Solution Finder Get Tech Support

12W TinySW-III StackFET Design

Posted by: bott on

PowerInt,
Attached are schematics and PIXls files for a power supply I have designed. Intended to operate from 208-480VAC +/-20%, 12VDC out @ 1Amp. TinySwitchIII + StackFET.

Prototpes appear to operate correctly at lower(<240VAC) input voltages, but when input voltage increases, at a certain point, the output will begin to drop(threshold based upon load). At full-load, regulates fine up to about 300VAC input.

Are any issues seen in the design data which could explain this? Also, the UVLO does not appear to operate correctly as the supply runs fine at 125VAC when ULVO set to 177Volts.

Regards,
Brian

Files

Attachment Size
PowerSupply2012-BLO.pdf 62.66 KB
TinySwitch-III-EE19-1.pxlt3 40.32 KB

Comments

Submitted by VCastrellon on 05/09/2012

I was looking your design. What I would do is to simplify the circuit.

One suggestion I have (I am not sure if it will help ) is to remove R4 from the circuit. You will have less snuber for the DRAIN of Q1 but is ok for now.

Second thing to simplify the circuit is to remove D9 and R16 from the circuit. Let the Tinyswitch to bias itself.

Once you have simplified the circuit, run your test again and see if any improvement.

You can measure with the oscilloscope your IDRAIN and VDRAIN at pin 4 of U1. Look the behavior of the waveforms when the supply is properly regulating and see how they change as the output voltage drops.

If you can post the waveforms I may get a clue what to do next.

Thanks, I'll remove the suggested components to simplify the circuit and retest. Is it fair to assume you did not see anything wrong with the design of the flyback transformer? In previous testing, at lower input voltages, the drain voltage waveform with load looked about as one would expect(discontinuous, with a fairly regular duty cycle, but switching frequency was more like 66KHz). As the input voltage was increased the duty cycle obviously got narrower, and when the output fell the switching frequency measured at 132KHz, the drain voltage took on a look more of continuous conduction, and the pulse width became very narrow(I'm guessing the pulses were being terminated by curent limit). Mixed in with this would be the occasional 'full' pulse width and the discontinuous look of the drain voltage.

I removed the components R4, D9, and R16 as suggested. Currently the scope I am using doesn't have the ability to save waveforms that I can share, but I'll work on that. This did not fix the issue. Follows is a description of what I saw on the scope.

Regulating Correctly:
Vin = 220VAC
Vout = 12.2V
Iout = 12.2/16 = 750mA
Vdrain switching between continuous and discontinuous
Discontinuous switching freq = 66KHz
Continous switching frequency = 145KHz
Discontinuous pulse width = 3us
Continuous pulse width = 2.5us

Not regulating:
Vin = 300VAC
Vout = 10.2V
Iout = 10.2/16 = 640mA
Fully Discontinuous
Switching Freq varies from 45-65KHz
Pulse width 2.4us

The simplifications suggested did not improve the performance. I should get waveform data of U1.4 later today to post. Any other suggestions?

Submitted by VCastrellon on 05/10/2012

what I suggest is to measure the DRAIN waveform with your oscilloscope (lets say channel 1) and at the same time, with a second channel, see the waveform at node of D1. Both waveform should be in phase. Now move your second channel probe and measure anode of D3. Should be in phase too.

The other thing you can do is to measure your transformer primary inductance.

If you do not have an inductance meter, you can see the DRAIN current at pin 4 of U1 using an oscilloscope. See the sloop of the DRAIN current waveform ans using L= V*(dt/di) you can fairly approximate the primary inductance.

Attached are Vdrain(Q1.2) waveforms taken at various input voltages. 200VAC was the only point at which the output appeared to be in regulation with load. Few others showing pulse width and frequency.

Help appreciated.

Submitted by bott on 05/11/2012

Thanks, I'll measure as suggested to insure phase correctness of transformer.
I've also attached the construction print from the transformer mfg for your reference and review.

Submitted by bott on 05/14/2012

Attached are waveforms. windings appear to be in phase.

Submitted by VCastrellon on 05/14/2012

I tried to send you a private message to your email but for some reason I can not find how to send it to you.

Could you please send me an email to vcastrellon@powerint.com

Thank you
Victor