PMP8740: PMP8740 Reference Design PFC Status I/O Issue "not working"

Hi Meet,

If you read a "low" condition on PG_PFC, it means U9 is conducting continuously, which is really weird.

If, instead, you read a continuous 5V on PG_PFC net, then something is not working.

Let's suppose you have always zero on this net.

It means there is a continuous current flowing through U9, right? This can happen if U9 is damaged, or there is a non correct resistor value for R43, R40, R37 or R46.

I suggest to do the following to test this circuit and double check both thresholds.

1) Disconnect R36 from "+390V" net.

2) Supply the disconnected terminal of R36 with variable DC power supply, ranging zero to 400V.

3) Supply the net VCC_Primary with 12V

4) Supply the 5V for pullup resistor PG_PFC

5) Increase the 0-400V external PSU from zero to the two threshold voltages.

You should see PG_PFC switching from 5V down to zero.

Please consider that there might be a bit different threshold voltage compared to what is described on the schematic: in fact, the LMV431 has 1.24V threshold voltage;

Considering that when the "+390V" net" voltage is lower than 385V, the hysteresis circuit driven by Q7 is not active, therefore the high-threshold voltage should be calculated considering 1.24V on the LMV431, and the voltage divider R36 + R37 + R40 together with R43; this voltage divider provides 1.24V on the reference of LMV431 when  "+390V" net (or the external DC source) reaches 388.7V (sorry, I wrote > 385V, but it should be actually 388.7V)

Please let me know which thresholds you are measuring.

Best regards,

Roberto

Dear Roberto,

I applied a 338VDC and 391V with PFC ON in the input of PFC Monitor Circuit, which exceeds the threshold of 388.7V. Despite this, I am consistently obtaining a Low Logic signal. Additionally, the reference voltage across the LMV431 is measured at approximately 0.6V also when I turn off the 12V it exceeds 0.9-1V, rather than the expected 1.24V.

I have verified that the components used match those specified in the reference design. Specifically, all resistors have a tolerance of 1% and the capacitor has a tolerance of 10%. Measurements taken with a digital multimeter confirm the presence of 5V, 12V, 338VDC, and 391VDC at the respective points.

Could you please provide guidance on why this discrepancy is occurring? Below is the circuit diagram including the exact part numbers of all components for your reference.

Thank you for your assistance.

Best Regards,
Meet Pandya

Dear Meet,

It's actually strange that you get 0.6V on the reference pin of LMV431; I mean, if you have 391V on top terminal of R132, you supposed to have ~ 1.247V on reference of LMV431. Maybe the LMV part is damaged or something similar.

Anyway, if the LMV431 is working fine, there might be a situation where the CTR of U12 is very high (600, worst case) that the worst case minimum cathode current of U12 is transferred through the optocoupler. In order to solve this possible case, can you please add a 10 kOhm resistor in parallel to the LED emitter of U12 (therefore between pins 1 and 2 of U12)? This way, at least 120 uA are flowing through U14 but not through U12. Then, when Vref of U14 is > 1.24V, Q10 will be activated and U14 in saturation: this way ~ 1mA will flow though the optocoupler LED, and the output will be low.

Another difference from my circuit is that fact that the pull-up resistor R140 is 10 kOhm, and pull-up voltage is 3.45V, while in your case R140 is 47 kOhm and pull-up voltage is 5V. In this regard, can you please do one test and reduce R140 down to 10 kOhm...just let me know if that helps.

Thanks,

Roberto