TNY280 15W 24VDC supply output Ripple

Hi,

Can you attach the drain voltage and current waveforms?
That output ripple may be the result of using a zener clamp.

One thing that stood out is that the line frequency component of your ripple voltage is very high. This is possibly caused by a slow control loop. You could try reducing C6 to 100nF + 3.k in series, and reducing R9 and R10 to 10k and 1.15k, respectively.. Reducing the value of R6 may also be an option.

Was this schematic generated by PI Expert?

My mistake - I mis-read the value of C6. 470pF is too low, and can cause problems that look a lot like classic loop oscillation. Increasing the value of C6 to 100nF should slow the TL431 down sufficiently for it to stay of of trouble. Ideally, the TL431 loop around the control pin should be sufficiently slow such that the TL431 looks like a fixed zener at high frequency.. To start with, I would leave the R9 and R10 values alone..

Thanks for the tips everyone. I'll be back in the lab next week to try them out. To answer your question, yeah the PI design schematic gave me the 470pf value for C6. I'll let you know what I find next week. Thanks again.

OK, replacing the .47n with a .1uf greatly reduced the ripple. It was down to ~20mv now, which is pretty good @ 24VDC out.

I have two issues now:

1. When I am at full load ~15W / 120VAC in, the transformer ringing on the output grows at least 6x larger. from when its loaded around the midpoint ~8W.

I've attached screenshots of each.

Things I've tried:

I tried a RCD clamp., which did not result in any noticeable change. Screenshot of the clamp below.

I've changed the Zener clamp to different value Zeners. 100v / 150v / 180v. No Luck

I've changed the clamp diode to a slower 1N4007 type.

I've added more capacitance on the input.

I've added more capacitance on the output post filter.

2. The transformer is making an audible noise. I see that may be because of a flux density that's too high.

I removed a ceramic output cap to make sure it wasn't the noise source.

The first tips were very helpful. If you have any others to help on these issues, that would be greatly appreciated. I'll continue to read the app notes and troubleshoot.

Thanks.

Hi,
Is the attached "output ringing" the 24V output ripple?
For audible noise, you are right that the flux density may be too high. Can you send the transformer spreadsheet?
Varnishing the transformer should also help with the audible noise.

It's the high frequency ringing from the transformer when the drain voltage drops. When viewing the 24VDC output, it grows greatly when almost at max load.

I put a .001u across the output of the transformer and it greatly reduced the ringing on the output.

I can still hear the transformer though. The supplier told me the its flux desnity is calculated to 1768 gauss. .

After evaluating further, at 15W, my supply is switching at ~70-80kHz, not 132kHz.

The transformer Flux density Calc was based off of 132kHz.

The value of the flux density at 70kHz is a little over 3000 Gauss. This may be the cause of the humming noise.

Is this switching below 132khZ because it's in a reduced current mode? Or because we're running it far below the I-limit min?

Hi,
Changing the current mode will only affect the maximum Ids limit.
A reduced switching frequency is most likely the result of operating the device at a lower output power than the circuit is designed for. Please see TinySwitch-III Family datasheet page 5.
But based on your Vds waveform at 15W, the device seem to be switching as intended with a maximum frequency of around 132kHz.

Also, can you point out in the schematic where you are probing for Ch2 "OutputRinging"?

Thanks. I'm probing the output of the filter at J1.

Do you see any issue with putting a 1nf Cap across the output of the transformer? That greatly reduced the amplitude of the ringing.

I did see the note on page 5, but I incorrectly assumed it was referring to designed loading max, not TNY loading.

Hi,
You could put a capacitor there but the efficiency would probably decrease and the temperature of TNY280 might rise.