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

Re: [TCML] Re: Hot Primary



Hi Greg

On Mon, Feb 22, 2010 at 9:56 PM, Greg Leyh <lod@xxxxxxxxxxx> wrote:

> Dave Leddon wrote:
>
>  As I push my DRSSTC into ever higher currents (presently limited to 3000
>>> amps) I am experiencing some significant heating in the primary circuit.  I
>>> can see that in the future, as we push the envelop with even higher power
>>> silicon, we may have to resort to water cooling.  Even with 3/4-inch od type
>>> L soft copper water pipe (was that ever a bear to bend) the primary becomes
>>> too hot to touch after running the coil for one minute.  I even changed the
>>> interconnecting  wires from #4 to #2 welding cable to prevent meltdown, but
>>> oddly the larger wire gets just as hot.  I wonder if there isn't some sort
>>> of rf heating effect unrelated to voltage drop which is generating some of
>>> the heat.  A related issue which has been bugging me for some time, why are
>>> the active primary turns progressively warmer as you move inward toward the
>>> secondary?  One would think that all turns carry the same current.
>>>
>>
>
>
> I'm unfamiliar with the particulars of your primary drive circuit, but it
> seems to me that there shouldn't be *any* noticeable heating of the copper
> at all, otherwise the unloaded Q of the primary will be very low.
>

Hopefully we can increase your familiarity ;-).  The DRSSTC has the capacity
of running much longer drive periods, whereas transient excited systems are
limited by losses and coupling coefficients.  The capacity for running many
cycles of RF per spark event, allows the operator to really drive up the
primary RMS current to huge levels because the duty cycle can be
significantly higher.  But back to numbers... the Q for this system is
easily more than 10, as im betting his primary voltage is about 10X the
H-bridge output voltage when he hits 3kA, and with no secondary in place im
guessing the primary envelope is a fairly straight ramp, meaning the Q is
likely in the 30-60 range at least.  Perhaps if Dave has a DSO he could
capture his primary current shape (with no secondary in place) ramping up to
3kA.  It would be important to capture the bus voltage as well (because bus
voltage sag would make the primary current ramp taper off, artificially
looking like lower Q).  From there we could determine the actual Rac of the
primary, at least to some degree of certainty based on how good the data is.


>
> When I first brought up the twin prototype, I noticed the unloaded Q of the
> primary circuits was horribly low... about 10.  Even then, the IGBT heatsink
> was the only thing that got warm to the touch.


Can you give us some idea of what the RMS primary current was?  I doubt its
near the 100's of amps that Dave is likely operating at.  DRSSTCs do pay the
penalty for higher RMS currents, though this is often a design issue, or the
operator just pushing the system very hard.  I bet Dave could run at half
the primary peak current and a lower duty cycle and still make really nice
sparks at a higher efficiency, but sometimes its more fun to push things
:-).


>  To raise the Q, I had to beef up the copper in key areas.  I discovered
> the weak points by passing 100A DC continuously through the entire primary
> circuit, and surveying IR drops point-to-point with a DVM.  After a bit of
> copperwork I got the unloaded Q above 30 and stopped there, since the loaded
> Q is well below 10.  The primary is a single turn of 6" Cu strap, as can be
> seen near the bottom of this pic:
>

I like the idea of this method, since it allows for convenient and safe
measurement of purely resistive loss.  I might give that a try on some of my
systems.

Steve
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla