Magnavox DRV power supply SMPS

[budm]

Re: Magnavox DRV power supply SMPS

I used this method with the vom and it passed.
http://electronicsbeliever.com/how-t...-is-defective/
The 520 failed it and i trust it now over the vom. I used new fets as a comarison test. Now on to figure out whats destroying the fet.

No way it could have pass the Diode mode test to check S and D if you FOLLOW that instruction.

From your link:
A good MOSFET should have a reading of 0.4V to 0.9V (depends on the MOSFET type). If the reading is zero, the MOSFET is defective. When the reading is “open” or no reading, the MOSFET is also defective.

When you reverse the DMM probe connections, the reading should be “open” or no reading for a good MOSFET. If the reading is zero, the MOSFET is defective.



Step #2 on How to Know if MOSFET is Defective

:Resistance Check

The next method to use on how to know if MOSFET is defective is to conduct a resistance check. A good MOSFET should have high resistance between drain-source regardless of the DMM probe polarity.

[johnfin]

Re: Magnavox DRV power supply SMPS

I only used the 520 on the mosfet. I used an esr, analog meter, and sprague teleohm on caps. DVOM diode test on diodes. I dont want to lift the legs of the tiny diodes on the bottom of the board to test them. Now that the mosfet is out the resistance is high. Whats the best way to chase the source of the problem. This is the second MOSFET so something it taking it out.

[momaka]

Re: Magnavox DRV power supply SMPS

p.s. I actually pulled c2109 last night and checked it. What else could take out the mosfet?

I can think of three things only:

1) The MOSFET was shorted before you did the incandescent bulb test. With the incandescent bulb, there's very little chance the MOSFET can short and this would be only if 2) applies below.

2) Open resistor between MOSFET Gate and primary-side ground. If you look at the hand-drawn diagram I posted, that would be 82-kOhm resistor connected to the MOSFET Gate.

That said, on the schematic of your power supply that Budm posted, there isn't such a resistor. Instead, there is a Zener diode (D2107) connected to primary ground. Its function is to clamp down the high voltage from the 4 biasing resistors, R2103, R2104, R2105, and R2106. If this Zener diode is gone open-circuit (very unlikely), then the voltage over the MOSFET's Gate could go too high and short-out the MOSFET. BUT! Diodes and Zener diodes typically go short-circuit, so this is unlikely.

Worth mentioning here is that there is a mistake in the schematic - it shows that Zener diode D2107 has its Cathode connected to primary ground and Anode connected to the MOSFET's Gate. The diode should be connected the other way around in the schematic. The PCB/board view shows that the Zener diode is connected properly, though. I'm just mentioning this in case someone sees this schematic and gets confused.

3) Open resistor or shorted diode in the snubber network. This snubber network typically consists of a diode in series with a capacitor-resistor parallel network. On my hand-drawn diagram, that would be the 10 nF 1 kV cap, 22-kOhm resistor, and diode - all in the upper left-hand corner of the schematic.

On the schematic for your power supply, I don't even see a snubber network. That is a really shitty design! If you do manage to fix this PSU, I suggest you copy the snubber network from my hand-drawn schematic and put it across the transformer's primary-side power pins.

Now that the mosfet is out the resistance is high. Whats the best way to chase the source of the problem. This is the second MOSFET so something it taking it out.

Put a new MOSFET in. It should not blow out with the incandescent light bulb if the above items I mentioned were checked.

I only used the 520 on the mosfet. I used an esr, analog meter, and sprague teleohm on caps. DVOM diode test on diodes. I dont want to lift the legs of the tiny diodes on the bottom of the board to test them.

Forget about all of that equipment for the moment. A regular digital multimeter with resistance and diode test functions can do over 90% of the troubleshooting work.

For MOSFETs, I prefer to use the resistance setting and check for short-circuit between Source-Drain or Gate to any of the other two pins. If that passes, I also check the body diode in the FET with a diode test, just in case. These tests can often be done in circuit.

As for bipolar transistors, diodes, and Zener diodes, I use the diode test function to verify that the junctions have proper voltage drops (typically 300-800 mV, depending on the diode type). This test can often be done in circuit.

[johnfin]

Re: Magnavox DRV power supply SMPS

Good catch man, you are right about that zener. I checked it and its ok. My snubber section (upper left) is different then the one I have. The schematic is for a 515, I have a 557. I checked those components, caps and diodes are large in size and easy to check. What I have been trouble checking is the small diodes on the bottom left of the schematic. D2108, D2109, D211, D2113

[momaka]

Re: Magnavox DRV power supply SMPS

Good catch man, you are right about that zener. I checked it and its ok. My snubber section (upper left) is different then the one I have. The schematic is for a 515, I have a 557.

Ah okay, I see.
I pity whoever has to work with the 515 PSU, lol.

What I have been trouble checking is the small diodes on the bottom left of the schematic. D2108, D2109, D211, D2113

Well, I can tell you straight off-the-bat that D2108 will appear as a short-circuit due to being in parallel with resistor R2108. That said, this looks like another crappy design flaw, because if R2108 goes open-circuit due to a blown FET, it will likely damage D2108, and then the PSU would have no feedback to turn the main FET off during the start-up sequence.

So to check D2108, you need to take R2108 out of the circuit. With R2108 out, you should be able to see a good diode reading on your multimeter in the forward direction of D2108. Meanwhile, R2108 should measure less than 1-2 Ohms, depending on how low resistance your multimeter an measure.

Next, diode R2110: again, because of the parallel resistor R2109, you probably won't be able to get conclusive results if D2110 is good or not. So you need to remove either R2109 or D2110 out of the circuit to check them.

While R2108 is out of the circuit, you should also be able to check diode D2109 for a proper forward voltage drop in circuit with your multimeter. Putting the probes in reverse may give you two "diode" forward voltage drops due to transistor Q2102's B-E junction and diode D2108.

Finally, diode D2111: you should get a forward voltage drop both in the normal forward direction and also in reverse, again due to transistor Q2102's B-E diode junction.

[johnfin]

Re: Magnavox DRV power supply SMPS

You mean diode D2110 not R2110 right? I need to wait till my new mosfet comes to do testing. After I install the MOSFET should I test the resistance across the filter cap and if its ok, put the bulb in series with the fuse and try it before I start pulling the micro chips. I just know I will screw those up if I start fiddling with them.

[momaka]

Re: Magnavox DRV power supply SMPS

You mean diode D2110 not R2110 right?

Yes.

I need to wait till my new mosfet comes to do testing.

Ah, I see. You rather try the PSU first again before removing those SMD components. Makes sense if you haven't done much SMD work. That said, SMD isn't hard to do, even with a regular soldering iron - at least the 2-legged components are pretty easy. On my soldering iron, the tip got very work out over time and became in the shape of a spoon. At first, I though that was a bad thing, but it actually makes it even easier to do SMD components and soldering in general. So if you have a pencil tip iron, you can grind down the tip either to a chisel or spoon-like shape. Bevel-shape okay too. Then use lots of flux and solder to remove SMD components... but really mostly flux. Without flux, SMD work is an extreme pain. Not impossible, but close.

After I install the MOSFET should I test the resistance across the filter cap and if its ok, put the bulb in series with the fuse and try it before I start pulling the micro chips.

Yup, that's how I'd do it.
Just make sure the PSU is *not* connected to the DVR.

[johnfin]

Re: Magnavox DRV power supply SMPS

I removed the R2108 resistor which is huge (it tested ok) and diode 2108 was shorted out. I tried to unsolder it and it flew apart. Either it was fried or I botched it. So can I replace it with a very tiny conventional diode?

[momaka]

Re: Magnavox DRV power supply SMPS

Lol, what happened to waiting for the MOSFET to show up?

Anyways, what's done is done.

No, you *cannot* use a small regular diode in place of D2108, because D2108 is a Zener diode. According to service manual, looks like 338sb-t26 (though I can't find a datasheet for that). But you can try it if you like. Worst case, the PSU won't work again.

[johnfin]

Re: Magnavox DRV power supply SMPS

The mosfet came last night and i installed it but i need to find out what is causing it to blow. The resistance on the power cap with new mosfet is 4 megs, thats before the diode incident. I may have some small zeners laying around (clear looking). I may try it.

[johnfin]

Re: Magnavox DRV power supply SMPS

With the new mosfet and zener in the light came on full blast across the fuse. Zener blew but mosfet is ok. Ohms across power cap bounce from zero to inf. On an analog like it should. I cant find a match on that chinese zener either. May try a regular diode there.

[johnfin]

Re: Magnavox DRV power supply SMPS

The diode is 33SBS-T26 Its some chinese zener diode and the online data sheets are garbage. I believe its a 33v zener. I think some matches are
BZX55/C33 33V
1N5257B
568-5899-1-ND

but the original one is SMD mounting.

[momaka]

Re: Magnavox DRV power supply SMPS

The mosfet came last night and i installed it but i need to find out what is causing it to blow. The resistance on the power cap with new mosfet is 4 megs, thats before the diode incident.

So nothing is shorted across the primary anymore then - that's good.

I may have some small zeners laying around (clear looking). I may try it.

Doesn't work like that - you need to know the Zener voltage of the Zener diode, otherwise it's of no use replacing it. Also, not all clear (glass) didoes are Zener diodes and not all Zener diodes are clear (glass) diodes.

With the new mosfet and zener in the light came on full blast across the fuse. Zener blew but mosfet is ok.

In that case, remove the Zener and test the PSU without it. Report back what you find. With the Zener diode shorted, there would be no voltage generated across the current sense resistor, R2108, which means Q2102 may not get turned ON, which in turn means MOSFET Q2101 may not get turned OFF - hence the bulb fully lit.

Ohms across power cap bounce from zero to inf. On an analog like it should.

Yup, that's normal.
Means you've gotten rid of the short-circuit from the bad MOSFET, and not it's just a matter of finding the faulty component that's causing the MOSFET not to turn OFF / switch.

If I had to guess, your Q2102 may have shorted or open pins. Looks like a TO-92 transistor, so you should be able to remove it from the PSU board and test it. Finds its datasheet and test using Diode Test mode on your multimeter.
For an NPN transistor, you should get a diode reading / voltage drop (500-700 mV) when you have the leads connected in the following way:
1) Positive (red) multimeter probe connected to Base lead and negative (black) multimeter probe connected to Emitter pin.
2) Positive (red) multimeter probe connected to Base lead and negative (black) multimeter probe connected to Collector pin.

Also, for an NPN transistor, you should NOT get any resistance reading or diode voltage drop between Emitter and Collector leads.

All of these measurements are out of circuit, of course.

I cant find a match on that chinese zener either. May try a regular diode there.

Just leave it out for now. Zener diode Z2111 should be enough to protect Q2102 from over-current / over-voltage. Test the PSU without it and report back what you find. If still getting a bright bulb, test out Q2102 as I mentioned above. Then while at it, also test some of the diodes I mentioned in my earlier post.

but the original one is SMD mounting.

Don't worry about that. Bad comes to worse, you'll just mount your through-hole diode on the bottom of the PCB. Ghetto, but as long as it works and it's insulated safely, it won't matter.

[Relayer]

Re: Magnavox DRV power supply SMPS

Hello johnfin,
Could you please check any resistors that lead from Pin 1 (Anode side) of the Opto-coupler to see if any have gone high. You need to lift one leg of each resistor to get an accurate reading.
Also, check any zener diodes on the same circuit on the cold side of the PSU.
I remember a particular Magnavox CRT monitor, during my computer monitor repair days, that one of the feedback resistors went high and kept blowing the mains fuse quite violently.
Please let us know how you get on.
Regards,
Relayer

[johnfin]

Re: Magnavox DRV power supply SMPS

I will test those item but in the mean while I am in deep doo doo. The ribbon cable pins broke off with all of the messing around I have been doing on the board. Any thoughts on repairing it. There are about 40 pins.

[momaka]

Re: Magnavox DRV power supply SMPS

The ribbon cable pins broke off with all of the messing around I have been doing on the board. Any thoughts on repairing it. There are about 40 pins.

As long as the connector it goes to is still good, you can find generic ribbon cables like that on eBay and Ali Express for cheap. If not, you can always "make your own" by directly soldering wires between the two boards. Ghetto, but it will work too. Just focus on fixing the PSU first, then you can get to that ribbon cable later. I *don't* think it's gave over just because of that.

[johnfin]

Re: Magnavox DRV power supply SMPS

This is what I have tested lately:

MOMAKA: Tested transistor Q2102, its fine.

RELAYER: Resistors R2233 and R2243 from the opo-coupler are fine, 200ohms and I didnt have to lift them.

Iam going to wait to get the zener replaced before I plug it in.

[momaka]

Re: Magnavox DRV power supply SMPS

MOMAKA: Tested transistor Q2102, its fine.

You sure?
If you tested it with the method I suggested above, then it likely is good. Otherwise, if you used that transistor/component tester again... well, I can't help but be skeptical.

[johnfin]

Re: Magnavox DRV power supply SMPS

Did it the way you said. Can you find a match for the 33sbs diode?

[momaka]

Re: Magnavox DRV power supply SMPS

Can't find anything meaningful for that Zener, but other online posts with similar Zener diode numbers tell me that you're probably right and this is a 33V Zener. Just grab the above one from Digikey you posted and try it.

Like I said, the PSU should work without that Zener diode. There is something else going on is making that bulb turn ON.

If you have a 5-12V double-insulated (non-grounded) power adapter and a 470-1000 Ohm resistor, try this: connect the adapter's minus to the transformer's pin 8 (I'm using the schematic in post #50 as a reference). Then plug in the Magnavox SMPS, with the series light bulb, of course. If it glows, now connect the 470-1000 kOhm resistor to that power adapter's positive pin and connect the other end of the resistor to transistor Q2102 Base. This should make the bulb go off instantly. Let me know if you can do this test and/or if you can confirm the results from it.