TOP265KG design doesn't regulate
Using PIExpert I spec'd a universal PSU, 10V @ 2A nominal, using TOP265KG and the Zener feedback (no opto-isolator). Well we built it and it doesn't regulate. With no load, the output voltage is 10V well enough, but the output declines steadily to 6.2V with a load of 2A.
We hacked up a version with the optoisolator and related (myriad!) components and it doesn't show this problem. So what's going on? I really want to avoid the opto - they are short lived and expensive.
I'll attach the design (pdf report as well as PIE file)....
and yes, double and triple checked the prototype is exactly like the schematic, and a good layout....
Comments
This design is exactly output from PI Expert. So you're saying the software has given me a faulty design?
Hi,
The Primary side feedback is only suitable for low power applications and will suffer from cross-regulation effects causing lower accuracy for voltage regulation. The PI EXPERT Schematic is correct. I misunderstood your requirement as being non-isolated design without opto.
If you want +/- 5%, I would recomment using our latest Primary side Regulation controller LNKSW-HP. http://www.powerint.com/en/products/linkswitch-family/linkswitch-hp
Sorry about the confusion.
rgds..
If the schematic is correct, then please explain how the voltage can drop from 10V to 6.2V with increasing load 0-2A.
40%? that's no regulation.
In fact, I did not select that IC, the PI Expert software suggested it for me. It seems this software is not trustworthy.
Unless you have some specific recommendations to easily repair the current design, there is no other conclusion, is there?
If I need to find another chip, then I should consider other manufacturers too... are you sure you want to recommend that I completely change the design?
Hi,
What is the measured leakage inductance of your transformer? The reason the voltage drops is because your primary feedback is the bias winding not the secondary output. Your controller does not see the voltage drop on the secondary and hence can't react to its changes.
What Switching frequency do you measure with and without loading the 10V otuput?
Regards
Hi Scotty, sorry for the delay... internet problems.
The frequency doesn't seem to change.
The leakage is about 15-16uH at 100KHz, a bit more than the desired 12.
regards,
Marc
hi Scott,
ignore my other statement, it was a miscommunication with the testing engineer... i looked at the frequency myself and it's about 26KHz with no load, up to 132KHz with 2A load.
Some obvious conclusions to be drawn from this thread:
1. Do not use the PI Expert Software. It's unreliable, it's design output is not to be trusted.
2. Don't expect timely support from PI via this board.
Fortunately for PI, they have a few good FAEs in China. If there are any in the NY Metro area, I haven't been able to find them.
Hi,
You are regulating the bias winding and secondary winding is cross regulated. So, as you load the secondary the secondary will start to collapse.
You must regulate the secondary winding (output) not the bias winding output.
You need to put the correct valued zener on the secondary winding and also use the output as bias for the TOP265KG.
rgds..