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

Re: Passive Ballasting for DRSSTC - My thoughts before Ed Wingates Teslathon



Original poster: Jimmy Hynes <jphynes@xxxxxxxxx>



On 8/27/05, Tesla list <<mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx> wrote:
Original poster: "Daniel McCauley" <<mailto:dhmccauley@xxxxxxxxxxxxxxxxxxxxxxxxxx>dhmccauley@xxxxxxxxxxxxxxxxxxxxxxxxxx >



> First of all, I don't see how the drive pulsewidth can be different > than the ON time, especially if you are not using a JK flip flop to > ensure a soft turn off. Even given a constant ON time, there was a > noticable effect. I wasn't claiming that the effect was purely > different ON time, but that if you are gonna try to compare them, you > should at least compare them honestly and not try to cheat a few cycles.

I can't either.  But if i keep drive pulsewidth constant, and simply change
the wire gauge on the
primary, cycles do get cut off in the smaller gauge case.  Definitely weird,
as i would not
expect that to change.



Something's just not right there. If you trace it out, you have to be able to figure it out. As soon as the enable pin goes low, it should stop driving, and ring down.

> Do you still notice the effect when you don't change wire length? I'd
> find that hard to believe. The only other change is the resistance,
> which is a simple ohms law calculation.

Yes.  Wire length can be constant, and the effect is still seen.



Hmmm... Sorta hard to believe, because that would mean that it's only resistance, and that just "shouldn't" do anything with the extra R you added.

> Not really a 'fluke', just that it works with your coil, with your
> tuning. Of course identical set ups will perform identically. It
> depends on how you tune the coil in the first place (running a
> slightly lower tuning effectively cancels out your 'ballast'
> inductance). In other designs, it may not be effective, and in the
> case of primary current feedback, it will do nothing, since it will
> simply drive it a bit slower, and keep building up current.

What ballast inductance?  Granted, parallel strands of conductors (larger
wire with more strands) flowing current in the same direction
will increase inductance, but not enough to make any noticeable difference,
especially if wire lengths are kept constant.  Also, i can tune
the coils anywhere (i did that this weekend to further investigate this),
and it does not affect this ballasting effect one way or another.


I was assuming it was extra length of wire. It just doesn't make any sense the other way


> First of all, we have to look at the possible effects of adding some
> small wire to the setup. The effects can be inductive, or resistive,
> or a combination thereof.

No, its not adding any small wire.  Wire length can be kept constant, routed
exactly the same way, and the ballast effect occurs.

> The extra inductance both changes the primary resonant frequency (as
> does changing the tap point), and reduces the coupling (as does
> raising the secondary). Therefore, you shouldn't add thin wire for
> the inductive effect, since you can simply retune (and raise the
> secondary, although it's probably more about the tuning), and thinner
> means more loss.

Again, using identical lengths of wire under same routing conditions, the
only difference i can see are in the R of the wire, and just a slight bit
of inductance change due to the variation in parallel, current paths from
the changes in number of strands.

Thanks for the response.  I'm still a bit stumped.  Even after additional
tests this weekend, i just can't figure out why simply replacing about two
feet of 4 AWG connections with 8 AWG connections reduced ground strike
current by 300A ! ! !  Even when routed identically with identical lengths,
the change still exists.
Oh well . . .



By my calculations for 2 feet of 8awg replacing 2 feet of 4awg at 150khz, I get another 2.23 milliohms, which is really tiny.

For some rough calculations of how much you should need to get that effect, say you're getting a 30% reduction in voltage, and therefore 30% in current. Then the current drops from say 1000a to 700a, and the drive voltage drops from say 300 to 200v. 100v/300a is 0.33 ohms, which is a couple orders of magnitude more than what I'm predicting for your wire.

If you feel like looking into this more, the next step should be to run it like that for a while, and see where the heat is actually going. If you're running 1kw, then the wire should see a couple hundred watts if it's really pulling the current down.

Dan