Dips in power supply voltage

Posted by: offurm on 09/27/2017

I am using a TNY290 in a DC-DC converter flyback application. I am seeing a dip in the output voltage which appears to be caused by a spike in either the EN/UV pin or feedback via TL431. Any idea to figure out which is causing the issue?

Comments

Hi.

Thanks for your inquiry and using TNY290 in your application.
May ask for the test condition where you are seeing the dip? Can you probe the EN/UV and the BP/M pin during this dip? Appreciate if you can send the schematic.

Thanks,
PI-Picard

Hi,
See attached.

Attachment	Size
EN-UV.pdf (87.37 KB)	87.37 KB
supply.jpg (215.2 KB)	215.2 KB

Hi,

Thank you for the attachments.

I have the following questions.

1) Which output rail were you seeing the dip? What is the frequency of this dip occurring?
2) Is it happening during startup only or periodically at normal operation?
3) Are you seeing this dip as well across C64?
4) Try unloading the outputs one at a time.If you lower down the load current will the dip disappear?
5) Is the input to your power supply a controlled voltage supply? Can you try increasing the input capacitor?

Regards,
PI-Picard

Hi,
1) I am seeing the dip in the 20V/up to 4.5W rail
2) It happens periodically during operation.
3) C64 is connected in parellel with a bank of 4 560uF caps, which are fed by a single phase full wave bridge rectifier, the rectifier is fed by the AC mains. V+ is the + side of the bank of caps, HVGND is the - side of the bank of caps. I see some ripple across C64 (about 20V ripple)
4) The application is an inverter. When the inverter output power is low/off then the ripple is not present.
5) I did try increasing C64, the dip was still present.

Thanks!

Hi,

1) If the dip is happening during the switching transition of the C64 ripple, the problem could be due to the input. Is the ripple (undershoot) low enough to cause the dip? I would assume that if the inverter output power is low/off there is no dip consequently because there is no output voltage?

2) You mentioned the dip is periodic, what is the frequency of the dip?

3) Try increasing the C64 until you see a low ripple, put in parallel some low ESR electrolytic capacitors. Will the dip disappear?

4) You are seeing the dip in the 20V/up to 4.5W rail, how about the +20V/ISO is there also a DIP?

5) Can you probe the EN/UV and the BP/M pin? What do you observe during the DIP?

Regards,
PI-Picard