AVR-Chat

Kathy Quinlan wrote:
> Russell Shaw wrote:
> 
>>Kathy Quinlan wrote:
>>
...
>>>
>>>At least it does start, and sort of works, the output is a bit low, but I used 
>>>just 1n4007 diodes instead of shotkey for the rectifier and the ones in the 
>>>circuits around the primary (they were specked at controlled avalanche ones 
>>>(1n5062 IIRC) But I can not find any in stock and no where is really open till 
>>>Tuesday lol.
> 
> I have uploaded the design, PSU.pdf.
> 
> It should work ok, it is from the appnotes lol

Assuming the app-note worked ok for them (AN-1272), it's either a problem
with pcb ground layout, or a transformer problem such as saturation or
excessive leakage inductance. All those diodes *must* be faster recovery
than 1N4007. You can tell if the transformer is saturating by looking
at the waveform across R13. A ferrite core *must* be gapped for that
type of converter.