sold in stores everywhere! Buy it now, NOT!
We can get the fluorescent forever-bulbs from offshore suppliers via eBay. Still, those have an induction loop coil and expensive ballast driver circuit, and a torus-shaped glass globe. Not the simpler capacitive coupling done by metal rings outside the glass.
Oh, and about "mercury freakout?"
This used to be a real issue back in the 40s/50s when lamp internal coating included BERYLLIUM OXIDE, and the white dust from a shattered lamp was carcinogenic toxic inhalation danger. Not the mercury. People in the early-1950s factories were dropping like flies from beryllium poisoning, corporate coverup, big scandal. And everyone was supposed to leave the room if a tube shattered, until the BeO powder could be washed up. But then toxic fluorescent powder was replaced by other materials.
Yet everyone for decades later still goes crazy when a fluorescent tube gets broken. Very silly. But we can see where it all comes from, as the older "dire warnings" get passed down the generations, even though the lethal hazard was removed fifty years back.
+wbeaty I have played with the electrodeless lamps. In fact I burst a full-price expensive one when they were new while trying to open it. :( There was a company selling DIY neon signs for a while that used electrodeless letters that clipped onto a high frequency exciter backplane. I also recall that my first EPROM eraser had a totally sealed tube which got its power capacitively coupled from two ordinary fuse clips. Then there are those intriguing linear lamps that are basically based on xenon tubes, and have one electrode at one end and a clear conductive coating.
At the Edinburgh Tattoo we switched to using induction lamps for the house lighting. Fast warm-up time and low enough power to run on a battery powered backup system so we have full output in the event of power failure.
+wbeaty the mercury tubes do severe damage on submarines, long ago when I was in navy they had florescent tubes on subs and it killed one entire ship one lamp breaking. nice to know about the beryllium, thank you for that
+wbeaty are you talking about one of those "contactless" induction lamps?
+FoodOnCrack Nope. Tesla's lamps used external metal plates, capacitive-driven with no need for a glass torus. Demo: hold any discharge tube near a Tesla Coil, even one lacking electrodes, and it lights up. The version used in Tesla's laboratory ceilings were phosphor-coated capsule-pipes, with external metal near the ends, driven by HV at KHz. No-cathode, rather than cold-cathode. Last's nearly forever, eliminates the need for billions of dollars of "re-lamping" and hazardous waste stream. Mod your conventional fluorescent tubes by attaching foil electrodes outside.
But Tesla used all his outside funding to build his Colorado lab, when he'd promised investors that he was starting a fluorescent lighting manufacturing plant.
Coulda been like Elon Musk of 1895! Wiped out Edison corp's incandescent lamp sales, become a billionaire, and started any future projects he wanted. Oops.
Funny, he mentions vehicles at the end and that is actually what came to my mind while I watched the video.
Super video - you give others too much credit. I never found anything nearly as detailed as what you just went thru. Thanks for the share.
I had no idea there was so much to these little buggers, cheers Clive
Absolutely brilliant. Great vid. I have those very same tubes in my camper. Replaced them only 2 weeks ago. Worked fine until this morning: now the tubes are dim, slight purple colouring at the ends. I imagine the ballast units are defective. LED strips here I come...
I love the way you describe all the failed design to strike the tube.
As it turns out the newer state of the art driver is called a "Programmed Start"... it makes the tube last 20Yrs with unlimited On/Off cycles - Lots of ballast manufacturers make them, easy to buy on eBay.
Basically these ballast glow the filaments before striking them On with HV, then gradually power down the filaments as the tube keeps runing - No more black ends, endless life: Perfect!!!
Excellent video! No idea why I didn't see this in December!
I had a few sets of these T5 (maybe they were T4?) in varying lengths as under-cupboard lights in the kitchen; some daisy-chained, all 240V... The tubes "blew" with alarming regularity (at least one every 6 months!), and unlike the larger T8, the fittings didn't have a separate ballast / starter. Almost every time I visited the *same* local shop to replace the tubes, they ended up being a different manufacturer and slightly different dimensions (actually within a year or two even THE SAME manufacturer, had different dimensions too!) with each tube being a few mm shorter or longer... So not only were tubes replaced, but the fittings were too...
I gave up the last time - and instead purchased some LED strips and 12v transformers, which I have fitted. They give far better light, show no signs of becoming unstuck, are run cooler and use less power.
The T8s in the garage I also replaced, but with "replacement LED tubes" - which also replace the "ballast/starter" with a LED compatible driver - so are "drop-in" replacements.... Oh - the joy of "instant light" in the garage. :)
+Sean Pearce Quality T5s are supposed to give LEDs a run for their money on lifetime.
Hey Big Clive. I wanted to point out something since you mentioned it.
I just left the semi/truck driving industry, spent a year driving trucks of the likes of Internationals/Freightliners/Kenworths. I am happy to report that the manufacturers are moving towards using LEDs in more of their rigs.
Cheers, mate.
CLive, i used to repair the 24v lighting ballasts used to light up tubes in coaches and buses, it was part of being the electronic repairer of an auto factory.
The transformer with heater supplys was the exact way the ballasts worked, they operated at 20khz and a push pull two transistor symetrical oscillator with feedback windings and identical components was the way it worked.
They worked really well :-D
Your discription is spot on, no supprise there :-D
Very Interesting Subject Clive. Awesome Video. Thank you. Keep up the great work. Nick.
Once again, another amazing video to learn loads of information from. I can't watch one without learning something to help me with my projects. I have to say, this is likely my favorite channel on youtube.
great video! finally a simplified way to explain cold/warm cathode lighting
And now I finally know how those old starters work. Thanks!
Wow that video was packed tight with information and very well structured. You certainly know your fluorescent tubes, Clive. Lot's of interesting physics going on there.
Interesting. I didn't know these were warm cathode tubes. I have used various fluorescent tubes like this for moth trapping since about 2009 and never really looked up how they actually worked. I mostly use the high output t5 12" 15W visible blacklight tubes that produce a visible blue light. Since the end of 2014, I moved to using t5 12" 15W synergetic tubes, and these produce an almost lime green light. These produce two peak wavelengths of light at 368nm and 540nm, which is broader than the blue tubes which peak at 350nm, and as such they work far better at attracting moths than the blue tubes.
I run them on the battery powered car inspection lamps that use the same 8W white tube you have, just switch out the tubes for mine. I power them on 12v 24 amp motorcycle batteries and I can get around 19 hours before the battery needs charged. However, after a long enough use, the plastic tube that protects the florescent tube, does turn a dark dirty yellow colour. Maybe because of the UV light the tubes I use, produces compared to the white tubes that come with the inspection lamp?
I was waiting for breaking glass @ 4:40 :D
Thanks BigClive! Awesome video as always. Between this one and the GFCI video, two things I've always wondered about have been explained!
We purchased a house in 1964, which has a storeroom with a fair stock of spare fluorescent tubes. Over the years we used them to replace the bulbs in our kitchen, as they failed. These old tubes lasted for many years before they finally failed. Any of the newer tubes generally last less than a year.
Planned obsolescence? Hard to argue against the evidence.
See "The Light Bulb Conspiracy" documentary about planned obsolescence...
Flourescent tube that dies in a year is insane thing. If it was running 24/7 it can run more than 5 years i dont get it
Must be instant start ballast and awful ton of starts?
Thanks BigCliveDotCom. This brought back memories from 1986-1990 when I designed emergency lighting systems for every tube from 8w T5 to 100w 6 footers. I am not that clever, so we did it by reverse engineering the transformers and pcb's then melded different circuits for our needs and CAD them on a BBC Model B computer. We undercut the Chinese and Germans on many products. I also did mercury lamps, Disco lighting and strobes. Never had many fails though and they went into many new UK buildings! Remember that Jon Kay?
That clearly explains the problem I had with battery operated fluorescent tubes in the past as they kept blowing the tubes. The 12V models were even sensitive to being switched on & off. E.g. if switched off and back on within a fraction of a second, it generally caused its electronics to self destruct - a trail of smoke and that burnt electronics smell.
clive, you help me understand all sorts of things. thank you so much. Please keep making videos.
One thing that I learned about learning is that...
The more I learn about stuff, the less advanced I realize the human race actually is. Things are too alike.
I mean, it seems complex, but once you get it down or hear enough about it, you are like, duh. This other thing actually works just like that too.
And mostly, the thing with me about stuff is, I generally have no clue because it never came up in my life. But the second it comes up, I can figure it out. Almost like I knew about it for years already.
Some things do not stick with me though.
Like, is Psychology. I didn't find it important to remember who said what. All I cared about was why it was said. So when it came time to recall the dude, I was like, why?
But yeah, names and flashy words don't stick with me. But, when it is time to learn and develop new things, I am generally the first one with an idea.
Fantastic! Thank you for some answers to my lurking questions... when I see odd florescent behaviour like darkening or purplish ends... humming and lopsided operation... now I know why. You made the answers fun and easy..
Happy Christmas Big Clive
I like the way Clive says "tube"
I always was wondering how this worked. Thank You!!
Thank you so much, you have a really good way of explaining that makes it so easy to understand.
At 4:35 I was waiting for the crash as the tube rolled off.
Some older tubes had a phosphor containing beryllium which would be hazardous if broken but these tubes were discontinued years ago.
Back in the early to mid 1980's, my grandfather used to turn the heat off in his garage, next time he would come in and turn on the lights (USA fluorescent light bulbs) they would only come on dim of if used his cigarettes lighter to heat one or both ends, these were usually, 4 to 8 ft, why did this help, and does this still work to-day on straight tubes and curled household fluorescent light bulbs.
Thank you Dennis
+Dennis Friedman In very low temperatures the mercury will condense out completely and the tube will just glow with the carrier gas. Heating the tube will help get the mercury back into vapour form. You get thermal sleeves for tubes used in refrigerated cabinets.
I love those preheat fixtures with starters, I actually have one in my basement with a rapid-start fixture, the preheat fluorescent in my basement is a sears fixture
I've always wondered why some tubes have a moving pattern in them when lit, (the light is on but not quite as bright as normal but with a snaking pattern that shouldn't be there). Nice one Clive.
+Graham R Dyer This often happens with new tubes randomly throughout an installation, but usually stops after a while. (But not always.) I actually like the squiggling effect. I've never come across a decisive reason for it though, other than "impurities in the tube" which may be residue from the electrodes.
The neon industry tried to recreate the effect deliberately for many years without lasting success. You can do it with short xenon filled tubes and high frequency drivers though.
I build machines that made those lamps and you telling the right story. Good job ashame that they are slowly replaced by led.
***** led is not better
Its a strange color and the electronics heat up and dies after a while
+J IJzer No one even noticed at work when we switched to LED... except the power company who halved our bills.
What a knowledge on that topic ... Impressive
Opened my eyes to some of the complexity.
I thought I knew how they worked but like Jon Snow - I know nothing, although a little more now.
Brilliantly clear and informative - well done
I remember having one of those battery-operated flourescent lights at one time, it took 4 C or D cells, and never did work really well. At this point I'm not sure what happened to it.
I used to have a couple of 4-foot fixtures hanging over my workbench. These worked fine, but there were times when I noticed a certain "fuzziness" on my scope. Looking into it a bit, I found that the fixtures apparently had electronic ballasts in them which were radiating at around 38 KHz or so, not exactly the same frequency, but pretty close. After a move I have an LED bulb hanging over my bench, and those fixtures are in the garage, where they will probably get put up at some point.
In recent months I have scrapped out a fair amount of gear, including some flatbed scanners and some flatscreens. The scanners had small flourescent tubes in them, and in some cases I also snagged a driver board to run these. I'm not sure what they want for power, but am guessing that 12V is pretty likely. What's done in flatscreens can get pretty interesting. They have inverters of a sort, and I've salvaged some of those transformers to play with. Apparently the logic of the way those work is that if there's any kind of an issue with either the lamps or the inverters, the set will shut things down, rather than letting it be obvious as to where the fault lies. I haven't figured out a way to troubleshoot that yet, and probably won't pursue it much further as the newer stuff is all using LEDs anyway. I do recall seeing a gadget or two on ebay that was supposed to let you test various bits of this stuff, but never pursued it, and at this point probably won't...
I use as my desk lamp a 12VDC 36W fluro. it uses an AC inverter(electronic starter+ ballast)(That is the name given the electronics by the manufacturer). This thing uses standard 240V AC fluro lamps and doesn't blacken the ends. Anyway the things that are of interest to me are the fluro fittings that are dimmable.
You Sir, are awesome at explaining this. You should explain other things.
Thanks a lot for this.
Please make a detailed video on how today's uv ballast works. Like Philips, osram or other brands.
Btw This is the best video on uv light i have seen on RUclips.
I had one of the more weird electrical failures today. In my kitchen I have 2 compact florescent lights that hang over my counter top. They are connected to a dimmer and a aux switch (the cf used were rated to be on a dimmer circuit). I think installed them about ten years ago. The light bulbs were off and I went to turn them on from the aux switch. I could see a orange glow from the starter on both lamps but they did not start, after a second I heard one lamp pop and I turn off the aux switch. I went to the dimmer switch and saw it was on its lowest setting. Sometimes the lights don't turn on at that setting at first but eventually or if I raise and back down they turn on. So to see if I just burnt out one I raised to the max setting and turned it on. One second later the other bulb failed. White smoke and electronic smell. They both made a mess with white powder as I changed them out. Replaced one with a cree branded led from philips and it is working fine. Have to go buy another bulb but I find it strange that they operated for approx 10 years with no issues and been started in that condition many times and both failed in the same night.
Very interesting Clive ty once asked an electrician friend about this , but you know what the average electician is like ;-) I got the vague no idea answer, although , to be fair he was an apprentice at the time, but yours is the best explanation I hVE SEEN :-)
any chance of a video on metal halide lights?
Fantastic video!
I bought a cheapo 12v fluoro fitting (one of them sub-£10 things with a prismatic diffuser) to use in my shed in addition to my LED tape lighting last year, it barely lit up and was out-shone by my LEDs, so I gutted it and fitted the inside of the thing with leftover LED tape and now it works great, not really fixing the fitting I know, but, it works for me... :P
Absolutely fascinating video! Very well explained too, I certainly learned a lot tonight! =)
only this guy could make fluorescent tubes interesting XD
... and I thought I knew everything about fluorescents! very good!
Great video. We have starter lights in our bathroom
excellent! I now understand the parts of these fittings - cheers
Thanx about that part about symmetric and non symmetric waveform...
i never thought of something coming out of secondary of transformer could be non-simmetric...just dumb i know ;-)
it explaines so much weird things i observed i a quick and dirty PSUs i built for fluo. bulbs in a squatting complex without el. power where we run everything on deep cycle lead acid (getting slowly replaced by li ion/poly from laptop packs) + solar and/or gasoline generators...
ok we switched allmost all lights to dumpsterdived LEDs(from those powersaver bulb recycling containers we have at supermarkets/shopping malls, where all the good stuff is, like phones, tablets, powerbanks...
As they are big/deep bins you have to lean in with flashlight in your mouth/on forehead, like aladin's cave full of broken glass shards, magic smoke smell and mercury vapours hovering near bottom, where all the good/dense stuff is ;-), but anyway, it explains those WTF how is that possible at all moments...
Very interesting video. Thanks
I have just started to replace the fluorescent bulbs in my 4ft fixtures with LEDs. The first ones I replaced use the ballast. The next ones I replaced did not need the ballast, which was my desire since the ballast was not running both the bulbs.
How about a video explaining these two different styles of LED replacements.
I do love the replacement LEDs. Instant start and the one which does not need the ballast has no hum. Also much brighter since all the light is shining in one direction
More Sudden Science style videos please! Really enjoyed this informative video. Love your channel. All the best fir the New Year!
Used to get a lot of the Tubes on Southwest trains when half of the bulb rectifying and only half of the bulb lighting up. Also london underground used to have tons.
That was common on Glasgow trains too. Probably mercury migration through a non symmetrical waveform.
You forgot the tiny amount of Neon as well, droips the striking voltage. The mercury dose is under 4mg, or basically what is in a can of tuna. When the mercury all gets absorbed in the phosphor coat you get those tubes that are lit slightly but the middle is pink from the neon fill. Also happens on new tubes that have not heated up enough to boil the mercury fill out of the little glass dosing capsule, they start off new as pink then get to full brightness over 10 minutes.
With the older starters they also had a tiny amount of Amerecium added, so the neon would ionise, otherwise they would not start in a totally dark room until you had a stray cosmic ray pass through the starter to ionise the gas, then it would light.
Wow thanks for the great video, our RV has many of these 12 volt florescent lights throughout the 35 feet of home. We spend our winters living in her, and love the travel and adventure. We rarely use the florescent, just at bedtime, as they are handy to turn on to see the way, and have local switches so you can turn them off as you pass. If we did more boon-docking (camping away from shore power) I would consider replacing them with LEDS but we are usually hooked to 50 AMP 120 Volt shore power so don't have to worry to much about draining our two 12 volt house batteries. Nice to be able to understand what is going on with them, I have been maintaining such lights off and on most of my adult life, so I am familiar with the starters, but always wondered what was going on with them. Now I know, thanks to you.
As long as the flouros are working and arent causing any trouble (too much current draw), I would keep them and replace them only as they die.
Duuuude! Very well explained! Sputtering, rectifying effect, high frequency operation, couldn't think of something missing. OK, maybe resonant operation and stuff like striations... ;)
Thanks for the video.
Have you ever seen starters with a bulb attached to them? We had a couple of fittings in our house in the 'sixties. They were 4' and took 2'' tubes. At one end was a housing that took the starter. When switched on, the bulb would give a constant light, no flickering, whilst the starter made it's attempts to get a strike. By light output, I guess the bulb was about 40w. Although the bulb was constantly on, if the filament in it blew, the rest of the starter wouldn't work. So it must have had a purpose. The casing was like the old metal starters, with four pins. They became near impossible to source and so the fittings were changed out. I wish I'd have kept one now.
Well, from what I have surmised, the folks at the space program are not afraid of Mercury because they are planning a trip to it. LMAO 😹😹😹
Very interesting. If you do a follow up, please cover the various ways of dimming lights and which ones are better suited for this application.
Thanks! This was genuinely very interesting! :) I have a bunch of those handheld fluorescent lights, bought them on a outlet department in a local hardware store once. They use 2 AA batteries and many of them came with the tubes already considerably darkened at the ends, so it's probably safe to assume they use the bad way of driving the tubes.
I really like your channel, but I would love it if you started doing more of these explanatory videos!
Keep up the good work and lets get you 100k subscribers!
Thanks Clive
When I was 7 to 9 yrs old, we have a round florescent overhead light fixture that would only sometimes turn on. What I had found was if I flicked the switch a few times, it would eventually light (key word there: EVENTUALLY - not always as sometimes I gave up).
As a child in 2nd grade of school, my mom was not eager to get a ladder (not sure we had one) to have me climb up and look at it. To this day, I would tell you that the starter (that small round can with the neon tube) needed to be replaced, but in 50 years, I've never discussed it with anyone who might know. I know you could see the neon bulb glowing (but don't remember if it was blinking or just on solid).
Was my 9 yr old brain correctly guessing what needed to be done?
It probably was the starter.
4:38 the tube was like “I’m outta’ ‘ere”
Don't know if it's already been said, but now I know why some of the really old tubes had that black "stripe" down them! I'm assuming it's a conductive strip between the two metal caps on the end that would help with the strike :) - thanks Clive!
Yeah, some of the very open fittings needed those tubes with that stripe on them. Especially the 8' tubes.
all good stuff wound numerous transformers & made 12v-24v tube fitting, drivers & yes push/pull drivers work well & have less rf interference .
The shown color temperature is intresting. Normally in the U.S, the color spectrcturm starts from 2700k-6000k. In other countries it could be a little different. But very interesting video.
Nice, could have mentioned or touched upon the beginnings, namely why we have tube lighting in the first place and that without Tesla, we would still be mostly in the dark ages.
I think you said the main purpose of the inductive ballast was current limiting for the filaments. I also generates a high voltage pulse when the starter open that helps get the gas ionized. If you replace it with a equal value resistance the filaments would heat but the lamp would not start.
technically, you do want to leave the area and ventilate if a tube breaks, mostly to allow the vapor to dissipate as even in small quantities mercury is harmful. however, you wouldn't need to call in hazmat or panic, only calmly leave the area for a few minutes