I'm totally puzzled by this power supply!

[Wizard]

Re: I'm totally puzzled by this power supply!

All caps must be in place. There is no error of room to operate, replace the electrolytics and check other caps.

This is over 50-150KHz design, this is reason for need of ultrafast diodes/transistor that can switch fast. Not simple as old school SMPS. This is reason the windings are litz wires not simple copper windings to cut down on skin effect.

Litz wires looks like rope with high strand count.

Cheers, Wizard

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Thanks for the info, I'll put the caps back on before I test with the new diode. The original 220uf caps measure about 0.5 ohms ESR each. Not great, but should be usable. I would hate to spend $30 on capacitors until I know the supply will work.

Will using a 200-250W light bulb in series with the power supply for safety affect operation? I'm just trying to prevent the MOSFETs from blowing if there are other problems. Maybe I should just use a fuse.

All caps must be in place. There is no error of room to operate, replace the electrolytics and check other caps.

This is over 50-150KHz design, this is reason for need of ultrafast diodes/transistor that can switch fast. Not simple as old school SMPS. This is reason the windings are litz wires not simple copper windings to cut down on skin effect.

Litz wires looks like rope with high strand count.

Cheers, Wizard

[Wizard]

Re: I'm totally puzzled by this power supply!

start low watts (around 30W)

Cheers, Wizard

[Toasty]

Re: I'm totally puzzled by this power supply!

Did you manage to ring the coil as kc8adu recommended?

See this: http://www.repairfaq.org/sam/flytest.htm

Is the ESR you reported on those caps correct @ 0.5 ? That sounds high to me, but I don't any 450v caps to compare it against. Can you compare that to any other similar rated caps you have? I was expecting .05.

Toast

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

The caps seemed decent for ESR when compared to similar rated new ones.

I'll be trying out the new (proper) rectifier soon, and if it still doesn't work, I'll test the coil.

Hopefully the rectifiers will come today, if not tomorrow. I want to make some headway on this project. These three 42" plasmas are taking up a lot of space in my workshop.

I took some readings on the 4th plasma that was working, and the boost converter was running at 392 volts, so now I know how it should be operating at least.

Did you manage to ring the coil as kc8adu recommended?

See this: http://www.repairfaq.org/sam/flytest.htm

Is the ESR you reported on those caps correct @ 0.5 ? That sounds high to me, but I don't any 450v caps to compare it against. Can you compare that to any other similar rated caps you have? I was expecting .05.

Toast

[Toasty]

Re: I'm totally puzzled by this power supply!

Do you have a calibration output on your scope?

If the coil is shorted, you could take out the new diode(s) real quick. You did order spares, correct?

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Do you have a calibration output on your scope?

If the coil is shorted, you could take out the new diode(s) real quick. You did order spares, correct?

I think I do have a calibration output. I also have a signal generator.

And yes, I ordered 10 diodes when I only needed 2 :-)

[Toasty]

Re: I'm totally puzzled by this power supply!

Assuming 0.5vp-p square wave calibration signal:

Calibration output, scope lead, .05uf disc cap across coil.

Scope to 50mv, .1ms sweep. If no ring, try isolating cal. output with a .002uf disc cap in line.

You should get a nice sine wave "pyramid on its side" type of trace.

See 2nd scope picture, top trace in this link:
http://www.google.com/url?sa=t&sourc...0Pictures.html

If it doesn't do that and looks like it only does one or 2 swings p-p, then its like shorted.

Toast

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Cool, I'll try that!

If my scope cal output is a different voltage, would it still work the same? I could also use my signal generator if need be.

Assuming 0.5vp-p square wave calibration signal:

Calibration output, scope lead, .05uf disc cap across coil.

Scope to 50mv, .1ms sweep. If no ring, try isolating cal. output with a .002uf disc cap in line.

You should get a nice sine wave "pyramid on its side" type of trace.

See 2nd scope picture, top trace in this link:
http://www.google.com/url?sa=t&sourc...0Pictures.html

If it doesn't do that and looks like it only does one or 2 swings p-p, then its like shorted.

Toast

[Toasty]

Ring that coil!

If your p-p voltage is greater, then set the scope voltage scale higher. Voltage isn't the key in general, the waveform is. You can ring the coil with a very small waveform. A signal generator will work fine. Start low and bring the output level up until it rings on the scope.

Note that it needs to be out-of-circuit to test correctly. You don't always -have- to remove them from the device, but they do need to be electrically disconnected from the rest of the circuit. Sometimes there are trace jumpers nearby that you can lift, rather than unsoldering the the coil leg.

I attached a few photos of scope results I got from a similar coil from a computer PSU.These are using the 1KHz 0.5V p-p scope cal. output as input with a substitution box and .01uF selected.
1) .0022uF cap on input
2) .00056uF (560pF) cap on input
3) quick -n- dirty bench setup - Nothing fancy required.

Toast

[gonzo0815]

Re: I'm totally puzzled by this power supply!

You need the fast diodes due to the switching frequencies of this boost converter. Those PFC converters are switching in the range of 50khz +
A normal diode will short out the current, as it can`t switch off fast enough. A normal diode should run very hot and should die.
Furthermore, i would use a 20 to 60w bulb as ballast, dunno of such a circuit can run without it. May be the PFC controller is shot too.
But usually, a PFC stage is not that complicated. Basically it is the controller, transistor, energy storage coil and the diode. I would suggest to id the PFC controller and then looking into the data sheet.

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

The controller is a common UC3854N. I do have the data sheet for it, but I still would like to find the schematic for this power supply. I think they are more complicated than they need to be.

I think these power supplies use a lot of current. The main AC fuse is 8A slow blow, and all the switching MOSFETs are rated at 24 amps each.

The replacement diodes came today, so I'll be trying them out after work tonight.

You need the fast diodes due to the switching frequencies of this boost converter. Those PFC converters are switching in the range of 50khz +
A normal diode will short out the current, as it can`t switch off fast enough. A normal diode should run very hot and should die.
Furthermore, i would use a 20 to 60w bulb as ballast, dunno of such a circuit can run without it. May be the PFC controller is shot too.
But usually, a PFC stage is not that complicated. Basically it is the controller, transistor, energy storage coil and the diode. I would suggest to id the PFC controller and then looking into the data sheet.

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Good news, I got two more of them working

Some things I found out- The power supply will not run without the TV connected, and it will not run with any ballast. Even a 250W light bulb caused it to shut down.

I would have had all 3 running, but the 3rd one needs some more work. I had it turned on for a minute or so, and heard a ticking noise while I was right behind it. I got a bad feeling and jumped out of the way just before the diode exploded in a loud shower of sparks. Upon inspection I found that the MOSFETs for the +Vs (180v) supply had shorted. I replaced everything again, but I won't be testing it until tomorrow. It had a great picture before it blew out though

Too bad these sets are only 480P...

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Not bad for a 7 year old clunker that was killed in a bar... It's really hard to get a picture of these for some reason.

Now that I know my way around these power supplies a bit, all 5 of these plasma monitors are working now.

Thanks again for the help guys!

[Toasty]

Re: I'm totally puzzled by this power supply!

Great! Good Job!

What was the end conclusion as to the failure(s).

What were the parts & numbers?
Parts source?

It's good info if someone else has this problem.

Glad we could help.

Toast

[Maxxarcade]

Re: I'm totally puzzled by this power supply!

Great! Good Job!

What was the end conclusion as to the failure(s).

What were the parts & numbers?
Parts source?

It's good info if someone else has this problem.

Glad we could help.

Toast

Original cause of failure was probably a combination of bad caps and overheating. Or did you mean the trouble I was having with the repairs? The repeated failures during repair were mostly because of the improper rectifier I was using.

I can get the part numbers tomorrow some time, but I did mention the rectifier in an earlier post I think. I got all parts from Mouser.

The only bad thing is, I don't want to spend the money replacing all the large caps. The cost would probably exceed the value of the whole thing. Hopefully they will hold out for a while.