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

Re: [TCML] Spark gaps, Solid state switches and diodes



Install a hyperbaric spark gap and you will be really amazed at the
performance increase.

Dr. Resonance





On Wed, Feb 4, 2009 at 5:16 PM, Stephen Hiscock <stephenhiscock@xxxxxxxxx>wrote:

> Thank you so much Bart and others for all the help you've given.
> What you have written is very clear and helpful and makes a lot of sense to
> me.
> It's interesting that you say I should use 1.3 to 1.5 x Cres with the NST.
> I got a lot of my data from Richard Quick's FAQ.
> You may have heard of him.
> But what you say makes sense - I would prefer not to blow anything up!!!
> I might have to upscale the primary caps a bit and perhaps just replace
> them.
> Do you have any good cheap capacitors that you can recommend or even sell
> to me.?
> I will post a separate post after I post this on this topic after this!
> Regarding the tube size on the primary - I didn't mention it because I
> thought it didn't matter too much.
> I'm actually using a 1 inch wide Aluminum strap that is about that is about
> 3mm (2/16ths inch).
> I'm assuming it has enough surface area for a good conductor at those
> frequencies...
> I will have to do some redesign of my spark gap and check my ground
> connections and etc.
> I think I'll check the current each transformer is supplying because I used
> to get 3-4 feet out of my coil.
> Thanks again guys!
>
> Stephen
>
>
> bartb wrote:
>
>> Hi Stephen,
>> Stephen Hiscock wrote:
>>
>>> Bart and fin wrote:
>>> "I get around 600 something turns....... "
>>> "Yes, 648 turns at 32 tpi. I would expect Ipeak to range 185A to 1130A
>>> (15/30 up to 15/180). The primary details are important as geometry greatly
>>> influences Lpri and thus Ipeak. Top loads play a role also since primary
>>> resonance is based on an accurate f2 with all components in place. So if the
>>> system is listed in some detail, then Ipeak is better guesstimated."
>>>
>>> The figures I gave you are wrong, but How were the figures above
>>> calculated?
>>>
>> I assume your asking about the currents given.
>> Ip = Vp x sqrt(Cp/Lp), where Vp = 15kV x sqrt(2) = 21.2kV. C and L are
>> from the cap and primary figures.
>>
>> Say we look at five 15/30's with 32nF.  The  primary should tune in at
>> about 5.9 turns with an L of 22.3uH.
>> Ip = Vp x sqrt(Cp/Lp) = 21200 x sqrt(32nF/22.3uH) = 803A.
>>
>> I meant to say that the Primary is 30 to 100 micro Henries (not
>>> microFarads) sorry silly mistake.
>>> and the Secondary is 30.25 inches winding length at 32 turns/inch ( again
>>> sorry) so 968 turns total
>>> the former is 6.5inches diamater
>>> The top toroid is 6 inch diameter ring and 6 inch diameter inner ring and
>>> 18 inch total outside diameter
>>> eg: (circles represent the toroid) = in courier font it should look ok
>>>  OOO           OOO
>>> OOOOO---------OOOOO
>>>  OOO           OOO             |<-6->|<--6-->|<-6->|
>>> |<----18 inches---->|
>>>
>>> The secondary should be roughly 185kHz assuming 22 awg at 968 turns with
>> the following toroid.
>>
>> yeah the Primary circuit would have two of my transformers as a minimum
>>> (and 6 at a maximum)
>>> I fired it up the other night I used 2 transformers rated at 15kV and
>>> 30mA each
>>> so 15,000 volts at 60mA.
>>> 12nF is my primary capacitance (but as a maximum I could have 36nF)
>>> The primary was tapped at the 10th turn out from the middle. (and as a
>>> maximum could be the 4th turn - but its not working properly)
>>> The primary is 12.5 turns total Turns
>>> The dimensions are 13.5 inches inside diameter and 37 inches outside
>>> diameter.
>>> with 1 inch between turns.
>>> as it is at the moment I'm only getting 6 inch sparks at the best tuning!
>>>
>> Yipes! That's a problem to deal with, but not related to the question
>> (think "gap" for that problem).
>>
>> I'm seeing a tune point of 9.6 turns with the 12nF, so your 10 turn tune
>> point as listed above is close enough that I know I'm near.
>> So in this situation:
>> Ip = Vp x sqrt(Cp/Lp) = 21200 x sqrt(12nF/60.85uH) = 298A.
>>
>> And with this you can see how L and C dictate Ipeak as compared to the 5
>> NST case earlier. If a six NST case, then near 1000A.
>>
>> (but sometimes I would hook up extra capacitance and transformers)
>>> each transformer I have is 15,000 @ 30mA so my capacitance for each
>>> transformer should be 6nF right?
>>>
>> Not necessarily. 6nF is almost resonant with a 15/30 NST. We try to stay
>> away from transformer resonance as the voltage is allowed to rise above and
>> beyond what the NST secondary winding insulation can handle (NST's can die
>> if the static gap is opened up wide enough). So we try to go a little larger
>> than resonance if possible. Roughly, 1.5 x Cres value (but it does vary from
>> coiler to coiler at about 1.3 to 1.5 x Cres on average).
>>
>> You should adjust the gap without the cap across the NST first (just the
>> gap and NST on a workbench). Open the gap until the NST cannot fire. Then
>> close up the gap until the gap just fires consistently. Leave it there from
>> then on. This simple procedure ensures your gap voltage is clamped at it's
>> maximum cold value (and saves NST's from a needless death).
>>
>> The tuning point for the biggest capacitance I can't verify because I
>>> think my capacitor died.
>>> But it should scale up right when I repair the capacitor bank?
>>> just like the primary transformers? eg if I use 2 x the capacitance then
>>> the peak current should be twice as large?
>>>
>> Yes, if you go from 12nF to 24nF "and also" adjust Lpri accordingly for
>> resonance, then yes, the current will be 2X. But if you simply add the C
>> without tuning Lpri, then it will be less.
>>
>> Is that enough info for a ball park Ipeak figure so I can work what IGBT's
>>> I need for a SISG?
>>>
>> It's fine for now. However when listing the primary in the future, give
>> tube size and if the spacing is edge to edge or center of wire to center of
>> wire if you happen to list that spec (actually with ID and OD, tube size is
>> really all that's needed). The ID's and OD's were the most important and you
>> gave those. The info we had was enough to guess on the tube size and spacing
>> reference.
>>
>> Regards,
>> Bart
>>
>>
>>> _______________________________________________
>> Tesla mailing list
>> Tesla@xxxxxxxxxxxxxx
>> http://www.pupman.com/mailman/listinfo/tesla
>>
>>
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla
>
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla