Ferroresonant AC mains voltage regulator 2

Even more experiments with my German AC mains voltage regulator from 1965. The input is 220V 50Hz +10 / 20%, the output is +/ 1% accurate 160V / 200V / 220V 50Hz, 300VA continuously, 350VA peak. I do some oscilloscope measurements of the waveforms and I try to reduce the harmonic distortion of the output voltage. I also explain various versions of ferroresonant voltage regulators. I talk about the isolated and non-isolated versions, compensated and neutralized versions and also about its single core version - a constant voltage transformer (CVT).
The previous episode:
• Ferroresonant AC mains...
Please support my channel on Patreon:
/ diodegonewild
Instagram:
/ savage_danyk

[@adriansdigitalbasement]

To me this really is magic. Thanks for the explanation on how it operates. I don't really get it but that's OK, it was still fascinating!

[@xxexplosivexxxxexplosivexx8512]

Wow I wasn't expecting to see you here!

[@roseelectronics4582]

I don't wanna watch it because I know I won't get it lol

[@msansjr]

This specific type of magic was too much for me and was the reason I dropped Electrical Engineering college and decided to switch to computer sciences :D

[@uK8cvPAq]

Think of it as a magnetic circuit.

[@lakiza55]

Dear lord, i can't imagine the amount of calculations and designing that went into the damn thing. To top it off, this was in the 60s, they didn't have computers to simulate the design, most of it was probably done through trial and error.

[@user-ml8wm2nl9w]

Technically speaking they had computers... WITH MAGNETIC RAM/ROM (They had to be assembled by hand by weaving ferrite cores into a copper wire matrix)

[@windshield11]

Bro, they used slide rules for calculation, it is probably very simple as they made the everything in-house and had huuge knowlege on the subject. But for us, ALIEN TECH!!!

[@westelaudio943]

They did have transistorized calculators in the mid 60s but, those things were probably much older designs. The underlying concepts (LC filters/resonators, transformers and saturable reactors) were already very well understood by the 30s and calculation was pretty straight forward, so smartly combining those things should not have been a big problem for experienced engineers.

[@laharl2k]

It was a lot of work and math but at the same time the pay was higher so it didnt matter that you as an engineer took 3-4 weeks to design just this thing, you were still getting a promotion.

[@rtyzxc]

Everything nowadays may be made with random experiments without understanding, but physics and math were invented way before computers and they can be used to calculate these stuff.

[@Miata822]

Absolutely fascinating. Understanding magnetism gives my brain a cramp but it is at the root of how everything works. If is amazing to think of the engineering involved in designing such a product.

[@SeanBZA]

Good thing with the distorted waveform is that on the secondary side a diode rectifier will have lower peak currents, and a longer conduction time per cycle, reducing the stress on it. As a charger this means you get to have a higher current charge, without having to have a diode bridge that has so high a peak permitted current, so you get by with a smaller diode. Reduces harmonic generation as well.

[@windshield11]

Hillariousley true XD makes it into a squarewave

[@RavenLuni]

Even though most of this is beyond my level of knowledge, you still explained it in such a way that my brain did not overheat (just got slightly warm). You made it clear which parts of the circuit were having the relevant effects and I understood everything you said. Now I want to learn more about this subject and perhaps build a small scale model of an AC regulator to boost my own knowledge. Truly an excellent video and you are an excellent teacher. Thankyou!

[@Basement-Science]

The color code is a very old one from Germany. It looks like someone probably rewired the input/output cables at some point though. There is a dedicated screw with a ground symbol where you are probably supposed to connect both PE connections with ring terminals. So someone rewired it but didn't replace the cables. Pretty dodgy.

[@netman69]

This thing is just endlessly fascinating and you did a damn fine job explaining its function too :). I'd never think you can do all this with just magnetics and a capacitor.

[@Shmbler]

I haven't felt as dumb as I do now for a very long time.

[@andymouse]

This is up there with RF for me, total black magic, you have to know your onions to figure out how an ' air gap 'is gonna effect things, indeed a " certain level of insanity " is required, you appear to have reached that level, more of the same please !...cheers.

[@cyrusshirazi-od9xr]

With just couple of components, creating so much havoc. Excellent .

[@ebasoalto]

Excelente análisis y explicación de este interesante equipo, que usa las propiedades de ferrorresonancia para estabilizar el voltaje de salida, saturando parte del núcleo. Tecnología antigua pero muy confiable. Gracias por el magnífico video, como muchos otros que ha publicado.😊👍

[@tigerelectronics5966]

These are absolutely fascinating units. I recently aquired one of these myself, a 1 KVA fiskars unit! I am using it with my computer and older audio gear that does not like running on the modern 235-245V mains in my area. (the B+ voltage becomes so high that the bias gets thrown way off, and it is also unhealthy towards the rectifier tube and I have noticed the power transformer in the amplifier running EXTREMELY hot at 235-240V. at 220, it barely gets warm and the amplifier operates within it's proper B+ and bias.) However, I have noticed the ferroresonant voltage regulator running extremely hot under low load, or no load conditions. to the point of me worrying about it melting down internally... It runs at 90C, on the outside of the transformer core, and 127C on the windings inside, measured with a thermal imagining camera. after about 2 hours of low load or no load conditions. I do not dare leaving it plugged in with low or no load for any extended period of time. I understand from watching your videos that it will run hot with no load because of the core being saturated, but do you have any information regarding leaving these powered on unloaded? is it something that is generally not allowed? I can not find any information about it online. So I was hoping you would have an answer! Thank you beforehand, best regards from sweden!

[@eddydutton4614]

In Mexico, power used to fluctuate a lot when I was a kid, so the TVs had a “regulator” transformer similar to that one, no moving parts, but regulated the power saving the TVs.

[@timtim8468]

I once had my hands on a little voltage regulator from the GDR, intended for lab use. It used tubes to generate a current, to get the transformer into more or less saturation. A little EL Audiotube to regulate mains voltage...

[@NatesRandomVideo]

General Electric used ferroresonant power supplies on their two way radio stations well into the 90s. Nearly impossible to kill. Great supplies.

[@erikziak1249]

17:31 I asked about the capacitors, thank you very much for measuring them. As expected, the measured value is slightly higher, but I am not sure if the 29,2 uf is because the capacitors are deterioated or this was within the acceptable tolerances when this device was built. Most likely it is due to a combination of both factors. Maybe if it were exactly 28 uf, you would get exactly the rated output voltages. I think that this device was meant to be connected to resistive loads, so the waveform was not an issue and was thus not compensated. I however do wonder how it might look if the input is far from a nice sine wave. Thank you for another video with nice explanations, this is really quality content on KZbin.

[@laharl2k]

Caps are usually +/- 20% so my guess is thats they are still within spec. Caps are one of the most inaccurate components on anything electric/electronic, in basic accuracy, thermal stability, and also in age stability. Add time and heat together and it gets even worse.

[@erikziak1249]

​@@laharl2k Correct. I have the same experience. I mostly use bipolar electrolytic capacitors. They are the perfect blend of price/size/performance for my needs. Sometimes I use foil capacitors.

[@rohitkhanna]

Brilliant analysis.... As always!!

[@user-yv5mt9rm3d]

Thanks for the video, I really appreciate all the work that went into this!

[@kolinevans9127]

The power dissipation interested me greatly thanks for sharing.

[@mernok2001]

Now I want to see a 10kW 3 phase version of this :)

[@tristanginod9238]

Yes but I don't think this version already exists because three phase is intended for two things : reducing losses and get a powerful rotating field for AC motors which are frequency-controlled.

[@albinklein7680]

I had one. I bought a TV-broadcast-van at an auction many years ago. It had a 25 kVA 3 phase isolation transformer in it. With three big capacitor-banks for the regulation "magic". I converted the van to a camper and nobody wanted the gigantic transformer. So sadly I had to scrap it. Shame on me; but I had no storage space for it and the price for copper was very high at that time. It regulated from 350 to 450 Volts input to 380/400V output iirc.

[@nicholasvalentine3907]

I don't think DGW has a forklift??

[@godfreypoon5148]

When a magnetic component is saturating, it runs at ALL the flux density. ;)

[@Skracken]

I'm from Sweden and that looks like the old color coding for wires we had here too, with red PE. Our house is from 1970 and it's a bit in-between standards. Black live, white neutral, yellow/green PE, and switched wires are either blue, green or yellow. The incoming PEN in the fuse panel is also red. And then with later additions there's brown live and light-blue neutral. You need to think twice so you don't mix things up!

[@jeldrikpetersen]

Red was DIN VDE colour for GND untill the late 70s. Either black = phase and grey = neutral.

[@johnconrad5487]

Thank you! Thank you! Thank you!

[@shrikanthkannan8234]

Is possible to connect ground the zero tapping of the transformer, because I measured 100V between stabilizer phase and ground

[@Shipu96]

Really fascinating and simple voltage regulator, it was very interesting to look at such a rare piece of tech. It reminds me of the so called "Royer converter" used as an inverter for inductive charging systems in a parallel reactive power compensation configuration. It is a self-resonating inverter, which uses the center tapped primary coil and the parallel compensation capacitors as a resonant tank and can self adjust its resonant frequency, which comes in handy and simplifies the circuit a lot. It has also a choke with a ferrite core before the center tap like in your schematic, that acts as a current source. Maybe you could make a video about it 😁😁

[@mernok2001]

14:45 Even if it was isolated,the ground wire woul still need to be connected.Grounded sockets must always have the ground connected.For isolated use ungrounded sockets would make sense.

[@okaro6595]

Not in Europe. The grounded sockets are specific ally designed to fit ungrounded sockets. If course if the sockets are grounded you must ground

[@piconano]

My brain is smoking. I'm gonna go lay down... What a magical mystical mastery of magnetics!

[@Zebra_Paw]

And what about the schematic of the 12V 50A power supply? Please I need it for a project!

[@TheTruthSentMe]

Why do you think the harmonics are prime multiples and not just the odd multiples?

[@mysock351C]

Maybe there's a deeper mechanism at work, but it could simply be due to the fact that the distortion in cases like this will pretty much always be symmetric, which means odd harmonics. This in turn means that the first few harmonics will automatically be prime numbers such as 3, 5, and 7. The odd man out is the 9th harmonic, but by then the energy is already getting quite small so its not as easily seen in the waveform as the first few or encountered as often. If you think about it, up until the 13th harmonic, the 9th harmonic is the only non-prime number. Edit: Better way to see the harmonics is with something like an FFT on a typical digital oscilloscope. Id suspect that even in the original form the circuit would have more than just 3rd harmonic distortion, but the 3rd harmonic seems to be the most prominent.

[@felixcat4346]

Thanks for posting.

[@mgsp5871]

I love that the capacitors are labelled with 25%ED but used with 100%

[@TonyBarr99]

The mains at my house is far from a perfect sine wave, so we can't be too critical of this thing.

[@laharl2k]

Same here. With all the electronics nowadays the mains is noisy AF. Far gone are the says of a clean and smooth sine wave, all thanks to the unfiltered chinese garbage products with 0.00001 power factor.

[@mernok2001]

@@laharl2k Power factor is not a big problem.I tried connecting almost pure capactitive and inductive loads,but the waveform wa snot distorted.The problem in nonlinear,non sinusoidal loads,like switching power supplies.

[@laharl2k]

@@mernok2001 "The problem in nonlinear,non sinusoidal loads,like switching power supplies." You mean like 99.999% of all electronic gadgets? Nowadays everything has a non linear switching psu. Aside from maybe your oven, everything is electronically controlled, your lights, your tv, even your fidge.

[@mernok2001]

@@laharl2k I still use chargers and power supplies with classic iron core transformers.

[@kolinevans9127]

How would you design the functional device to dissipate less than 5watts with no load? Seems like it could be possible no?

[@TheSalemSaberhagen]

Přiznám se, že dost dobře nechápu k čemu toto podivuhodné zařízení je dobré.

[@cocosloan3748]

Amazing explanation as always !

[@Mohammed-cc1jz]

Nice

[@mysock351C]

That was a lot more distortion than I expected. But it was probably made for simple loads like lamps or small table-top tube radios. Wouldn't want to plug anything reactive into it. I dont think single-phase electric motors would take too kindly to it, but its only for a few hundred VA, and there wasn't much to power back then.

[@johnconrad5487]

simple lamps dont need regulation

[@mysock351C]

@@johnconrad5487 They would if you don't like dim lights or light blinking from fluctuations on the mains.

[@peplegal8253]

I designed some few ferroresonant ac stabilizer with discrete cores. Usually isolated and low power (

[@mcflapper7591]

Are those caps still good enough at full load, temperature and voltage?

[@kolinevans9127]

Can I ask a question The linear dissipation was measured as the main inductor has 1.6 ohm however what about the rotational offset? if we know the inductance which you did not display {that i recall in this video} we could know the LC frequency between the 28uf and the full winding which could be guessed at ? 600mh 1 H ? Maybe much less It's interesting because at 363mh that is 50hz lc resonance on the full winding It would be interesting to know the inductance as this is not a DC device. Regards

[@alnov91]

Sorry but the creator only replies to IEEE MEMBERS in good standing.

[@kasel1979krettnach]

hey can you do such a circuit for higher frequencies (thereby less iron) and use it with bridge rectifier output to get stable DC ?

[@godfreypoon5148]

This would make a good rectifier input stage!

[@Bata.andrei]

Jerry Walker made a series of videos about ferroresonant transformers also. Very interesting technology from days gone...

[@oscarfiala2104]

Skvělé video

[@ren-416]

this simple look like tech, but actualy super complex callculation to make it work even better than current chinese tech implement,

[@smartups1]

One more video on this .+ Diy

[@taharmouz8864]

We was using it for TV

[@padariczada1014]

Wow! Da sieht man wieder : Alte Technik hat eine echt lange lebensdauer! Simpel aufgebaut und fast Unzerstörbar. Schade, dass heute Geräte vielfach nur noch knapp bis die Garantie Zeit abgelaufen ist überlebt, und sehr oft nicht mehr Reparierbar hergestellt wird.

[@Amperekaefer]

Ich bin voll Deiner Meinung! Denke es werden wieder Zeiten kommen wie früher, z.B. als dieser Trafo gebaut wurde, allein schon zugunsten des Umweltschutzes. *DiodeGoneWild* : Einzigartig, diese tiefgründigen Videos die er macht!

[@Amperekaefer]

@G E T R E K T 905 Ich hab in den 70ern mal ein dt. Autowerk besichtigt. Da wurde die Fahrertür mittels Automat auf- und zugemacht mit einem Zähler gekoppelt. Bei meiner Frage wie hoch der max. Zählerstand sei bei ausgeleierten Scharnieren, dünkte mich die Anzahl rel. recht bescheiden. Auf meine Frage ob man da nichts verbessern könnte, bekam ich die ketzerische Antwort: ...

[@Amperekaefer]

@G E T R E K T 905 Ich behaupte mal die Zeiten werden wieder kommen, wo *Nachhaltigkeit* das beste Verkaufsargument wird 😍.

[@DolezalPetr]

solidní převaděč spojitosti

[@stalkerfromvoronezh4493]

How about low pass filter?

[@wechselrichterschaltungen]

Was für ein Tastkopf ist das am Oszilloskop?

[@Amperekaefer]

Ich rate mal, Du meinst wegen der U-Festigkeit des Oszis? Die meisten modernen DSO haben eine U-Festigkeit am Eingang von 400V. Mit TK-Einstellung von 1:10 könnte man Spannungen von 4000V messen natürlich unter Berücksichtigung P-t-P?

[@wechselrichterschaltungen]

@@Amperekaefer ja..genau deshalb kam meine Frage...OK. nun bin ich schlauer... danke!

[@EsotericArctos]

I wonder if the capacitors may be not quite as good as they were new. Could the device maybe perform better with new caps? I guess the distorted sineqave really wasn't an issue for the purpose this device was designed for

[@TzOk]

I had the same problem - subtracting, not substracting ;) but otherwise great video as always.

[@DevCristi]

Wow, I just ended watching the first episode and the appeared this one, lmao :D

[@wuukaa9079]

Hey DiodeGoneWild, could you build the best voltage regulator you shown on this video? You could build the core yourself? :)

[@vaclavtrpisovsky]

You should put a voltmeter or an oscilloscope with a sine wave displayed (not to reveal the actual waveform) in the thumbnail to capture the essence of the video and to differentiate from the previous one.

[@1pcfred]

DGW is like the reincarnation of Nikola Tesla. Pretty soon he'll be counting the things floating in his soup.

[@Amperekaefer]

👍👏

[@godfreypoon5148]

Ferroresonant regulator PART 2 ELECTROMAGNETIC BOOGALOO

[@gacherumburu9958]

👍

[@twocvbloke]

Red for Ground, yeesh, I recall my science teacher in school recounting how people blew fuses re-wiring German audio equipment they imported because they wired it up with the Red wire as we had it at the time, to the Live in the plug (old UK colours were Red = Live, Black = Neutral, Green or Green & Yellow = Earth), scary stuff, no wonder there was the harmonisation thing, though soil-brown for live and water-blue for neutral makes no sense to me...

[@mernok2001]

I hope nobody was electrocuted from a ground connected to live.

[@makusmati]

Red and black are more commonly associated with DC power or loudspeaker + and -

[@twocvbloke]

@@makusmati Which is a thing dating back many years, and countries like the UK used Red and Black for several decades for Live and Neutral (I.E. Red for Danger, Black for charcoal, the colour after a good electrical burn)...

[@mernok2001]

@@makusmati It depends on country and age of wiring.In Hungary in the 1970s,black was live,blue was neutral and red was ground.

[@kjm-ch7jc]

I tried to understand but my mind couldn't work out what was being said.

[@EdHayes3]

Energy Star compliant I see!

[@cameron20020]

Do you think the capacitors have dried out and affected the output waveform?

[@fabriziobrutti1205]

I think they are film capacitors... They are dry and sealed and meaby molded in resin too... They are immortal 🙃

[@DiodeGoneWild]

they are not electrolytic. They seem to have the original capacitance and it's not likely that they have a high esr.

[@cameron20020]

@@DiodeGoneWild wow. Capacitors that last this long and still test fine. More alien technology!

[@kpkpkpkpkpkpkpkpkp]

can ferroresonant transformer protection against lightening surges

[@fabriziobrutti1205]

Maybe yes, but you would need pretty badass capacitors

[@fabriziobrutti1205]

"The ferroresonant transformers, which are also known as constant-voltage transformers (CVTs) or "ferros", are also good surge suppressors, as they provide high isolation" Directly from Wikipedia 🙃

[@Basement-Science]

@@fabriziobrutti1205 Maybe the isolated models will, but I cant imagine this one would do much against this. Fast transients like from a lightning strike can capacitively couple across the windings to the output. The resonant capacitor is not directly across the output, so it wont help with that here. Essentially, not all "surges" are created equal. This can probably protect well against some spikes induced by motors, but probably not for lightning.

[@fabriziobrutti1205]

@@Basement-Science ow... Right.. I didn't think to capacitive coupling across the windings... A MOV or a Transil at the output will help a lot... Anyway an isolated wersion with an electrostatic shield would be perfect 😃

[@objection_your_honor]

Is your cat Persian?

[@taurielv]

British

[@objection_your_honor]

@@taurielv What's the name of this breed?

[@Rainbow__cookie]

I have one of those but i want to make it isolated

[@Basement-Science]

Your best bet is probably a separate isolation transformer on the output. That is, unless you want to spend days rewiring and measuring and experimenting with it.

[@eDoc2020]

@@Basement-Science If possible, I would put the isolation transformer on the input instead of the output; that way the voltage drop of the isolation transformer is corrected.

[@mohamedbadusha4970]

Who is the fan of "bloody hell"😁😁😁

[@jooch_exe]

Interesting, an electromechanical version of a switch mode power supply. I´m pretty sure they would have made a proper SMPS if electronics were cheaper back then, but it seems copper was cheaper and the electromechanical solution more robust.

[@amkp40technology]

Power losse device.🙄😂

[@nevillewaiting485]

not prime numbers mate. odd numbers

[@thisone1697]

+

[@ruimvp]

Modified sine wave converter... Lol