power supply with two outputs.
I developed a power supply with two outputs (12V) and (36V) using TOP270EG.
When I apply a load only at 12V, the PWM acts to keep the voltage constant, but this causes a voltage rise at the 36V output.
When I apply a load only at the 36V output, the voltage drops until it reaches 12V, the feedback responds from there and raises the voltage, which drops again and the output voltage oscillates.
How can I solve this?
评论
Hi Tadeu,
The main output will always have the best regulation. There are techniques to improve it, you may check this document: AN-22 - Designing Multiple Output Power Supplies with TOPSwitch | Power Integrations
Regards, Gavin
There was a specific reason why I asked that question. If the two outputs have separate ground systems, regulation will be much more difficult on the output that does not have the regulation loop, and some juggling of resistive preloads may be necessary. Also, it sounds like the loop is unstable on the output that is regulated, and that issue may need addressing. If the two output have separate returns, the output of one will be affected by the coupling between outputs, which may not be very good. This may explain why the 36V output oscillates when you load it. To help you better, I will need to see a schematic, which will tell me a lot more about your supply and what precisely you are trying to accomplish.
Attached is the power supply in question where the secondary windings are stacked.
There was an improvement with the repolarization of the PS2505, but it was not satisfactory.
I disassembled a transformer to check the assembly and the winding layers seem to be well distributed.
I am starting the process of repolarizing the reference voltage using the two outputs as shown in figure 4 of AN22
| Attachment | 大小 |
|---|---|
| power supply 12V_36V.pdf (2.36 MB) | 2.36 MB |
My first observation is that there is no preload on the 36V output, so that it is free to go overvoltage with moderate loading on the 12V output. I would suggest that you share feedback between the 12V output and the 36V output. This will make regulation on both outputs a little worse, but possibly also solve your issue with the 36V output, since it will now contribute to the feedback.
OK, here's my take on the situation - R17 needs about 221 uA passing through it to match the 2.495 V Vref of the TL431. First off, change R16 to 49.9k. That will give you about 190 uA through R17. Next, string a 1.07M resistor from the 36V output to the bottom of R17. This will give you close to the remaining 31 uA. Regulation will not be perfect, but at least you will be sensing the voltage on both outputs.
OK, here's my take on the situation - R17 needs about 221 uA passing through it to match the 2.495 V Vref of the TL431. First off, change R16 to 49.9k. That will give you about 190 uA through R17. Next, string a 1.07M resistor from the 36V output to the bottom of R17. This will give you close to the remaining 31 uA. Regulation will not be perfect, but at least you will be sensing the voltage on both outputs.
An R16 value of 56.2K and a resistor from the 36V output to the bottom of R16 of 649k would give the 36V output more say regarding the regulation.
Do the two outputs have separate grounds? Which output has the regulation loop?