[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Terry's DRSSTC actually hooked to a coil now >:-))



Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi Bob,

At 09:00 AM 2/1/2005, you wrote:
Hi Terry,

Nice clean wave forms.

It's amazing how clean they are!! I am not sure "I" can take much credit for it ;-) Toshiba, IR, and physics did all the work ;-))



>: Monday, January 31, 2005 9:46 PM
>Subject: Terry's DRSSTC actually hooked to a coil now >:-))

....
>
> http://drsstc.com/~terrell/pictures/FullSystem-01.gif
>
.....

 Why did it miss the zero crossing near the last part of the trace and
change phase so it started to suck out power as opposed to pushing it in?

That freaked me out the first time I saw it too ;-))

The IGBT drive shut off on that cycle (PW~110uS). The logic keeps driving until the zero current crossing and then the IGBTs all go open. But they have reverse diodes to the rail voltages. So after shut off, you are seeing the voltage bounce up against the rails. So you are seeing the system going from driving power, to absorbing it. This is the "recycling" Jimmy and others talk about. The energy stored in the system is being re-absorbed back into charging the rail buss caps through the reverse diodes. Not that it "needs" to do that, but rather that is just the way it simply "likes" to work. If you lock the two lower or two upper IGBTs on, then that will prevent the power going backwards. But the logic in that case is "messy" and apparently it does not make much difference. I was going to do that at first, but it would need programmed logic arrays or a fist full of extra logic. I just skipped that after studying the thing. Steve said it would not make much difference and I came to that conclusion too, although slower.

May not be a problem when you have break out if it occurs before that point.
Is that your scope you can do the FFT on? or was it MathCAD you did that
with?

The FFT on the scope is pretty poor compared to other methods unless you just want a quick idea. Best to pump all the 10,000 data points into a file and let MathCad play those games. Or, Paul's super sophisticated FFT programs could probably extract the name of the person that made the scope probe from the FFT data ;o)) But you will just see the usual two peak FFT unless Steve and Antonio know something at least "I" don't.



Can you give me a link to 8o

What?

Cheers,

        Terry


Robert