Battery fluorescent inverter + Royer oscillator explained

[@JOHNCHENSJCA]

It is interesting seeing a practical Royer oscillator in action.

[@kasel1979krettnach]

i so wish I had sent you my old slide projector. 8 separate PCBs in it, 5 different voltags, SCRs, Triacs, Mosfets, EPROMS, RS232 interface, microcontroller in a socket, light guides, photoresistors, plunger coils, little DC motors.... it was a blessing.

[@d.k.9406]

Danke! Thanks! Hell Yeah.. i love this Baxandall oscilator Very neat .. And seems quite common until nowadays? The Royer and Baxandall could get its own BASICS video like you started with THE 555 ❤

[@DiodeGoneWild]

Well, yes, maybe I might make a separate video about this one. Baxandall, which everyone calls Royer :D. And thanks for your support ;).

[@tajtrlik1111]

Opäť ti chcem poďakovať za toto veľmi pekné video. V teórii meničov mám obrovské medzery, takže vysvetlenie princípu činnosti spolu so schémou prišlo vhod, nie že by som sa tým v blízkej budúcnosti chcel zaoberať v praxi, ale rád som sa nechal poučiť.

[@Alexelectricalengineering]

Nice as always 😊😊

[@KryH85]

Niiiiiiice

[@chrisgj198]

Often you can fix a shorted cell by sending a large pulse of current through it. One method is to charge a large electrolytic capacitor and then discharge that through the cell. Once the cell is no longer shorted, you can charge it as usual, though it may not be reliable. The worst ones I had were some cells that were shorted with a somewhat resistive short that was a low enough resistance that it would self-discharge the cell in a day or so, but not a low enough resistance to be able to get much of the current pulse to go through it, so I couldn't blow away the short.

[@LMB222]

Not always - I've tried it. And the wires between the cell and the case simply burned out.

[@chrisgj198]

@@LMB222I guess if you use a big enough capacitor, destroynig the wires is possible! Certainly this method of removing shorts does not always work, but if the battery was otherwise to be scrapped, and is intended for a non-critical purpose, then it can be worth a try.

[@xenoxaos1]

They also will vent the electrolyte...

[@eDoc2020]

When I've done it I touched the shorted 1.5v cell across a 12v lead acid battery. It worked okay for a bit but continued to get worse.

[@chrisgj198]

@@xenoxaos1Never when I did it. I don't use too much energy though.

[@victorman2227]

Thanks for the detailed rundown. This sort of inverter with a resonant capacitor and supply inductor is properly called a Baxandall oscillator but everyone calls it resonant royer (even Jim Williams called it that). The Royer is the saturation commutated one. Its great how this circuit is so simple and can be very efficient with BJTs. Even some old soviet ones with Uce sat of 0.5v got me to mid 80% efficiency. Also what about the base-emitter voltage limit of the transistors(oh nevermind its not powered from the 8.2 supply but from the 4.8V battery, so it's fine), when i ran this kind of inverter from 20V supply my transistors gradually lost gain and i had to place a reverse diode for protection there.

[@DiodeGoneWild]

Well, I did a bit of a research, and you're right, it should be called a Baxandall, but everyone calls it Royer for some reason. I put a note into the description.

[@555-xd1fo]

I like your bench power supply its antique

[@KomiyanVT]

Those inverters typically put out AC voltages, so they are usually pretty good with fluorescent tubes - so long as you preheat the filaments! But here they rectified this output, so that tube will be hating it! Great stuff for LEDs though, you can essentially replace that tube with a series string.

[@schaltnetzteil495]

Thanks for uploading! I've also had rotten battery cells, but they were NiMh cells. They were AAA cells and had like 50mA short circuit current.

[@mernokimuvek]

I like high voltage high frequency power supplies because they can supply all discharge lamps: fluorescent, low pressure sodium, high pressure sodium, high pressure mercury vapor, metal halide, noble gas signs, plasma globes. It should be a symmetrical sine wave likea CCFL inverter.

[@mrlux0716]

Daniel, if you could show in a future video, I would like to know what the waveform is when you have a worn tube rectifying in a magnetic ballast, where the ballast has a very loud buzzing and the tube visibly flickers. Probably, even if you don't have a worn tube, maybe you can find one in a lamp bin

[@dogwalker666]

As i have replaced all the fittings at work with LED, I have about 100 of these modules, I am slowly de soldering them.

[@VC-oz5fi]

Nice royer oscillator Btw, do royer oscillator work with current limited supply? Because ZVS will refuse to start if the supply is current limited/soft start And what the purpose of C1 and C2 on the schematic?

[@tiagoferreira086]

I don't quite understood the role of the big inductor on the positive line. Is it for limit the peak current on the transistor? suppress interference? resonate with the capacitor next to the transformer?

[@DiodeGoneWild]

No. The inverter produces a sine wave, not a square wave, so it can't be connected directly to DC. You have to feed it via an inductor to allow for the sine wave halfcycles on the center tap. There would be no inductor if it was a square wave inverter.

[@tiagoferreira086]

@oneWild I see, i also had that suspicion in mind... Thanks for your clarification 👍😉 Edit: i made a bit of research and indeed the inductor combined with the capacitor makes a sine wave, old school yet simple and clever, according to wikipedia: "The Baxandall converter has been used recently in driving fluorescent tubes from low-voltage sources, often using rechargeable batteries, for emergency lighting and camping etc. Also in his 1959 paper, Baxandall described a voltage-switched variant of the sinewave oscillator. This variant seems to have been the forerunner of most two-transistor drivers for compact fluorescent lamps (CFL's) which has been extended recently to drive low-voltage LED lamps. "

[@thanhhuynh272]

I am seeking to build a Royer oscillator to use as a replacement EHT power supply in an oscilloscope that originally sourced its tube EHT off a 950v secondary on its 50Hz poer transformer…but that winding, many thousands of turns if hair-fine wire…shorted. It was the winding right on top so I pulled half of it off and left about 430v worth of it and reconected it as a voltage doubler…and it worked for nine months, but then it shorted again. Iv’e had the pip with it and will tear off the remaining 430v of wire and replace it with a Higher current 20v winding to provide the supply for a Royer high voltage. The BWD 845 Australian oscilloscope sources ALL its DC rails off a Royer Oscillator…it is a deceptively simple circuit, just an electrilytic decoupling cap, two fairly matched transistors, (MJE3055’s) and the trandformer. There are another two resistors, an electrolytic and a diode that make up the bias circuit. This circuit (in the BWD 845) has no inductor in series with the DC rail as here and the transformer has no air gap in the core. The Tektronix 7000 series mainframe oscilloscopes also use a similar circuit, again, deceptively simple, just two SLOW 2N3055-like transistors, a bias network and a bypass cap…here too, no series inductor. In the Tek scopes it only provides the EHT and 6.3v heater voltage for the tube heater…but in the BWD 845, it does everything! I wish to duplicate either of thise circuits and I know enough to know the “magic” is in the transformer, its lack of air gap, (so a saturating core) and I have heard that only a few types of ferrite will work, you just can’t get any old flyback SMPS transformer and grind down the core to close the air gap. In the BWD 845, this Royer oscillator had “failed”…actually, it still sort of worked, it oscillated but the secindary voltages were all really low and the two original transistors, TIP3055’s got really hot. I assumed that their gain (hfe) had somehow gradually decreased over the last few decades to the point they were unable to get out of yheir linear region…do I replaced them with the two matched MJE 3055’s…all I could get these days…and that fixed it. So the transistors seem to need enough gain to saturate. Unloke your circuit here, these produce square waves, lack that series coil and lack any resonator capacitors on any secondaries or between the collectors of the transistors. You did mention thst your circuit here is actually a Baxabdall, if so, are mine true Royers? I wish to sttempt to build one from scratch, so can you direct me to the type of ferrite i need for the transformer core? My only real other choice…because of my lack of understanding, is to purchase a Tek 7000 EHT transformer (120-0879-00) online snd build the Royer circuit sround that because I know it has the “correct” type of ferrite and lacks an air gap. So I am basically seeking to know how somebody designs such a circuit from scratch and I know they start with the transformer core…but how to choose it and how to move forward from there?

[@victorman2227]

Yep yours is a true Royer. Then there is the current fed Royer that has a series coil. And the Baxandall or current fed resonant Royer has the coil and resonant capacitor. Check out the schemes that are used for CCFL monitor backlight supplies. Might even be able to reuse the transformer, the CCFL tube voltage is about 1kV i think. The scientific correct name for a Baxandall is a "current-fed push-pull parallel-resonant inverter". Maybe some articles would have a section on transformer design. For the Baxandall transformer design isn't that hard, i have done a couple just by guesswork using subpar soviet 2000 relative permeability ferrite, make sure the transformer doesn't saturate.

[@victorman2227]

The resonant royers(Baxandall) work elegantly with 5V supply and low Vcesat BJTs, as they dont need a base resistor the efficiency is nice (you will need a base resistor and diode with higher voltage input as it exceeds the Veb limit). See Linear Tech AN65(to a lesser extent AN118) by Jim Williams. There are also MOSFET versions, but they have loss in the gate charging resistors if there is no driver IC. Also the Baxandall inverter is kind of ZVS so it has low switching loss.

[@xDR1TeK]

Man this channel isn't weird at all. It's nice to find a channel with similar interests as mine. Quick question, what happens if we remove the inductor that is connected to the center tap? Just curious. I think it is meant as an RFC.

[@snipersquad100]

"Why doesn't your cat believe you?"

[@ParswaNathTheo]

I've question: why is voltage doubler bad for driving fluorescent lamps?

[@DiodeGoneWild]

Because it produces a DC current.

[@teardowndan5364]

When a CFL is driven with AC, the plasma stays mostly still since it gets accelerated in alternating directions all of the time, even more so with an electronic ballast which operates at 20+kHz and gives ions even less time to accelerate in either direction which improves lamp life. When you apply DC, positively charged atoms in the plasma always accelerate in the same direction, slam into the negative electrode at higher speeds than normal, cause it to sputter and fail prematurely. That is why cold-cathode tubes have a much thicker cathode to accommodate the wear.

[@d.k.9406]

​@@teardowndan5364 Thx man!

[@tigerelectronics5966]

I built a even simpler 12V inverter for flouro tubes using just one power transistor, and some capacitors and resistors! It worked really well, but it completely ruined the tubes very quickly. Flouros indeed don't like running in cold cathode mode at all

[@mernokimuvek]

The problem with hat was DC operation. Flyback power supplies output an asymmetrical voltage and the tube will act as a rectifier.

[@lmwlmw4468]

Very nice.

[@carlosaugustorj]

cool

[@techtinkerin]

I really miss nicads they used to work brilliantly in RC cars and drills😊

[@simontay4851]

No, NiCds are crap. They leak, cadmium is toxic, they have low capacity and go rotton. NiMh and Lithium are much better.

[@d.k.9406]

Great Video... Lots of infos over different toppics. Flourouscent lamp Batterys Oscillators ... And this doubtful cat.. But luckely there was oscilloscope waveforms to convince it😂

[@ralfhartmann5050]

Hello, in which land do you come from?

[@twocvbloke]

Certainly a lot more complex than the cheap 12v drivers I've pulled from a few 8w T5 lights (they were faulty in one way or another), those are usually based on a TIP29B transistor and not much else, a trio of capacitors, a lone diode, a resistor and a transformer wound around a ferrite rod, the one I'm looking at right now flickered badly, but the fitting it came from now sports two 240v Philips Matchbox electronic drivers and two 840 tubes having started off with just a single murky yellow tube... :)

[@gordonwelcher9598]

How to get an an oscillator named after you?

[@abderrahimziky6538]

@DiodeGoneWild , I think this oscillator is similar to one you used in metal ditector project.

[@DiodeGoneWild]

Yes, it is that same oscillator design :)

[@anoimo9013]

what is the zener for?

[@DiodeGoneWild]

I guess it's to protect the transistors from the inductive spike from the inductor when the circuit shuts down. When the PNP transistor turns off, making both the NPN transistors turn off, the circuit suddenly stops drawing current from the inductor, but the inductor tries to keep the current flowing. This creates a high voltage spike, which would damage the transistors, so the Zener has to clamp it.

[@andymouse]

Nice !...cheers.

[@mernokimuvek]

I was hoping it would be like a CCFL inverter and drive the tube with AC, not a junk multiplier. They could have simply used a transformer with a higher secondary voltage.

[@DiodeGoneWild]

Yes, so did I. But putting 2x the number of turns on the transformer secondary was probably too expesive... or maybe this produces less interference. But in a long run, it ruins the tube.

[@WagTsX]

So basically the inverter is running all the time but only disconnects the lamp when not in use? Weird.

[@DiodeGoneWild]

This inverter is connected to mains all the time to keep the battery charged and also to keep the relay coil energized, which keeps the tube connected to the mains inverter.

[@LawpickingLocksmith]

Nice cat!

[@G1ZQCArtwork]

I think the fact the tube is poorly driven is only as a back up for power failure illumination. It is not likely to be run this way for long. Better to upset the tube than have injuries trying to exit a building in the dark.!

[@dogwalker666]

3 hrs as legal requirements.

[@DiodeGoneWild]

Drawing 1.5A from a 1.7Ah battery, this would work for a little over 1h.

[@michaelfisher9671]

The fixture may have been replaced by crappy Chinese LED’s but the original is marked made in PRC.

[@Mladjasmilic]

Nadam se da koristiš izolacioni transformator.

[@thisisdvd8094]

The battery looks like a pipe bomb

[@migry]

The scope screen and even the scope keys use that horrible standard font, so often used in Chinese products.