TOP245 no oscillating

Andreas -

I'm guessing that when you say the power supply isn't oscillating, you mean that it's not switching?

Without seeing your design files, I would start by disabling the more advanced features of TopSwitch like UVLO, OVP, etc. If you can, use the datasheet and put the IC into 3-terminal operation.  This should get you to a good starting point.  If you run into further issues, please let me know and please attach your PI Expert design files. 

-The Traveler

Hi The Traveler,

thanks for your reply. Yes I meant not switching.

I have no design files. I just got this broken power supply I like to get going again.

I got further and managed to wire in the scope on the 240V mains side and I do see the switching of 130 khz for a short moment. About 3ms. I can see the same signal shape on the secondary, including the feedback winding..

However it's stuck in that sort start sequence.

The C (control) pin does not go above about 5.2V. The DC on the feedback is going up a little bit but I don't think it's enough.

I shortened R6 (see the jpg above) to not limit the current, but no change.

It appears to me that I don't get enough on the feedback to push the C high enough to keep the switching going.

But all it should need is the TOP245 (new chip), the transformer (ohm resistance ok) and the opto coupler (tested) / shunt reference (tested) to make something work.

The way I understood it from the original Power Integrations app note going with that schematics, the Q3, R4 link to M pin is to use the current to control the loop, not the C pin as usual.

Maybe the problem is somewhere there. Q3 and all the R's and C's around there are tested too.

I have tossed the thing in the corner for now. Just got too frustrated with it.

But there is still a nagging in the background to find out what's going on here.

Cheers

Andreas

Q3 is used to reduce the primary current limit at light loads.

If the power supply is pulsing on/off, this is probably the auto-restart feature.  TopSwitch products will do this during a fault condition or if the control loop isn't functioning correctly.  The most common culprit is a broken feedback loop that won't sink current out of the control pin.   You might have a bad optocoupler, the IC could be damaged, feedback components might be damaged.   

-The Traveler

Thanks for quick reply.

I had been looking at that feedback loop over and over. I've replaced pretty much all in the feedback loop and still no go.

I try to understand the 'sink current out of control pin'.

The feedback voltage (positive on emitter of U2) is meant to pull up C (via R3 and C16) ? Correct?

I could leave out Q3 and R4 and shorten R6 for testing?

What's R2 for then?

I thought the feedback current is via opto coupler U2, R3, C  to S pin, and back via D6.

Cheers

A

TopSwitch products use the control pin *current* as the signal for the feedback circuitry, not the control pin voltage.  The opto feeds current into control pin (disregard my comment with sinking current from the control pin.  I've been working with another one of products a lot where this is the case).  In the 3-terminal operating mode for TopSwitch, the controller duty cycle is proportional to the control pin current.  See page 5 of the attached PDF for a graph of how duty cycle is related to the control pin current.

http://www.powerint.com/sites/default/files/product-docs/top242.250.pdf 

Here is the engineering report for the reference design you linked to from Digikey's site.  It has some thorough explanations of the control circtry, bias windings, etc.

http://www.powerint.com/sites/default/files/PDFFiles/epr32.pdf

-The Traveler

Thanks for that.

Looks like I need to pull the thing out again and have another go at it.

So to simplify the setup I should be able to get the feedback going by having just the opto coupler and a series resistor connected to C, and close to circuit via S? That would be the simplest 3 terminal mode?

Cheers

A

Ok, so I did a bit of reading on EPR/DAK 32 and the feedback and control circuitry is quite a bit different than PI's more standard method.  The fairly standard PI feedback circuitry for TopSwitch products is to control the primary side MOSFET duty cycle proportionally to the current being fed into the control pin.  This is basically our implementation of voltage mode control using a current based feedback signal as opposed to a voltage based feedback signal.

In EPR/DAK 32, the bias winding and control pin circuitry (components around Q3) implement a control scheme similar to current mode control.  A small amount of current is fed into the control pin to keep the controller operating and running at maximum duty cycle.  Any current in excess of the ~2mA required by the IC is used by Q3 to control the M pin current.  The M pin current sets the primary MOSFET current limit.  So instead of controlling the MOSFET duty cycle, this feedback circuitry controls the MOSFET current limit to control the output of the power supply.  

The engineering report has a bit more information on the feedback and control circuitry.

http://www.powerint.com/sites/default/files/PDFFiles/epr32.pdf

-The Traveler