basic SMPS circuit repair

Thread Starter

nick9one1

Joined Mar 22, 2016
12
I'm trying to repair an SMPS and have narrowed down the issue to a particular circuit on the board.

I believe this is what is called the startup circuit? - The board has a large bridge rectifier, relay caps etc. Then there is this smaller circuit with its own (much smaller) bridge rectifier, caps and transformer. it seems to control a relay with would then feed AC power to the large bridge rectifier.

There is 340v DC coming from the bridge rectifier which I can trace all the way to the transformer. Caps have all been replaced, resistors and diodes all test ok.

Is it common for the OB2358 to go bad?

Also, If it's not possible for me to repair the circuit, I'm considering removing the transformer and replacing it with a small standalone smps. I believe this should work? If I go down this route, I need to know the DC output of the transformer. Is there any wat to work this out?

2.png
 

bertus

Joined Apr 5, 2008
22,270
Hello,

What is the VDD voltage on the OB?
It looks like the OB is not switching.

There is an inconsistency in the text and schematic.
In the text you mention the OB2358, as the schematic says OB2353.
Wich is it?

Bertus
 

Thread Starter

nick9one1

Joined Mar 22, 2016
12
Hello,

What is the VDD voltage on the OB?
It looks like the OB is not switching.

There is an inconsistency in the text and schematic.
In the text you mention the OB2358, as the schematic says OB2353.
Wich is it?

Bertus
I apologise, it's the OB2358. I assumed (perhaps wrongly) that the circuits would be the same.

Voltage on the VDD pin is about 15 mV, 1.3mV on the VDDG pin, and 1.5mV on the FB pin.

Untitled.png
 
Last edited:

Thread Starter

nick9one1

Joined Mar 22, 2016
12
ok, thanks. I'll order a replacement!

Datasheet for smps ic...
https://www.google.com/url?sa=t&sou...FjAAegQIARAB&usg=AOvVaw1m8azbzXMsVfed-IaeVf-_

You can't just change the transformer, it's all been designed around it,,,start by checking the Voltages on the chip pins 1,2 as per datasheet,, pin 3 is the feedback pin from the opto-coupler.

I didn't mean I would just replace just the transformer from the circuit, but replace the whole circuit with a small smps that outputs the same voltage. It looks like all this circuit is doing is feeding the main relay and the LCD screen (relay is 12v but not sure about screen).
So I would bypass this whole circuit and where the transformer currently outputs 12v, I would feed with an external 12v.
 

LesJones

Joined Jan 8, 2017
4,174
I have seen a few cases in the past of the decoupling capacitor on the VDD pin becoming leaky or high internal resistance (ESR) and this prevents the circuit from starting up. I suggest that you try replacing this capacitor first. If it is leaky it could explain the very low voltage reading on the Vdd pin. Also check that the 1M5 resistor has not gone high in value or open circuit.

Les.
 

Alec_t

Joined Sep 17, 2013
14,280
Even if the OB isn't switching, I think the voltage divider on the Vdd pin should be giving >200mV. As Les says, check the 1M5 resistor.
 

Thread Starter

nick9one1

Joined Mar 22, 2016
12
I hadn't realised there had been more replies since I ordered the OB - thanks for the additional suggestions.

I replaced the resistor and ob and now it's working!
It's a power supply for a Reebok treadmill, previously completely dead but now powers up.
Bad news it looks like theres an issue on the other side of the SMPS (the bit that powers the roller) Doh!
 

DarthVolta

Joined Jan 27, 2015
521
1 thing is for sure, the MOSFET inside the OB chip works hard, and it's no doubt on the cheaper side, so I'm sure lot's can go wrong with the chip .
 

SLK001

Joined Nov 29, 2011
1,549
The curious thing about this type of SMPS, is that almost EVERY part must be functional for the supply to even turn on. The Schottky diode on the secondary side is the most stressed device. If it is bad, then the regulator chip (the 431 device) isn't powered and the feedback (the optoisolator) is railed. The opto circuitry is a weak link in this chain, also - check the electrolytic cap at its output. If it fails, the driver IC may go into shutdown mode and make the driver appear to be defective.

Believe it or not, but the switching MOSFET inside the driver chip isn't stressed all that much. It operates as a switch, either full on, or full off. It will be a device with a very low Rds-ON value.
 

Thread Starter

nick9one1

Joined Mar 22, 2016
12
I've nearly had a breakthrough. After replacing the main bridge rectifer it all worked. Powered up motor turned etc.

So I reattached the heatsink, and mounted it back into the treadmill. Powered up, then when I selected a mode it blew a fuse.

It looks like the transistor is to blame. I can see marks on the heatsink and transistor where there was a short.
 

Thread Starter

nick9one1

Joined Mar 22, 2016
12
Should the transistor collector have continuity with the metal back of the transistor? it's a G40N60UFD

I've just remembered that originally there was one of those thermal pads between the transistor and heatsink. I just threw it away when I was took it apart (paste is much better). Now im wondering if it had a dual purpose and was also an electrical insulator?

 

Thread Starter

nick9one1

Joined Mar 22, 2016
12
I believe I caused the failure with my transistor incompetence! After a little googling it turns out transistor insulators are a thing.
 

SLK001

Joined Nov 29, 2011
1,549
The phototransistor in the optoisolator receives its bias from internal circuitry in the IC. With the Schottky diode on the output bad, no voltage is powering the reference IC, the OB431, so nothing is powering the LED in the opto, which means that the phototransistor is in the OFF condition (zero conduction). That means that the bias voltage on the phototransistor isn't being pulled down to the normal operating range, but is instead still at the rail.
 

atferrari

Joined Jan 6, 2004
4,764
The phototransistor in the optoisolator receives its bias from internal circuitry in the IC. With the Schottky diode on the output bad, no voltage is powering the reference IC, the OB431, so nothing is powering the LED in the opto, which means that the phototransistor is in the OFF condition (zero conduction). That means that the bias voltage on the phototransistor isn't being pulled down to the normal operating range, but is instead still at the rail.
Quite clear now.
Thanks.
 
Top