Author Topic: PSU failed due to overheat  (Read 1258 times)

0 Members and 1 Guest are viewing this topic.

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
PSU failed due to overheat
« on: May 22, 2018, 01:59:08 am »
Hello everybody,
My 220v 12v 1600w power supply just failed, my best guess is due to overheating, I wasn't at the place when it failed. Now I checked every component I could do with my dmm (caps, inductors, resistors, diodes, relay) and except for the 100c thermal fuse (edit: even that was doing fine) every item was OK, replaced the fuse but it still doesn't work. Its relay doesn't click and output is 0.00 volts.
Also, the filtering capacitors get charged up after I plug it in, so I think the bridge rectifier is also ok.
There's also a NCP1399 IC which the thermal fuse was connected to its SKIP pin, and I don't know any way to test this guy.

Any guesses which would be the cause?
I could post the pics if needed.
« Last Edit: June 14, 2018, 05:08:40 am by m.m.m »
20 years old Electronics Lover
 

Offline Samogon

  • Frequent Contributor
  • **
  • Posts: 425
  • Country: us
Re: PSU failed due to overheat
« Reply #1 on: May 22, 2018, 05:34:54 am »
Pictures will help indeed. Do both sides of PCB.
 

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #2 on: May 24, 2018, 01:15:29 am »




20 years old Electronics Lover
 

Online wraper

  • Supporter
  • ****
  • Posts: 8654
  • Country: lv
Re: PSU failed due to overheat
« Reply #3 on: May 24, 2018, 01:25:47 am »
There's also a NCP1399 IC which the thermal fuse was connected to its SKIP pin, and I don't know any way to test this guy.
What thermal fuse? I guess you opened the datasheet but did you try reading what that pin is used for?  :palm:. If there was a resistor which blew up, there is no way that NCP1399 survived.
« Last Edit: May 24, 2018, 01:29:11 am by wraper »
 

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #4 on: May 24, 2018, 01:51:22 am »
What thermal fuse? I guess you opened the datasheet but did you try reading what that pin is used for?  :palm:. If there was a resistor which blew up, there is no way that NCP1399 survived.
The thermal fuse is NC, so if the SKIP pin on the IC opens up the IC dies? What a shitty design, they should have connected a large resistor in parallel then shouldn't they?  :o
It says in the datasheet "Sets the skip in threshold via a resistor connected to ground – version NCP1399Ay"  :-\

edit:
the thermal fuse is connected in series with a 2.7k resistor to the ground pin.
« Last Edit: May 24, 2018, 02:03:17 am by m.m.m »
20 years old Electronics Lover
 

Online wraper

  • Supporter
  • ****
  • Posts: 8654
  • Country: lv
Re: PSU failed due to overheat
« Reply #5 on: May 24, 2018, 02:06:30 am »
There is no thermal fuse an cannot be by design. Dunno what you mean by
Quote
and except for the 100c thermal fuse every item was OK
What the hell is that 100c?
Quote
so if the SKIP pin on the IC opens up the IC dies? What a shitty design
:palm:
There can be either resistor connected or pin left unconnected. Dunno out of where you pulled a stupid idea about some thermal fuse.
 

Offline HecticZA

  • Contributor
  • Posts: 41
  • Country: za
Re: PSU failed due to overheat
« Reply #6 on: May 24, 2018, 02:46:28 am »
There is no thermal fuse an cannot be by design. Dunno what you mean by
Quote
and except for the 100c thermal fuse every item was OK
What the hell is that 100c?
Quote
so if the SKIP pin on the IC opens up the IC dies? What a shitty design
:palm:
There can be either resistor connected or pin left unconnected. Dunno out of where you pulled a stupid idea about some thermal fuse.
I think he is referring to a Ksd thermal switch and a temperature of 100°C.

Sent from my SM-N950F using Tapatalk

 
The following users thanked this post: m.m.m

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #7 on: May 24, 2018, 02:50:45 am »
What the hell is that 100c?
I meant 100 degrees Celsius, sorry if that was confusing,
:palm:
There can be either resistor connected or pin left unconnected. Dunno out of where you pulled a stupid idea about some thermal fuse.
There was a 100 degrees Celsius ceramic NC thermal fuse connected to the skip pin in series with a 2.7k resistor as I said, which blew up and now I replaced it with a metal thermal fuse which you can see in the top of the third picture I posted.

I think he is referring to a Ksd thermal switch and a temperature of 100°C.

Exactly, thanks for the name, didn't know it's called KSD switch.  :D

---
EDIT:
There's also a PFC ic with PN: ice3pcs01g
« Last Edit: May 24, 2018, 05:25:06 am by m.m.m »
20 years old Electronics Lover
 

Online Gregg

  • Frequent Contributor
  • **
  • Posts: 416
  • Country: us
Re: PSU failed due to overheat
« Reply #8 on: May 24, 2018, 06:33:20 am »
D13 almost center on back of the board looks like smoke was issued at its expense
 

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #9 on: May 24, 2018, 06:53:53 am »
D13 almost center on back of the board looks like smoke was issued at its expense

It's the flux remains on the part, it shows fine on the meter, 0.400v forward drop just like the D14. Cleaned it up with my finger nails and it looks fine.
20 years old Electronics Lover
 

Online wraper

  • Supporter
  • ****
  • Posts: 8654
  • Country: lv
Re: PSU failed due to overheat
« Reply #10 on: May 24, 2018, 10:24:10 am »
Check voltage on HV input pin of NCP1399. BTW I guess IC should be NCP1399B which has REM, not skip pin. Because for skip pin NC thermal protector does not make sense (A version). Are you completely sure that original one was not NO thermal switch?
 
The following users thanked this post: m.m.m

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #11 on: May 24, 2018, 05:37:51 pm »
Check voltage on HV input pin of NCP1399. BTW I guess IC should be NCP1399B which has REM, not skip pin. Because for skip pin NC thermal protector does not make sense (A version). Are you completely sure that original one was not NO thermal switch?
Yes you are right, I first thought the switch is broken but now just double checked and realized it's fine, meaning it's a NO switch. sorry for that.
For the voltage on HV pin I'm a bit scared to probe it when it's connected to the mains :-[, so first unplugged it and then probed it, shows the same voltage the filtering caps are at (quickly discharging, around 20v when I put my meter).

Also another interesting thing just noticed is that I checked its power factor when plugged in and realized it's at 0.02! meaning that the PFC section is dead?
« Last Edit: May 24, 2018, 06:31:00 pm by m.m.m »
20 years old Electronics Lover
 

Online wraper

  • Supporter
  • ****
  • Posts: 8654
  • Country: lv
Re: PSU failed due to overheat
« Reply #12 on: May 24, 2018, 07:32:27 pm »
Also another interesting thing just noticed is that I checked its power factor when plugged in and realized it's at 0.02! meaning that the PFC section is dead?
No, it means that circuit is not consuming power. In any case, without measuring voltages it would be hard to find the issue. Don't forget to return back old thermal switch.
BTW, PFC needs minimal load, so it normally should not work when PSU is not running for some reason.
« Last Edit: May 24, 2018, 07:42:38 pm by wraper »
 
The following users thanked this post: m.m.m

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #13 on: June 13, 2018, 12:03:23 am »
Hello again everyone, I've been troubleshooting the PSU for a while now, and yet I haven't found any faulty components. I've also tested every single transistor and varistor (for varistors just checked for shorts). The only thing that caught my attention was this tiny TEA1995 IC which controls the output-side MOSFETs which has turned its color, and might be dead.
Checked the Vcc on it and it's sitting at 0v, seems to be connected to the output positive side. The GND on the IC is also connected to output GND.

edit:
The PFC section also has to turn on an inrush-relay after a few ms connected to the mains but it doesn't, which I guess would be because of it being not loaded  :-\
And for the MOSFETs I just checked them with my dmm diode test and they were OK too.

edit2:
Also the Vref pin on the PFC IC is at 5 volts so I think that one's normal too.
« Last Edit: June 13, 2018, 01:13:34 am by m.m.m »
20 years old Electronics Lover
 

Offline shakalnokturn

  • Regular Contributor
  • *
  • Posts: 194
  • Country: fr
Re: PSU failed due to overheat
« Reply #14 on: June 13, 2018, 08:35:21 am »
Did you check the through-hole resistor hidden in shrink-tube that's parallel to the relay contact?
Relay contact resistance is also something to consider.
 

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #15 on: June 13, 2018, 09:21:11 am »
Did you check the through-hole resistor hidden in shrink-tube that's parallel to the relay contact?
Relay contact resistance is also something to consider.
Yup, it's a resistor in parallel with the NTC and relay contacts, and it reads just fine.   :-DMM
20 years old Electronics Lover
 

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #16 on: June 14, 2018, 01:56:16 am »
Another update, just measured some voltages on some pins, and unfortunately I don't have any differential probe or pocket oscilloscope around to check the waveforms, but what I measured with my dmm is heavily unstable and rippling. for both NCP1399 and ICE3PCS01G ICs the VCCs ripple between 14v to 9v! And the HV pin on NCP1399 also ripples between 400 to 280 volts!  :-//
Any idea about what's wrong?
20 years old Electronics Lover
 

Offline shakalnokturn

  • Regular Contributor
  • *
  • Posts: 194
  • Country: fr
Re: PSU failed due to overheat
« Reply #17 on: June 14, 2018, 02:39:28 am »
PFC hiccuping.
Where does the ICE3PCS01G get it's Vcc from?
 
The following users thanked this post: m.m.m

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #18 on: June 14, 2018, 05:06:38 am »
PFC hiccuping.
Where does the ICE3PCS01G get it's Vcc from?
The both ICs share the same Vcc and GND and the Vcc is connected to Q19 collector, then passed through D9 and connected to Q21 emitter. I think it just uses a simple BJT linear regulator configuration since I see no regulars at all.
« Last Edit: June 14, 2018, 08:31:22 am by m.m.m »
20 years old Electronics Lover
 

Offline shakalnokturn

  • Regular Contributor
  • *
  • Posts: 194
  • Country: fr
Re: PSU failed due to overheat
« Reply #19 on: June 14, 2018, 09:13:26 am »
On most designs there is a start up circuit that will supply the primary side IC's (insufficient current for steady running the PWM, but enough for a couple of seconds) once the switching has kicked-in there will usually be a dedicated winding from the transformer to take over supplying the IC's.

I'm guessing that will be the two thin wires on the transformer, anything wrong between there and the IC's Vcc?
You have checked the secondary side for short circuits, haven't you?

Another thing I would check on a half bridge supply is that the values of the two grey MKP capacitors (the ones wedged against the transformer) haven't gone too low, they are responsible for getting all the energy into the transformer so they have a hard time!

 
The following users thanked this post: m.m.m

Offline m.m.m

  • Regular Contributor
  • *
  • Posts: 83
  • Country: ir
  • EE Student and Hobbyist at IUST
Re: PSU failed due to overheat
« Reply #20 on: July 06, 2018, 05:10:03 am »
On most designs there is a start up circuit that will supply the primary side IC's (insufficient current for steady running the PWM, but enough for a couple of seconds) once the switching has kicked-in there will usually be a dedicated winding from the transformer to take over supplying the IC's.

I'm guessing that will be the two thin wires on the transformer, anything wrong between there and the IC's Vcc?
You have checked the secondary side for short circuits, haven't you?

Another thing I would check on a half bridge supply is that the values of the two grey MKP capacitors (the ones wedged against the transformer) haven't gone too low, they are responsible for getting all the energy into the transformer so they have a hard time!


Sorry for the delay, I had a bunch of finals I had to study for.

I still couldn't find any problems. MKPs are fine, the ohm meter reads short circuit on the secondary side because of the huge output capacitance, after a while the resistance shown rises though. the two thin wires on the transformer as you also said feeds the ICs, but there's no problem between there and the ICs too. the problem is that non of the transformers pairs (including the one feeding the logic circuits) carry any voltage (AC) at all. :-\
20 years old Electronics Lover
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf