AVR BC100 minimum pulse time is 1us !?

Go To Last Post
6 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I cannot get the output voltage low enough.
When the PWM output pules from ATtiny is 200ns, the power FET is on for 1us. This results in a too high voltage (appx 10V) for my need (8.5V).

I measured input to driver at TP410 (CH1), output from driver at TP411 (CH2) and Power switch at TP408 (CH3). I will try to attach the scope traces here.

Attachment(s): 

Einar Sjaavik

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think you have a classic "gate charge" problem.

To switch a power FET fast, you need to be able to source or sink a LOT of gate current (often, 1Amp). A processor port pin will simply NOT do that. Not AVR, not PIC, not FreeScale, not anyone.

If you need to do that, then you need a gate driver. They are out there, though I don't have any to recommend. Probably TI, maybe ON or Fairchild or NXP.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

There is a gate driver on the BC100.
It was thoughtless of me not to include the schematic.
Here is a crop showing the relevant measuring points.
BUCK_PWM comes from the ATtiny 861. C425 is also connected to V_BUCK_IN_C. After the inductor there's 40uF of caps.
J404 selects the 20V/1A buck switch. In my case it's shorted 2-3.

Regards
Einar Sjaavik

Attachment(s): 

Einar Sjaavik

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So it looks like the gate driver is to slow and to weak. The gate drivers output is allready more than 500 ns and the output pulse is even longer.

Anyway, one should avoid such short pulses. Instead one should use some pulse skipping to reduce the frequency if the pulse length would go below about 1-2 us. This gives better efficiency and less EMI.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree. While the gate drive is much better than a port pin, it still does not have the speed you are expecting.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks guys.

I was aware of the gate charge problem. I just thought the circuit shown would be capable of that. But reading the big writing from Atmel, this is a charging design. I tried to coax it into a service it was not designed for. Thanks for the input, I'll stop wasting time on that.

I'll probably check out Kleinstein's suggestion. Using that strategy, it should be possible to get down to any voltage.

Einar Sjaavik