[Home][2015 Index] Re: [TCML] Question for solid-staters [Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] Question for solid-staters

Ken, that's correct; Ultimately, it's the total ampere-turns applied to the secondary that matters.

Long ago at SLAC we developed solid-state modulators that used many primary turns in parallel on the main HV transformer. Each primary and it's solid-state driver could then be electrically isolated from the others. You would add more primaries in parallel to increase the total drive level and throughput power.

Here's a solid-state design from 2001, which used 76 primary circuits in parallel:


This approach was called a 'fractional-turn transformer' and worked ok, although HV IGBTs back then were still improving at a rapid pace. We were using 3300V IGBT modules, at around 2500V each. There were some odd failure scenarios though, which prompted me at that time to start cracking open and studying faulted IGBTs:


With sufficient care and synchronization, think that paralleled primary circuits should work ok on a Tesla coil. -Greg

On 4/13/2015 12:42 PM, tesla-request@xxxxxxxxxx wrote:


Message: 1
Date: Sat, 11 Apr 2015 17:17:41 -0700
From: Ken Herrick<kchdlh@xxxxxxxxx>
To: TCML<tesla@xxxxxxxxxx>
Subject: [TCML] Question for solid-staters
Content-Type: text/plain; charset=utf-8; format=flowed

...If there are any left, that is, paying attention to this List.

1.  Years back I made a way-too-complex s.s. coil, involving several
multi-TO220-MOSFET assemblies, daisy-chain-connected into one equivalent
primary coil.  It finally worked but its mtbf was way too short and I
finally tossed it.
2.  Now I'm trying to re-create the general idea, with just a single
pair of much-huskier MOSFETs per module, and employing such 4 modules.
Ref. my earlier posting this year. Each is to operate from the rectified
and doubled mains--in my case around 300+ V.  A single one currently
works on the bench, driving at least ~200A p-p~5 ms pulse-bursts into a
single 12" diameter coil, at ~100 KHz and ~5% d.c.

3.  It occurs to me that rather than daisy-chaining the ultimate 4
modules into 1 equivalent full-diameter primary coil (to yield an ~1200V
source applied to the primary) it would be muchsimplerelectrically to
establish 4 synchronized parallel primary flux-paths passing through the
one secondary.  Four primary coils would be used, each one occupying a
90 degree quadrant of the full secondary diameterand situated in the
same plane.  All4 MOSFET-module/primary-coil groups would be identical
and driven from the same ~300V and l.v. gatesources. Eachprimary coil's
turn-quantity would be sized to yield the maximum
pulse-burst-duration/duty-cycle that the MOSFETs would accommodate.

Seems to me that the respective synchronized fluxes would have minimal
interaction while trraveling through the secondary, so the resultant
total flux would be of more or less the same magnitude as if I were to
daisy-chain the 4 modules into a single equivalent primary coil of ~4x
the c.s. area.

Can anyone shootthis down?

Ken Herrick


Tesla mailing list