SMPS low Vout

[R_J]

Re: SMPS low Vout

If you remove D661, that will eliminate any problem on the 20VPFC. The power supply will work with D661 removed (no shorting)
This is such a simple circuit there is not much left, Maybe the transformer primary was damaged?

[madan1]

Re: SMPS low Vout

Could you be measuing those opto pins 3 & 4 wrong? reversed? The reason I ask is that the top247 should be outputing around +5 volts on its own

You are totally correct.
Indeed the pins are reversed.. I got it wrong because on the schematics the IC pinout is mirrored.
So on IC652.3 and IC651.1 there is 5.2V.
On IC652.4 - 0.5V
On c661 - 0.39V

[R_J]

Re: SMPS low Vout

Yes, C661 negative is what we refer to as HOT ground.
So pin 1 (C) of ic651 is only .46 volts? that can't be right

[madan1]

Re: SMPS low Vout

Yes, right - c661 - 0.4V
For the measures on the primary side I used for reference the negative pin of c661 ( I do not know if the term in English is "hot ground" ).
I hooked the scope to the secondary side of the transformer and yes - something is going on ( see photo ), but still no voltage on the rails ( c707-0v, c708-0.07v, c709-0.15v )
From the VPP on transformer pins 2 (0.22), 3(0.4) and 4(1) at least it seems that the secondary coils are fine.


edit -

..Try removing one end of D661, ...With it removed, see if you get your +12v

Please clarify - remove the diode and leave the trace empty, or remove the diode and shorten its pads?

[R_J]

Re: SMPS low Vout

You must mean C661 0.4v
Is pin 4 of opto 5.2 volts measured to hot ground? If it is, the power supply must be trying to work and giving you that voltage, that means you should have something on +12, I would think at least 4-5 volts
Could you be measuing those opto pins 3 & 4 wrong? reversed? The reason I ask is that the top247 should be outputing around +5 volts on its own

You should also be getting more than .4 volts on C661
If D662 is producing 5.2v. D661 should produce at least that much, Try removing one end of D661, Its only needed for the pfc circuit which is not working now anyway. With it removed, see if you get your +12v

[madan1]

Re: SMPS low Vout

If this top247 circuit is not working, you won't have 20VPFC to turn on the pfc circuit.
You may want to check D681 and make sure nothing is shorted off that line,
Are any of the 12,5,2.5 volt lines shorted or is the power supply removed for these tests?

I also thought the same about the 20V rail, but was not sure.
D661 works as a diode when tested by DMM and external power supply, what other tests I can do?
I also injected 16volts on c661 and got only 2.3mA leak.

The psu is tested totally out of the circuit - it does not require control signal to start, so no need for the rest of the boards. There is no load on the lowV connector. Visually everything on the secondary side looks fine, but I have not tested almost anything there.

I just measured some more voltages:
on c681 - 0.4V

on pins 1 and 2 on the optoC - 0.13V
on pin 3 - 0.46v
on pin 4 - 5.2V

[R_J]

Re: SMPS low Vout

If this top247 circuit is not working, you won't have 20VPFC to turn on the pfc circuit.
You may want to check D681 and make sure nothing is shorted off that line,
Are any of the 12,5,2.5 volt lines shorted or is the power supply removed for these tests?

[madan1]

Re: SMPS low Vout

maybe? what resitance do you get? as in xxxΩ
The transformer resistances are likely ok, and not open.
Are you getting any voltage at all on the secondary? even a couple volts

for the octoC
when using the positive probe on pin 3 and negative on pin 4 I get between 120ohms and 38k
when the positive probe is on pin 4 and the negative on pin 3 - 60ohms - 10m
the values change parabolically to the applied voltage ( start high, go low and then again high with the V increase ).
On the pins for the voltage output I get millivolts - 0.15-0.20.
The high V line still is 320 instead of 380.

[R_J]

Re: SMPS low Vout

Just tested and the optocoupler - when voltage is applied on the diode side, it changes its resistance on the other side. Is it normal to be conductive in both directions ( on the transistor(?) side )?

maybe? what resitance do you get? as in xxxΩ
The transformer resistances are likely ok, and not open.
Are you getting any voltage at all on the secondary? even a couple volts

[madan1]

Re: SMPS low Vout

Most of the time, yes you can detect leakage with a DC ohmmeter though it can be a test of patience to wait for the capacitor to charge. Ideally leakage is tested at the rated voltage which a simple ohmmeter can't do.

I tested all elements from the TOP247 circuit with an external power supply and did not find any leaking diodes or caps.
The resistors checked with DMM and again - all read fine.
Just tested and the optocoupler - when voltage is applied on the diode side, it changes its resistance on the other side. Is it normal to be conductive in both directions ( on the transistor(?) side )?
Got a new TOP247 and it shows higher resistance between S and D, so I guess the other one was faulty.
Unfortunately in the local shop they did not have smd caps, so for now I'll be troubleshooting without the c653.
Tested again out of the board c654 and c661. Used external power supply to charge them and a transistor tester. Again - the capacitance is fine, the ESR is fine, they do not selfdischarge or conduct current. Anyway, this time I'm switching them with new ones just because I'm fed up with testing them .
Should I desolder the transformer to test it?
On the board I get the following readings:

7-8 - 0.8ohm
8-9 - 0.1-0.2ohm
9-10 - 0.1-0.2ohm
8-10 - 0.3-0.4ohm

5-4 - 0.1-0.2ohm
4-2 - 0.1-0.2ohm
2-1 - 0.1-02ohm

[R_J]

Re: SMPS low Vout

Yes you can use a .1µf, check scrapped boards, it helps if you have a schematic of the scrapped board to find a smd cap.

[eccerr0r]

Re: SMPS low Vout

Most of the time, yes you can detect leakage with a DC ohmmeter though it can be a test of patience to wait for the capacitor to charge. Ideally leakage is tested at the rated voltage which a simple ohmmeter can't do.

[madan1]

Re: SMPS low Vout

I would replace the broken cap
This is from the pdf:
An external bypass capacitor closely
connected between the CONTROL and SOURCE pins is required
to supply the instantaneous gate drive current. The total amount
of capacitance connected to this pin also sets the auto-restart
timing as well as control loop compensation.
If R653was open I would also suspect the optocoupler may have been damaged, you should also check the resistance of the transformer between pins 9 & 10

And bad caps CAN show a higher value on a cap meter.

I guess I was looking at this:
"In addition to the 47 μF CONTROL pin capacitor, a high
frequency bypass capacitor in parallel may be used for
better noise immunity."
I will try to find one, but what do you think 0.1UM means ( quoted from the SM )? Is this 0.1uF? The cap is of the smallest smd type and is real pain in the a** to solder it in that overpacked space.

[madan1]

Re: SMPS low Vout

yes, but if it's leaky it will show higher capacitance.
the meter charges the cap and times it - the longer it takes then the higher the capacitance.
if it's leaky it takes longer to charge and fools the meter.

Will a simple DMM resistance test show the leak?

[R_J]

Re: SMPS low Vout

In the process of testing, I broke c653.

I would replace the broken cap
This is from the pdf:
An external bypass capacitor closely
connected between the CONTROL and SOURCE pins is required
to supply the instantaneous gate drive current. The total amount
of capacitance connected to this pin also sets the auto-restart
timing as well as control loop compensation.
If R653was open I would also suspect the optocoupler may have been damaged, you should also check the resistance of the transformer between pins 9 & 10

And bad caps CAN show a higher value on a cap meter.

[stj]

Re: SMPS low Vout

yes, but if it's leaky it will show higher capacitance.
the meter charges the cap and times it - the longer it takes then the higher the capacitance.
if it's leaky it takes longer to charge and fools the meter.

[madan1]

Re: SMPS low Vout

I checked them with "transistor tester" and by "higher capacitance" I mean like 50uF and ESR 0.38 ( real values for c654 ), when it should be 47uF.
This is the only electr. cap in the TOP247 circuit. There are two more which I tested outside the circuit - c661 and c620 and they also showed similar values.
By the way, is it possible to have a leaky cap and still having capacitance detected by a multimeter?

[stj]

Re: SMPS low Vout

It's an old PSU indeed, but all caps are nichicon, I have desoldered some of them and they all read fine ( even more than what is stated on their labels )

that's a red flag!
did you check ESR & Leakage, or just the capacitance?

when caps go leaky they often start to read a higher capacitance.

[madan1]

Re: SMPS low Vout

It's an old PSU indeed, but all caps are nichicon, I have desoldered some of them and they all read fine ( even more than what is stated on their labels ).

Unfortunately I'm still not able to fix it so some assistance will be appreciated.
OK, so the story so far - the topic label is wrong. All voltage rails were present and it was my bad not measuring them fine.

After some testing of the unit, the PSU just popped. I guess I have shorted something, but no obvious signs - no burn marks, no smoke.. nothing, but a loud explosion and pieces of TOP247 everywhere.
I replaced the TOP + R653 ( it was reading 40+K instead of 6.8R ). All other components seem to be fine - the resistors show the expected values, the diodes work in only one direction, the caps have their capacitance.
In the process of testing, I broke c653. Should I replace it or it will work without it ( the default TOP247 datasheet does not show it )?
The 20VPFC rail is missing, thus the 380V controller also does not work and there is only 320V on the highV rail.
When powered up, the TOP247 gets hot, but that is the only sign of life I got from it.
Is it possible that I have bought a faulty IC or there is something else?

[ktmmotocross]

Re: SMPS low Vout

recap is most common repair on PSU so i recommend to do it as first

especially if it is 10years old