From: Chris Carlen
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1) Gecko/20020826
X-Accept-Language: en-us, en
Subject: Re: 87% All that vector calculus paid off
Date: Mon, 23 Dec 2002 04:05:35 GMT
NNTP-Posting-Date: Sun, 22 Dec 2002 20:05:35 PST
Winfield Hill wrote:
>>Chances are there is something wrong with your measuring technique. ...
> An easy trap to fall into, calibrated setups are necessary.
Yeah, I measured it in a more proper manner, and what I saw before
really wasn't happening. You can read about it in my response to Fritz.
But now I have discovered something really wierd:
The ringing of the drain circuit after the diode turns off, is clamped
by the FET body diode, at least this is how I understand it at this
level of my experience with this stuff. So we see a series of decaying
ringing oscillations, but the negative swings are clipped off.
You can see an example here:
Scroll down to the links to my scope traces. Those traces are not from
the exact same setup I am using now, which has much better gate drive,
but it is close enough to follow what I am about to describe.
Now that I have a PWM chip, I can vary the duty to my heart's content,
without changing the frequency. What I have discovered is very
interesting, and it brings up serious questions about what will happen
when I try to close the regulator loop around this thing.
When I adjust the duty%, and let's say there are three ring spikes
present, and say I am increasing the duty% by hand. I don't see the
output voltage changing proportionally to the duty%. Instead, the
output voltage changes slightly, then all of the sudden it lurches
upward as one of the ring spikes disappears.
Thus a plot of the output voltage vs. duty cycle would look like an
upward staircase, with the steps coincident with the changes in the
number of ringing spikes.
Ugh. This is a wrench in my works!
Christopher R. Carlen
Suse 8.1 Linux 2.4.19