JVC VCR (2000 era) power supply fails to start up

[robjordan]

Hi, may I declare from the outset I'm a hobbyist and this is my first time delving into an SMPS. It’s a learning exercise, but also a repair I want to achieve. Hope that's ok?

I'm trying to get an old JVC S-VHS VCR running. It dates from around 2000 and it sat unused for many years. When I tried to use it again, it won't power up. I have schematics, an isolation transformer and a scope. If I understand correctly, it’s a flyback design, with a MOSFET switcher and optocoupler feedback, but without an oscillator IC; the oscillation is started by discrete components.

A helpful AVForums post suggested that most failures after long downtime are due to a particular startup capacitor (C5104, 1uF), so I replaced that without too much investigation in hope it would fix the problem. The removed capacitor measured ok for capacitance, but the ESR – as measured by my cheap Chinese component tester – was 13 ohms, and the replacement measured 3 ohms. However the replacement hasn’t improved things, so I’m digging deeper to find the problem.

On the primary side, I scoped the MOSFET GS and I found that the MOSFET turns on for very short pulses, with a switching frequency initially around 10kHz, but then reducing in frequency until switching stops completely after about 200ms (see attached scope traces mosfet-gs-zoomed-out and zoomed-in).

On the secondary, I can see voltage climbs during this short startup period. An unregulated output which is meant to reach 48V climbs to about 36V and then decays (trace D). The main regulated output, which is meant to be 5V, climbs to about 1.4V then decays (trace I). I scoped the anode of the feedback optoisolator LED, and there’s some very small fluctuation in voltage (about 220mV p-p) but it’s at 50Hz and I think it’s just mains noise.

Do these symptoms point to any component in particular, or should I just replace all the caps and hope for the best?

Thanks for any advice.
Rob

[amyk]

This is a simple two-transistor self-oscillating design. C5104 would be my first suspect too. I'd also check for any cracked/dry solder joints, and then check the electrolytics on the secondary side.

I suggest disconnecting the PSU from its load and then giving it a light (few tens of mA) load on the AL5.8 rail, so that in the event the feedback goes open --- which might be possible if the original fault was a bad connection somewhere --- it won't damage the rest of the circuitry.

[robjordan]

Thanks for advice. I've ordered a full set of replacement electrolytics and we'll see how that goes.

For my education: any thoughts on why it would start to oscillate but grind to a halt after 200ms?

Rob

[inse]

Did you check those two resistors that charge C5104?

[robjordan]

I checked them in-circuit... they weren't a perfect match but not shorted so I guessed other paths were affecting the measured resistance; moreover the scoped voltage across C1054 rises to 45V, which is what I calculated it should be by dividing the potential according to the resistor spec (see the scope trace referenced points A and B).

[Update: Checked those two resistors in-circuit again and resistance slowly rises toward 220k, so I guess that's the big storage capacitor charging up?]

Is my reasoning sound?

[ResR]

[Update: Checked those two resistors in-circuit again and resistance slowly rises toward 220k, so I guess that's the big storage capacitor charging up?]

Is my reasoning sound?

A capacitor in circuit can mess up the resistance reading, specially if it still has some charge. I would also bet that electrolytic capacitors are at fault, my old Samsung Syncmaster 204b didn't turn on also due to the output capacitors being dried out (too high esr) and replacing them resolved the problem.

[robjordan]

I replaced all the secondary electrolytic capacitors but unfortunately it didn't fix anything and the behaviour is exactly the same. Oscillation in the primary starts up but then slows down and stops. I'm trying think of reasons why that might happen.

If I'm understanding correctly, there should be negative feedback from the secondary to the primary, but isn't that only supposed to happen when the regulated output reaches or exceeds the intended voltage? In any case I have monitored the input and output of the optocoupler and I only see millivolts, so I don't think it's feedback which is making the primary shutdown.

Rob

[fzabkar]

Try this fault database:

http://www.e-repair.co.uk/faults/VIDFAULT.HTM

[sean0118]

Is there a load connected?

edit: I guess it probably would not effect the oscillator, so might not be relevant in this case. But I fixed a 1980's SMPS once where the original issue was the secondary caps but it still wouldn't start up because I had disconnected the load to protect it, once the load was reconnected it worked fine.

[robjordan]

Is there a load connected?

I put a couple of LED/resistor pairs across the 5V output to draw 40mA as suggested by amyk above. But the 5V line never gets above about 1.5V.

[robjordan]

Try this fault database:

http://www.e-repair.co.uk/faults/VIDFAULT.HTM

It's a good resource, thanks.