Lynx RS4-2000 switch mode PSU does not start

[dicky96]

Hi guys
This PSU is actually in an amplifier, but it is the PSU itself that has a problem

I have spent quite a lot of time trying to diagnose this and the more I get into it the more puzzled I become

It has a Standby, or Startup supply (15V) generated by a TOP245GN and this is working

The main SMPS is controlled by a SG3525A and it does not start up

According to the datasheet
https://pdf1.alldatasheet.com/datash...A/SG3525A.html

This SG3525A will shut down if pin 10 is high, or pin 9 or pin 8 are low. Further it says that pin 9 & 8 have an internal pull up and can be taken low (disable) by drawing as little as 100uA externally.

OK so I have pin 10 low, Pin 9 high (this pad does not go anywhere) and pin 8 low - should be high

First I changed the capacitor from pin 8 to ground in case it was leaky (just tagged a 4.7uF electrolytic on there as the example circuit showed 5uF) but it made no difference

Then I changed the SG3525A - no difference

I then reverse engineered the circuit connecting to pin 8 (a couple of G1 SMD transistors and the opto-isolator)

From what I worked out (see attached pic) pin 8 is held low because the transistor connected to it is turned on (base connects to the 15V Standby via a resistor and has 0.47V which seems a bit low but I guess it is enough).

There is also the opto-isolator connecting here to pin 8 but the LED side has no voltage supply so it must be off.

The only way I can see to turn this transistor off is to turn the other transistor on, this transistor connects via 3x 150K resistors and a diode to 240V AC (not the +ve of the bridge rectifier - one lead of the actual AC cable!)

However it does not seem to have enough base voltage (0.35V) to turn on.

There is a resistor divider 150K-150K-150K-1K5 that supplies the base, I removed the transistor and checked the divider. I removed the 150K resistors they all check OK, I measured the voltages across each resistor and applying ohms law this should indeed supply 0.34V to the base (see pic).

** Note I originally had 0.26V on the transistor base, and tried strapping a 10uF electrolytic across the capacitor in parallel with the 1K5 resistor and the base voltage went up to 0.34V

The 1K5 to hot ground reads correct. I thought the capacitor across the 1K5 may be faulty so I removed it. It wasn't leaky but only measured 0.1uF so I replaced it with a 10uF. That gave me 0.34V on the transistor base.

So now I just can't figure out why the circuit does not work.

My best guess of the intentional operation of this circuit is that for the main PSU to power up it needs:

1. 15V standby supply
2. Sufficient voltage on the AC lead (is it intended detect a brown out as a fault condition and shut down the main supply?)
3. The opto-isolator is off - I assume this is driven by some error detection circuit in the amplifier section.

Anyway, could anyone confirm this is the intended operation of the circuit, and explain why it doesn't work or what to check next? Because it is frying my brain LOL!

Rich

[stj]

Re: Lynx RS4-2000 switch mode PSU does not start

presuming you draw that correctly, it makes no sense

[dicky96]

Re: Lynx RS4-2000 switch mode PSU does not start

I could double check What do you think I may have missed that would make sense of it?

Or what part in particular does not make sense?

[stj]

Re: Lynx RS4-2000 switch mode PSU does not start

the 50hz clock pulse from the chain of resistors going into the switcher i.c.

[dicky96]

Re: Lynx RS4-2000 switch mode PSU does not start

It goes to one of the transistors rather than to the IC

I can 100% confirm I have direct continuity from the top of this resistor chain via a diode to one of the AC input pins on the bridge rectifier and to one wire in the mains input cable

[CapLeaker]

Re: Lynx RS4-2000 switch mode PSU does not start

why is that through hole resistor (bottom left) so tarnished brown? Maybe check the smt caps if one is shorted? If all fails, inject power and see what is getting hot.

[dicky96]

Re: Lynx RS4-2000 switch mode PSU does not start

@CapLeaker
That through hole resistor is an inductor - it connects to the TOP245GN and generates the 15V acting kinda like a buck converter as best I can tell - the TOP245GN does not drive a transformer like you would expect - just this inductor.

As the 15V supplies SG3525A and associated circuitry shown in my drawn schematic, it is all on the HV side of the power supply so I guess isolation was not important

@stj
The top one of the two large SMD diodes in the photo, just below where it says 'PAD3' and '120' is the one that connects to the AC. The left hand pin of that diode marked H28 U1J connects directly to the third pin up from the bottom of the bridge rectifier. The bottom pin of the bridge is + as I marked on the PCB and the top pin is hot ground. The middle two are AC in

The other side of the diode that connects to the AC line may look like it goes to the capacitor marked 47 16V, but it doesn't. It goes to a via which is just under the edge of the black plastic base of the 47 16V capacitor, and this then goes back through another via to the top end of the 3x 150K resistor chain. I have checked this on my meter and also visibly followed the track through the two vias to the resistor chain as I couldn't figure out where the resistors were going at first, though it was obvious from the values they must go to the high voltage supply. At first I expected +ve of the bridge, but no.

Using ohms law I can calculate that if I had about 120V on the top end of the resistor chain I will have 0.5V on the base of the transistor that needs to turn on. See the little diagram to the right of my photo - this is where I calculated the voltages. Originally I had 78V at the top of the resistor chain and 0.26V on the transistor base (ohms law agrees).

You can check my calculation that shows 120V in would give 0.5V out (if base current was negligible)

Now the odd thing is, when I replaced the capacitor in parallel with the 1.5K resistor - which does read 1.5K by the way - with a 10uF, the voltage on the transistor base increased to 0.35V and the other end of the 3x 150K changed to 103.7V.

Quite how changing this capacitor this could change the voltage at the other end of the resistor chain I can't quite figure out - unless the whole circuit is very high impedance? But it is supplied by a rectifier directly from mains AC so it would be a low impedance source. The diode itself checks OK on my meter in diode range but could it be faulty somehow?

Also I can't figure out how to calculate what voltage I should see from the diode to hot ground. So one end of the diode is AC mains and it is a half wave rectifier, but my meter reference point is other end of the resistor chain 3x150k + 1.5k which is hot ground which is floating at half mains potential?

So really what voltage should I expect to see across the resistor chain? Is it 103.7V?
I can do ohms law as you can see but I don't understand how to make this calculation.

As I have 220-240V mains I can't quite figure out why I only have 103.7V here and not more.