I'm not familiar with that chip, but the first thing that comes to mind is =
to
put a scope on the RC components and look at the freq (should be a sawtooth
most likely). If it varies in freq, and +V doesn't change, it should be the=
 R
or the C. That's probably what I'd look at first, even if only out of
curiosity... ;)  In the process, I'd also verify the V driving the RC
components for shifting.


-Skip

gardenyu wrote:
> =

>    I built a power-factor-corrected battery charger using Fairchild's FAN=
7529 PFC chip and Onsemi's UC2843 constant frequency PWM chip. Basically it=
's a boost converter driven by FAN7529 followed by 2843 chip driven flyback.
> =

>    It worked fine on my bench for hours. But the customer made a test on =
a very deeply discharged battery set (somehow discharged to 8V for 24V batt=
eries, near dead ones to me) and they said something was not ideal. =

> =

>     When I re-test the returned unit, there was something weird. It worke=
d fine for 2 minutes or so, then the 2843 chip gate drive frequency ramped =
up from 58k Hz to about 68k very fast in seconds ( remember it's a constant=
 frequency chip programmed by RC components outside), thus giving more load=
 power. Its duty cycle was not changed, however, thus only the load current=
 increased. The input power increases linearly thus the efficiency is simil=
ar. When restarted, it repeats the whole process. I can't allow this, since=
 overheating might occur.
> =

>     I probed the oscillator pin of that chip and apparently the oscillato=
r frequency (should be constant) varies along with the changing gate drive.=
 Also the power factor stage always works.
> =

>      Any suggestion on what could be wrong before I replace that chip?
> =

> _________________________________________________________________
> Live Search=CA=D3=C6=B5=CB=D1=CB=F7=A3=AC=BF=EC=CB=D9=BC=EC=CB=F7=CA=D3=
=C6=B5=B5=C4=C0=FB=C6=F7=A3=A1
> http://www.live.com/?scope=3Dvideo
> =

-- =

http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist