EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: zape on November 10, 2017, 07:50:58 am
-
I'm trying to repair this SMPS
It shorts when I plug it in with a bulb in series.
I replaced Mosfet and switching transistors, and also sense resistor because they were blown. also replaced a few zener diodes and shorted zener diode on the secondary side.
I even bought a scope and found out that the switching frequency is 20k and the MOSFET is almost always closed, the PWM pulses on the gate are almost 100% duty cycle. What could be the reason?
Attaching a schematic in the PDF
-
Missing feedback or shorted/overloaded output
Check the circuitry around D660 and the optocouplers.
-
I checked and actually replaced D660 it works fine.
Will check the optocoupler
The output voltage seem to be correct +21V but voltages on the optocoupler (on the secondary side) are 1.2-3.2v so it looks like D660 is pulling it to the ground.
-
Hi,
You should check the ohms of R654 and R655 with power off.
;D
-
R654 and R655 are good. I think I checked every single resistor on the primary side and around feedback area on the secondary
-
I now checked the optocoupler (it was bad) and replaced it.
The same result.
Checked the optocoupler and it still works.
Any ideas?
-
R654 and R655 are good. I think I checked every single resistor on the primary side and around feedback area on the secondary
Yeap, the secondary side should be In order, 50mA flowing across the resistors is well within the capability of SL431A.
Edit: You sure you checked D605? How did you check it?
Is the output voltage correct?
-
I checked it in-circuit with multimiter like a normal diode. It conducts one way. I did not check the Zener voltage.
The output voltage is +21V and it is correct.
-
I checked it in-circuit with multimiter like a normal diode. It conducts one way. I did not check the Zener voltage.
The output voltage is +21V and it is correct.
So your question is actually on the PWM, Why it's on 100% duty as you purported by measuring the voltage of 4.9volts. The gate is always power "ON" by the resistors divider and zener diode at the gate. You can see for yourself. So voltage measurement at the Gate is surely incorrect. It will be pulsed "OFF" by the transistor.
No it is not 100% duty, You should actually see it on a oscilloscope probe it onto R611 [with isolating transformer].
The wave is actually by number of pulses. As you load it more, more pulses rises to appear. As you load it less, less pulses will appear.
This power supply is BAD, at most 55% efficient and will fail horribly.
Edit: To put it simply, this is not a PWM switching power supply, it is a PFM switching power supply. ;D
-
Here are some photos:
I used X10 probe (and I use Backup UPS instead of isolation transformer, because I don't have one)
-
Load up the power supply say 0.5A, 1,A, 2A [check whether it is able] and probe across R611.
Then you see the PFM properly.
Edit: I am not sure, no knowledge about backup UPS or whether it's isolated. ??? :-// :-// .
-
Ok the only load I was able to find around was electric boiler which measures around 10Ohm
The output voltage drops from 21 to around 3v when I connect it.
Attached scope photos:
-
Ok the only load I was able to find around was electric boiler which measures around 10Ohm
"Electric Boilder" and its 100 ohm ???????
How did you measure the ohm? By multimeter?
Edit: Measure the voltage at the opto-coupler, whether the SL431 releases the current.? or same as before?
-
10, not 100
Yes used multimiter
-
10, not 100
Yes used multimiter
Measure the voltage at the opto-coupler, whether the SL431 releases the current.? or same as before?
BTW: what is the Max current of this power supply?
-
Yes it does.
With load connected the opto-coupler is off. No voltage drop on it, and all voltages are equal to ~3v output.
The TV is rated at 90W so I guess the max current is about 2-3A x 21v
-
Yes it does.
With load connected the opto-coupler is off. No voltage drop on it, and all voltages are equal to ~3v output.
The TV is rated at 90W so I guess the max current is about 2-3A x 21v
So is a TV power supply... There are the command On signal from the TV. I suggest you plug this power supply back into the TV to test it.
There are also the 3.3v, 13v etc...
-
I forced the PSU to turn on.
I don't think it should drain so much power when powered so the bulb lights up
-
you should measure all the secondary power whether is OK, e.g 13V, 3.3 V, 21 V etc.....
Anyone shorted.?
Edit: What bulb lights UP? What is so much power?
Edit: I know this type of power supply is 50% efficient. From your scope trace, I deduced 19W is expended somewhere, don't know if you have connected something consuming 19W or secondary voltage of one of them is shorted.
-
Nothing shorted on the secondary side. all voltages are fine
I learned this "bulb trick" here: you connect your device not directly to mains but through a 100W light bulb in series that acts as protection in case the power supply is draining too much power (as in my case)
If I connect it directly to the mains the fuse will blow up and probably mosfet and others (I learned this the hard way lol)
-
Nothing shorted on the secondary side. all voltages are fine
I learn this "bulb trick" here: you connect your device not directly to mains but through a 100W light bulb in series that acts as protection in case the power supply is draining too much power (as in my case)
The trick is for initial testing of power supply incase short circuit. If is not short circuit anymore, you should take the series light bulb out.
-
but it is shorted - the bulb is very bright
-
Nothing shorted on the secondary side. all voltages are fine
I learned this "bulb trick" here: you connect your device not directly to mains but through a 100W light bulb in series that acts as protection in case the power supply is draining too much power (as in my case)
If I connect it directly to the mains the fuse will blow up and probably mosfet and others (I learned this the hard way lol)
If all the voltages are fine, just plug it back to the TV. ;D
WITHOUT THE SERIES LIGHT BULB...... hahahahahah LOL.
-
but it is shorted - the bulb is very bright
Hello, how many watts is the series light bulb you used.? What bulb is that? Is it incandescent bulb?
The mosfet is only firing at 0.25A. [from your scope].
What is the AC current flowing through, can your backup power supply show the current, if not, measure the AC current.
Don't let the bulb trick you.
-
Man!..... to think that you want to do a load test with a series light bulb dragging the voltage in circuit...... |O
-
I have two bulbs 40W each so it is 80W
UPS shows 73W and it does not change if I connect the load
So the AC current is 0.6A
-
I have two bulbs 40W each so it is 80W
UPS shows 73W and it does not change if I connect the load
So the AC current is 0.6A
It is not short circuit. the bulb does light up in normal situation. You can take out the light bulb and plug it directly into the main to see what brightness is short circuit intensity.
Yes you learnt the bulb trick, but the teacher was not complete...... hahahahahahah LOL.
Edit: You are not using incandescent light bulb right...? hahahahaahahah LOL
Edit: And then maybe your backup power supply is charging the internal battery....... hahahahahaahha LOL....
Oops, sorry,..... just a lighter moment for all of us...
I have never used a UPS before, hence cannot recognise all the problems associated with it... example power limit etc...
-
It is an incandescent light bulb of course.
HMM let me check one thing
-
It is an incandescent light bulb of course.
I lights up exactly the same if I just plug it into mains
it depends before the UPS or after the UPS.
Do you think 0.6A is a short circuit?
Edit: 0.6A is not a short circuit. the series light bulb is only used for initial testing for short circuit. In this case, the light bulb as you used it with the UPS cannot discern or provide you with a useful cue between short and 0.6A.
So, you should plug back to the TV and turn it on.
:D
-
HMM let me check one thing
Don't say you wired it in parallel?
-
IT WORKS! :-+ :-+ :-+ :-+ :-+
So how should I properly use this light bulb trick then?
-
:-+ :-+ :-+ :-+ :-+ Great news.
The light bulb trick is for initial testing in case of short circuit.
Purpose is to limit the voltage and current through the device under test.
So for me I would size the light bulb to be 20% of the full load power of the device under test.
So, if for example the DUT is 20 watts, so the light bulb should be 100W [around to your discerning ability of shorts].
The sizing is not for the light bulb to lights up or maybe abit.
the light bulb must be incandescent light bulb.
you must wired in after the isolating transformer [after any device that consume power] and in series to the DUT.
The isolating transformer must have the power sufficient for the DUT X 2.
I have 1500VA isolating transformer to meet the occasion. ;D
-
The function of Q621 makes no sense. Its Collector would have to connect to the Gate of the MOSFET to reduce the MOSFET's on time, in case of excessive line voltage.
-
The function of Q621 makes no sense. Its Collector would have to connect to the Gate of the MOSFET to reduce the MOSFET's on time, in case of excessive line voltage.
its "over voltage protect" not for "output over voltage" but for Q602 if the induced voltage at the Base continue to rise even after it had pulsed. ;D
-
The function of Q621 makes no sense. Its Collector would have to connect to the Gate of the MOSFET to reduce the MOSFET's on time, in case of excessive line voltage.
its "over voltage protect" not for "output over voltage" but for Q602 if the induced voltage at the Base continue to rise even after it had pulsed. ;D
Yes, that's the way it might seem from the way it's drawn. Not useful at all.
D621 rectifies the + half cycle or the forward half of the waveform rather than the fly-back half. The voltage rectified by D621 will be proportional to the incoming line voltage.