Switching Power Supply - SMPS Transformer Low Output

[2X]

Hello,
I have a switching power supply that after the SMPS transformer on the LM7812 input I measure 5.4V instead of 17V. I changed all the electrolytic output capacitors (besides that all the old capacitors where I measured didn't lost their capacitance). I measured all the resistors, diodes and capacitors that you see on the schematic where I recreated. I measure the "KA5M0380RTU" with a multimeter in diode mode and it seems ok, but when I measure on the "Drain" pin with the EEvblog 121GW Multimeter in duty cycle mode and it shows 95% (like a DC?) - in DC Volt mode shows 315V at the "Drain" pin. I can't find an obviously faulty component... in my opinion seems that the "KA5M0380RTU" don't oscillating (his internal oscillator). Also, I can't find any datasheet for the SMPS transformer so to see how the windings are wired. Does anyone have a datasheet for this transformer or knows a way to find how the windings are formed? Also, I measure with the resistor mode on the multimeter the output of the optotransistor and at 5.4V on his input it shows O.L. but if I inject 17V from my bench power supply the resistor between C-E goes down to Kohm.  I don't know what else I have to measure to find the fault component. Any ideas about how to find what is the fault are welcomed.

[fzabkar]

Is there a small electrolytic capacitor on the Vcc pin? If so, I would replace it.

To me, it sounds like the supply may be hiccuping.

[2X]

Is there a small electrolytic capacitor on the Vcc pin? If so, I would replace it.

To me, it sounds like the supply may be hiccuping.

These capacitors I changed. There is a capacitor at the input pin of the LM7812.

[asis]

Hi,

This document contains typical wiring diagrams for the KA5xx series.

[fmashockie]

This is more just a general tip because it's hard to provide specific troubleshooting info based only on the info you've given. But in my experience, I've had similar issues with SMPS where an interruption in the feedback circuit was the problem. Such as a zener diode going faulty.  I would check the feedback circuit for faults.

[magic]

Not a bad idea. Check what's the voltage on the input and output side of PC5 optocoupler.
If you have a scope, check if the output rail is stable or if it pulses.

I measured all the resistors, diodes and capacitors that you see on the schematic where I recreated.

I think you forgot to upload the schematic.

Regarding the transformer, it should have three windings. One primary, one secondary, one auxiliary which supplies VCC of the KA5M0380RTU.

[asis]

Hi,

@fmasockie
There are general principles for constructing such PSUs and the circuitry is approximately the same.
If you don't see a specific circuit diagram, you need to reverse engineer to read the circuit in detail and find the correct solution.
This is a difficult job, but it educates and gives the necessary experience and confidence in their actions.
Nobody bothers you to apply voltage from a laboratory power source to the circuit in which the TL431 is installed and through the increase / decrease voltage to determine the response threshold of the optocoupler on the side of the photodetector.
Naturally, the source under study should not be connected anywhere.

@X2
Your photos do not show the entire power supply.
Possibly active PFC.
Specify which components are on the PCB's which are mounted vertically.
Dismantle the ECAP C5 and check its condition (preferably with a dial gauge).

[2X]

Not a bad idea. Check what's the voltage on the input and output side of PC5 optocoupler.
If you have a scope, check if the output rail is stable or if it pulses.

I measured all the resistors, diodes and capacitors that you see on the schematic where I recreated.

I think you forgot to upload the schematic.

Regarding the transformer, it should have three windings. One primary, one secondary, one auxiliary which supplies VCC of the KA5M0380RTU.

"I think you forgot to upload the schematic."
You have right.. for some reason didn't upload. Thanks for your reply.


"Regarding the transformer, it should have three windings. One primary, one secondary, one auxiliary which supplies VCC of the KA5M0380RTU."
The SMPS transformer has 5 pins from each side, so how you beleive it goes (what is the purpose of the 5pin?)? If it has six pins I would say that must be 3 windings on the primary side. The Vcc of KA5M0380RTU suppluied from a voltage divider.

[fmashockie]

I made my post before OP shared the schematic he had written out. As you said asis, troubleshooting a SMPS is difficult without a schematic, let alone trying to troubleshoot it from someone's photos online. Which is why I gave a general tip. Was not applying your suggestion or mine was the correct one. Just thought I'd share a similar experience where I had this exact problem and feedback circuit was the issue.  Did not mean to apply not to check anything else.

[magic]

OK, so first of all, GND is not the same as GND, because primary "ground" is the negative output of the rectifier, at mains voltage during negative half-cycles.
Hopefully you know about it already, but it's important to point it out

The power supply is not a voltage divider. The 200kΩ provides small current which charges a Vcc capacitor (there should be one somewhere) and when voltage is high enough, the chip starts working. Then a winding between pins 2 and 4/5 supplies much more power through D19 and R43. A failure of this circuit would cause the converter to work in short pulses, discharging its Vcc supply and waiting for it to recharge through 200kΩ. If R43 and D19 are good, check solder joints and continuity through related PCB traces and from pin 2 to pin 4.

On the secondary, check if Vref is 2.5V or less and check voltage across R55/LED.

Check voltage at the FB pin (carefully, it's all high voltage stuff).

[2X]

OK, so first of all, GND is not the same as GND, because primary "ground" is the negative output of the rectifier, at mains voltage during negative half-cycles.
Hopefully you know about it already, but it's important to point it out

The power supply is not a voltage divider. The 200kΩ provides small current which charges a Vcc capacitor (there should be one somewhere) and when voltage is high enough, the chip starts working. Then a winding between pins 2 and 4/5 supplies much more power through D19 and R43. A failure of this circuit would cause the converter to work in short pulses, discharging its Vcc supply and waiting for it to recharge through 200kΩ. If R43 and D19 are good, check solder joints and continuity through related PCB traces and from pin 2 to pin 4.

On the secondary, check if Vref is 2.5V or less and check voltage across R55/LED.

Check voltage at the FB pin (carefully, it's all high voltage stuff).

Thanks for the point "OK, so first of all, GND is not the same as GND, because primary "ground" is the negative output of the rectifier, at mains voltage during negative half-cycles." I corrected it.
Finally the problem was on the capacitor at the VCC pin on the "KA5M0380RTU" switch. When I measured the voltage on the capacitor "C5-47uF" the voltage trembling between 12.8V-13V. I disconnected the capacitor and I measured it and from 47uF were 0.76uF. With the new capacitor the voltage at the Vcc pin was 13V stable. Thank you very much for the help. I have one more question (or I have to create a new topic?)... I can't understand how the SMPS transformer windings are. I disconnected the transformer and I measured all the pins (the results are on the attached picture on the table).

[magic]

It looks like there are three windings on the primary:
1 - 2
2 - 3
4 - 5

The first two are connected at pin 2 and can be used as a single longer winding. This appears to be the primary.
The 4-5 winding should supply power to the chip through D19. I think you have made a mistake on the schematic.

[fmashockie]

Glad you were able to find the source of the problem! I attached a schematic of the transformer below (it's based on your measurements).  You can clearly see from your photos which pins aren't a tap of the transformer because there are no windings soldered to them (pins 10, 8, 6).  They are probably no connection (NC) and are just for structural support or to ground the ferrite core. 

[2X]

I finished the complete design. I beleive my schematic is right. I want to ask you one more thing about the SMPS transformer. What is the "right" names for the pins on the primary side? On the secondary I put N2+ & N2-.

P.S. I changed the pins sequence (pinout) of the transformer in order to much at my transformer.

[2X]

It looks like there are three windings on the primary:
1 - 2
2 - 3
4 - 5

The first two are connected at pin 2 and can be used as a single longer winding. This appears to be the primary.
The 4-5 winding should supply power to the chip through D19. I think you have made a mistake on the schematic.

Thanks again for the reply. Do you believe it is impossible the pins 4-5 (1-2 in my new design) are not in use?

[magic]

They are in use.

On the schematic, the green 2 near D19 and the transformer means that there is a connection between these points.
Such connection makes perfect sense and I'm pretty sure you will find that it exists on your board.

This winding has small number of turns, like the secondary, and it produces a relatively low voltage. It's used to power the chip.

[2X]

They are in use.

On the schematic, the green 2 near D19 and the transformer means that there is a connection between these points.
Such connection makes perfect sense and I'm pretty sure you will find that it exists on your board.

This winding has small number of turns, like the secondary, and it produces a relatively low voltage. It's used to power the chip.

From the pictures you have right (again) and also as I saw the pin 4 is connected nowhere. The strange thing if you see on the top layer of the PCB the pins 1,9,10 are connected together which means there is no isolation (or am I wrong - why to do this?)

[magic]

 

Right, it looks like the secondary is really connected to mains rectifier negative.

I don't know why, but it's not a problem if this PSU sits inside something and nobody can touch circuits powered by the secondary from outside.

[2X]

Right, it looks like the secondary is really connected to mains rectifier negative.

I don't know why, but it's not a problem if this PSU sits inside something and nobody can touch circuits powered by the secondary from outside.

There is a reason (advantage) to do this (the secondary is really connected to mains rectifier negative)? If you have any link to read about this "technic" please post it.

P.S. I truly thank you and the other guys for all the help and your points.

[asis]

Hi,

It is possible that the PSU is used with some kind of thyristor/triac drive.
Then it's justified.
Can you unravel the cryptic phrase: "(to another mini extra PCB)"...

[2X]

Hi,

It is possible that the PSU is used with some kind of thyristor/triac drive.
Then it's justified.
Can you unravel the cryptic phrase: "(to another mini extra PCB)"...

"(to another mini extra PCB)"
I have it on my first post... I don't have the power supply anymore because I gave it back so to take more detailed photos.
The "thyristor/triac drive" can't work with isolation?

[asis]

Maybe if the triac has an optoelectronic isolation.
But, in any case, the circuit (integrator) that controls the ignition, directly or indirectly, must be tied to the AC network.