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

Re: [TCML] Pig SISG: Destroying the Indestructible

In a message dated 3/22/08 8:09:12 P.M. Eastern Standard Time,  
pterren@xxxxxxxxxxxx writes:

>I have purchased from Mark Dunn many of the parts for a 20 stage  SISG 
    Keep up the good work, Mark!
>to run on a smaller TC 
    It seems to me the SISG seems to work well in a  certain "middle-sized" 
range of coils with the current cost-effective IGBTs  available. 
>but also later on my 18 inch coil that runs 2 PT's.  So it  will be a bit 
mini piggish.

    I'd think more than a "bit"!

>Hence I am interested in the results of higher powers.

OK, I'll try to make a concerted "scientific"  effort here.

>Are you sure that stage still works? 
    No idea yet, but as I have to make a repair to  the board, I'll check it 
as I go. Keep pestering me if I don't report back.  

>The IGBT would have been over volted and destroyed if the voltage  across 
the SIDACs went higher >than 1200V.

I'd think so. On the other hand, the IGBT appears  to be intact. Most things 
in the path of the primary current, especially  electronics, tend to fail 
spectacularly. Plus, I think when the middle SIDAC  in that section went, it was 
effectively a short (much soot, flame, and  carbonized remains). Shorted SIDACs 
are no problem, since they *drop* the  section firing voltage. An open SIDAC 
might be a problem, since it  means that section will never fire (although I 
wonder if the 1 Meg balance  resistor across an open SIDAC would screw things 
up). If the other sections  fire, the non-firing section might become 
over-volted and fail?
    Now that I took a closer look, not only did that  one SIDAC fail, but the 
24V 1500W TVS (D31 in Mark's schematic) failed as  well. The case is split 
axially from one lead to the other, with a bit of  scorching evident as well. I 
can only assume it failed as short as well: since  the coil continued to run 
just fine, with no other spectacular failures, the  primary current was still 
flowing through that stage. Unless the IGBT (or body  diode) was shorted with 
no externally visible damage, the only primary current  path around the TVS 
would have been through the 1N5819 diode (D32) and then  through a number of 
smaller components that should have been instantly blasted  by that current.
    Everything else in the entire SISG array except  the SIDAC and the 24V 
TVS looks just fine. FWIW, the connecting lead between  SISG boards would be an 
attractive point for arcs to hit under the primary,  and that TVS has one lead 
on the same trace as that jumper wire connects to.  I'm really starting to 
think a direct hit from the topload was to blame. 
>I don't recall the details of your coil. Could you give me the basic  

>and any info you have on spark length and power  input?

    Sure! And I'll bet Bart will enjoy doing his  calculations on these specs 
as well. ;)
    Power source: 240VAC (measured) residential single  phase (120VAC to 
neutral/ground each leg)
    Line filter: 100A single phase commercial  unit
    Ballast: handmade, open "E" core of laminated  transformer steel, 
quadrafiliar wound 12 AWG, limits line current on  Pig to 60A max
    Voltage control: four ganged 30A variacs, two per  leg in parallel, no 
balancing chokes, but calibrated to all run within 1V of  each other. Allows max 
output of 280V from 240V line.
    Ammeter: old moving vane type, direct in line
    Voltmeter: old, moving coil, calibrated
    Transformer: 10kVA 14.4kV / 240V Wagner pole  pig
    Rectifier: full bridge, two legs still 40  seriesed UF4007 fast-recovery 
1kV diodes, other two legs "bar" rectifiers  unknown specs (20kV 1 Amp 
fast-recovery?) but very heavy duty.
    No DC filtering, charging chokes
    Each DC leg has a 200 Ohm 100W wirewound power  resistor in series 
immediately after the bridge to protect the rectifiers  against current surges. 
These get merely warm on runs of a few minutes.
    Primary cap: two Maxwells in parallel. One is a  31976, .075uF, 40kV; 
other is the ubiquitous 37667, .030uF type. So total  .105uF
    Primary: shallow inverted cone, 7" ID, .375" OD  copper tubing, 3/4" 
spacing on center, currently tuned at 9-1/2 turns (so now 7  unused outer turns!)
    Primary inner (bottom) turn 1-1/2" below bottom  turn of secondary
    Gap: SISG, six of Mark Dunn's SISG4 boards in  series, nominal 21600V 
breakover voltage. For (Rx5), I'm  using 5kOhm trim pots, DigiKey #CT6EP502, left 
at their factory center  position settings (2.222kOhm +/-1 Ohm). For (Rx4) 
I'm using the 47 Ohm 2W 2%  metal film resistors, DigiKey #P47W-2BK. For what 
it's worth, I'm using 1% metal  films for (Rx1,2,3), DigiKey #1.00MXBK  Using 
AAVID Maxclips to  mount the IGBTs to the heatsinks. 
    Secondary form: 6.25" OD thinwall PVC sewer  pipe
    Secondary winding: 25 AWG, .019" wire OD with  insulation, 30.5" winding 
length. Deduct 1-1/4" of winding length from the top  half due to repairs. 
    Corona ring at top: 3" minor diameter, 12" major,  centerline 2" above 
top winding of secondary
    Main toroid: 8" minor diameter, 36" major diameter,  centerline 9" above 
centerline of corona ring
    Hope this helps! 
-Phil LaBudde
Center for the Advanced Study of Ballistic  Improbabilities

**************Create a Home Theater Like the Pros. Watch the video on AOL 
Tesla mailing list