Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

I didn't say it was worthless...
You can't really test things correctly in circuit, unless you know the circuit or have a schematic for it. Right now you don't know what you are measuring. Yes, a fishy testing component, you have to take out and test it out of circuit.
A SMD diode with a marking of T4 (the bar marks the cathode) is a stinking ole 1N4148 signal diode. Very common.

To me I am not convinced that the 1.1 volt difference between them two optocoupler pins is enough to turn this thing successfully on and stay on. Typical this opto has a FV of 1.2V. You got 1.1V and I'd say you should get way more that but under 5V in between these.

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

that diode tests good out of circuit, but with the diode removed should there be 10ohms from the base of those transistors to the PWM_OUT pin? - 10ohms seems quite low for a path where one side is connect to two transistors, and the other side is connected to two mosfets(?) that are inside the chip.

check all them opto isolators. Once when the psu works ok and then again when it shuts down.

CM6800 PWM Feedback opto: [works ok] 1.135V [shuts down] 0.000V
Middle opto: [works ok] 1.33V [shuts down] 1.091V (briefly), then 0.136V
Opto that's closest to the transformers: [works ok] 1.046V [shuts down] 1.046V

All three optos are the same part number

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

10 to 100 Ohms should be o.k. depending on components used.

These opto's... always compare the voltages both pins on the secondary to secondary GND, PSU turned on. Then check the opto in diode more. That will tell you where the problem lies. Could be the supervisor or the opto or something else. All depends on how old the PSU is and ICs used.

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

yeah, after removing the diode again, and trying to fathom why the 10ohms was still there, despite no continuity through the two transistors, i noticed that there was a 10K resistor in parallel with the diode, so that makes sense to me now, and that's not the problem.

i again measured the voltages, each opto leg to secondary ground, in the same order as before
3.29 4.34 = 1.05V
0.939 0.2 = 1.139V
6.97 8.06 = 1.09V

in diode mode, should i be checking these optos out of circuit?

the PSU ran for about 15mins this time (but shut down about a minute after that, after plugging in a hard drive) - maybe the ambient temp plays a part, it is now about 20 deg C (cooler than before)
i noticed that the 3.3V was only 3.0V, and thought that that rail might have to be loaded more, so i restarted it with a 4.7V 0.75A bulb - the PSU shut down as the bulb lit up

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

The lightbulb in series trick is for repairing things on the primary side. If the primary works o.k, which it does, remove it. When you load the psu the lamp trick will limit the current going to the psu.

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

seems that there is a misunderstanding/confusion...
i'm not using my mains dim bulb tester
the little 4.7v bulb was connected across the 3v rail - i wouldn't do something as stupid as connecting it to mains, as that would blow up the bulb.

how will checking the opto in diode mode tell me where the problem lies?
the optos appear to be working OK, so do you mean test them in circuit, in order to identify anomalies?

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

Post 44, second opto… is there a typo or something?

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

yep, "or something"
i copied my calculation instead of the measured values
same voltage drop as before though

3.29 4.34 = 1.05V
0.061 1.2 = 1.139V
6.97 8.06 = 1.09V

[stj]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

what component tester do you have?
there is an adapter for testing optocouplers on some

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

just a multimeter,
and a 328 transistor tester, doesn't have a optocoupler tester though,
also have an old CRT oscilloscope if that's useful

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

I think there are vids on YouTube on how to test an opto.

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

yeah, i watched lots of them already
might as well do it properly, since we want to know the minimum working voltage
i'll order parts to make a tester
and post results, probably in a few weeks

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

Found that i already had the parts to make a tester

Using that tester, all three optocouplers are all working with 1.11V voltage drop across them , and with same intensity of LED glow

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

So one is on, one is right at the threshold and the other one is off. Right?

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

No, don't think so
In post #53 i accidentally measured the live voltage drop with it on diode tester mode (so that just created a bit more supply voltage)
So i switched the meter to DC volts and checked the voltage drop:
With a 9V battery as the supply, i got 1.071V across the optocoupler, and the LED was lit,
and after replacing the 10Kohm resistor with 17Kohms, i got 1.046V (of course the LED got dimmer as i added the extra resistance)
I did this with all three optos, they all measured 1.046V, and at 1.046V all the LEDs lit up

So, in summary, since all of the optocouplers are on at a 1.046V voltage drop, which is lower than the three measurements taken from the turned on PSU, i think that the optocouplers are OK
Does that sound right?

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

Where did you read to use 9V as supply? This little test jig sez 5V which makes sense to me.
I would redo that test with one opto.

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

I read it on another webpage with a similar circuit - in the parts list it had 9V battery, but in the description it said 4V - so it seemed to me that it wasn't that much of an issue, since current was being limited (in the circuit that i'm using) by a 10Kohm resistor
Anyway, i tried it with 5V rail of ATX PSU, and i got 1.032V on all three optocouplers
and after adding a 3.7K resistor in series with the 10K resistor, the voltages decreased to 1.022V
In both instances the LED lit up

[socketa]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

I took another look at this, and noticed that the the 3.3V rail goes down to 3.04V not long after powering up the PSU
The P229 Supervisor datasheet says 2.4V max for Under voltage threshold.
So it doesn't seem like it this supervisor chip is the cause for the PSU shutting down, since the voltage on the 3.3V monitoring pin is about 3.02V. Is that correct?

[Per Hansson]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

3.3v, 5v and 12v all have a tolerance of +/- 5% according to ATX specifications.
You should be using a decent load when testing the power supply, it could be as simple as shutting down for not enough load causing under or over voltage protection.

I also read through this entire thread again and one thing is strange:
The APFC boosting of voltage from 340VDC to 390VDC should only happen when you start the main power supply.
It sounded in several posts like you had 390VDC with the power supply in standby mode (only 5VSB present) and that is wrong.

[CapLeaker]

Re: Zalman ZM500-GV stops after 40s and 5VSB drops to 2.5V

I am starting to wonder if something is wrong with the TNY2xx IC.