[SOLVED] Power supply repair on Sony SVO-9620

[decoder_82]

Hi,

I am trying to repair a power supply unit from a videorecorder Sony SVO-9620.
There is a service manual for the device but unfortunately no schematics for the PSU block.

There is only reference about the voltages expected at the connector:

pin 1 -7/-6v
pin 2 12v
pin 3 GND
pin 4 5.3v
pin 5 5.7v
pin 6 GND
pin 7 GND
pin 8 GND
pin 9  12.3v
pin 10 12.3v
pin 11 7v

Initial symptoms: no output from the power supply.
What I did: replaced power resistor on the primary side that failed open, replaced IC101 (MA2830) than failed shorted. Replaced all electrolytic capacitors (some were leaky).

Now I get correct voltages output when there is no load. As soon as I plug it to the VCR, it makes noise like Tweet-tweet-tweet or chirp-chirp-chirp which probably indicates overload. I know there is no other issues with the VCR it works fine if I use a spare power supply.

I preventively replaced the PC111 but it didn't change anything.

Any advice on what I could do next to further troubleshoot ?

[eurgenca]

Is ic 201 TL431? If yes replace it. Also check sec. caps.

[decoder_82]

Thanks for the reply.
All caps have been already replaced, including on secondary side.
Not sure what is IC201. Here is a picture of the IC. How can I check it?

[eurgenca]

It is TL431 replace it and test ps. It is hard to test this ic. You have test circuits in datasheet but it is better to replace this ic.

[decoder_82]

Thanks for the reply. Will find a replacement IC and see if this solves the issue.

[decoder_82]

Alright so as suggested I did replace the TL431 IC. Unfortunately the power supply still has the same issue. Any further advice on what else I could check ?

[eurgenca]

Can you post pic from other side.

[decoder_82]

Here is a picture of the other side.

[floobydust]

I would suspect the VCR has a short causing the PSU overload. See which rail collapses when it is in hiccup mode. It could be a motor H-bridge IC that has failed.
The PSU should be OK now but not liking the overload, don't run it too long like that IC101 under stress.

[decoder_82]

I would suspect the VCR has a short causing the PSU overload. See which rail collapses when it is in hiccup mode. It could be a motor H-bridge IC that has failed.
The PSU should be OK now but not liking the overload, don't run it too long like that IC101 under stress.

That's also something I considered but since it runs fine when I put another power supply block we can rule out this cause.

[eurgenca]

What is voltage at pin5 (ic101) with and without load?

[jwet]

The MA2830 is an odd duck.  It just contains a couple of HV transistors and some diodes.  The circuits are self oscillating- hysteretic like.  Considering your symptoms, I would expect that a small passive, (probably a cap) is out of tolerance and not allowing it to operate at full power (at the proper frequency).  It operates no load but can't handle a load.  You might want to A/B test on a known good unit and yours since it seems you have on available.  I've attached a couple of docs that I googled.  You can likely find better on the internet.  Shindengen is a power supply module company and makes some of the lowest cost consumer power supplies in existence.  They have a bunch of parts like the MA2830 that are very lean on details in the data sheets- I think they're mainly designed for in house use in their own supplies.  I found a lot of Russian language circuits that use them.

See attachments- good luck.  Might help.  The first is a generic flyback using the MA2830 and the second shows the internals of MA2830 in context.  You can likely find better images on line- they get butchered on save...

[decoder_82]

What is voltage at pin5 (ic101) with and without load?

Many thanks for the suggestion.
I will check pin 5 of the MA2830. I just have to figure out the right dummy load as it will be difficult to measure with the power supply in the VCR

[decoder_82]

See attachments- good luck.  Might help.  The first is a generic flyback using the MA2830 and the second shows the internals of MA2830 in context.  You can likely find better images on line- they get butchered on save...

Thanks that's very useful. Indeed not so much details documented for the MA2830.

[floobydust]

Sorry I missed that you have a second PSU that works.
I would guess next the electrolytic capacitors are old and could be low value, especially being so close to the heatsink. The secondary-side ones are important.
The board appears to regulate the +12.3V rail as the master rail and other rails can move around a bit in voltage.
+12.3V rail C201, C205.  C204 does something around the TL431. These caps are important for stability.
You can run the PSU on the bench using small loads, light bulb etc. as well.

[decoder_82]

All the electrolytic capacitors have been replaced by new ones.

At the pin 5 of Ic101, I measure 0v when there is no load. When there is overload, I observe some pulses of negative voltage (mV).

[eurgenca]

Yes. According to dataset -0.13V is limited to stop ps from working. If the value is higher than go in this direction to the sec. side of ps.

[decoder_82]

Yes I will look again on the secondary side. All electrolytic Caps are new. I checked resistors and transistors and did not find bad ones.

[decoder_82]

Yes. According to dataset -0.13V is limited to stop ps from working. If the value is higher than go in this direction to the sec. side of ps.

Do you have other suggestions of things I can check on the secondary side?

[floobydust]

Did the changed caps include the two small ones on the primary side, by IC101? These are also critical. Look like C111, C113. If those are new, I would keep looking on the primary side.
The shorted MA2830 could also damage any parts connected to it. This circuit has extra 2 small transistors Q101, Q102 I would see if they are OK and not damaged/weak.
On the secondary side I can only see an open rectifier half on D201, D202 upsetting things.

Another approach is to put the PSU's side by side and take comparison ohms/diode test readings.

[decoder_82]

Thanks yeah I did check all the diodes and transistors (in circuit with diode tester). Everything looks ok.
Small caps were replaced as well.
I did a side by side comparison of the two PSUs with ohmeter. Couldn't find any difference.
I will try to take some voltage measurements of the two PSU units live with some load on it.

[floobydust]

If you have replaced and checked so many parts, either something was missed or tested wrong. What size is the fuse on it? Have you measured/compared operating frequency.

Rarely, I have seen an home AV receivers the SMPS IC shorts and blows and it damages the transformer primary from the overcurrent. You get shorted turns, and I use a Ring Tester to flush that out. I hope it's not the problem here. Something to keep in mind if this repair keeps being so difficult.

[jwet]

A side by side load test will likely be fruitful.  You don't have to load all the outputs- just one- the one that has the TL431 and opto feedback.  This gets the circuit working at a reasonable power level.  An incandescent lamp from a car is  a nice load usually and comes in many sizes.  Be careful with big resistors as loads unless you support them in a little vise, etc- they get hot and will quietly burn a hole in your bench or start a fire.

Draw a little schematic of how the 2830 is connected, just trace out its pins.  Pin 3 is usually key and is often how the oscillator works.  If there is a cap on that pin, replace it.  It needs a to be a high voltage unit.

[decoder_82]

If you have replaced and checked so many parts, either something was missed or tested wrong. What size is the fuse on it? Have you measured/compared operating frequency.

Rarely, I have seen an home AV receivers the SMPS IC shorts and blows and it damages the transformer primary from the overcurrent. You get shorted turns, and I use a Ring Tester to flush that out. I hope it's not the problem here. Something to keep in mind if this repair keeps being so difficult.

Thanks - indeed there is always a chance something was not measured properly. I will double check there are not some many components.
I don't have a ring tester but indeed the transformer primary failure should be kept in mind.

[decoder_82]

A side by side load test will likely be fruitful.  You don't have to load all the outputs- just one- the one that has the TL431 and opto feedback.  This gets the circuit working at a reasonable power level.  An incandescent lamp from a car is  a nice load usually and comes in many sizes.  Be careful with big resistors as loads unless you support them in a little vise, etc- they get hot and will quietly burn a hole in your bench or start a fire.

Draw a little schematic of how the 2830 is connected, just trace out its pins.  Pin 3 is usually key and is often how the oscillator works.  If there is a cap on that pin, replace it.  It needs a to be a high voltage unit.

My hope to that a side by side load test will help to understand. Could be some component that fails under load or temperature.

Good idea the diagram. It will for sure help.