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

Re: [TCML] The Dreadful Task of Ballasting (longish)




--- On Mon, 5/11/09, jhowson4@xxxxxxxxxxx <jhowson4@xxxxxxxxxxx> wrote:

> From: jhowson4@xxxxxxxxxxx <jhowson4@xxxxxxxxxxx>
> Subject: Re: [TCML] The Dreadful Task of Ballasting (longish)
> To: "Tesla Coil Mailing List" <tesla@xxxxxxxxxx>
> Date: Monday, May 11, 2009, 6:22 PM
> Alright I have to ask... 
> 
> Where have you gotten all those Laminated Cores? 
> 
> I think I have seen 8 different cores now. 
> just curious is all ... 
> 
> thanks 
> Jay Howson
I recently purchased J class ferrite in a 2.5 OD toroid from Amidon. This J class ferrite is advertised by spec.s to have a high permeability factor. 35 turns of common insulated 14 gauge wire painstakingly wound around the 2.5 in. OD toroid yielded ~45 mh. To achieve the same air core  inductance value would involve stacking four 500 ft spools of 14 gauge wire as sold in hardware stores. I have used these spools in source frequency resonance investigations with a garage set-up with a AC car alternator delivering 465 hz. The actual q factor in series resonance @ 465 hz for this J class core toroid was only 3. In comparison half the inductance when it is air core can deliver Q series resonant voltage rises in the 15-20 range.
However as I have recently discovered the "source frequency" series resonant Q factor may be entirely different from the parallel tank circuit Q factor used in power factor correction schemes.
> ----- Original Message ----- 
> From: "Phil Tuck" <follies@xxxxxxxxxxxx>
> 
> To: "Tesla Coil Mailing List" <tesla@xxxxxxxxxx>
> 
> Sent: Monday, May 11, 2009 3:50:58 PM GMT -05:00 US/Canada
> Eastern 
> Subject: RE: [TCML] The Dreadful Task of Ballasting
> (longish) 
> 
> Bert, Scott, Harvey, Bart et al 
> The situation has taken a slight turn for the better now. I
> am basically 
> waiting on the company who sold me my DVM to get back about
> a replacement, 
> so in the meantime I wound a permanent ballast 
> http://homepage.ntlworld.com/follies/tesla/inductor.jpg
> by the 'suck it and 
> see ' method. Eventually after gradually chopping bits off
> the winding I got 
> to my desired 16 amps (this gives the secondary a
> reasonable current). Out 
> of curiosity more than anything I tried it on the coil and
> the safety gap 
> firing is significantly reduced. There are a few too many
> safety gap firings 
> to be happy with, but it is on the way to success. I am now
> wondering that 
> maybe my previous temporary ballast's, that also gave
> 16amp, were in fact 
> saturating.
An interesting experiment with tuning forks, which are magnetic, being suspended in a air core 2.4 H air core series resonance showed that it was attracted to the core magnetically at 60 hz resonance, but at 465 hz resonance the magnetic attraction was weak and the tuning fork heated up quickly. Noting the 60 hz actions, three or four of these tuning forks might be placed around the core and exterior winding of the 2.4 H coil to make a somewhat complete magnetic loop of iron. However this small of a cross section of magnetic material around such a huge amount of 9,000 turns of 23 gauge wire to produce 2.4 H involved with this volume of space would probably lead to wonderings about saturation theory to begin with. In short I tend to disbelieve that putting iron circles as a toroid formation around such a comparatively large inductance will reduce that inductance to nothing, as saturation theory implies. This is because the speculations of saturation
 theory imply that after a certain amount of amp turns, the ferromagnetic core becomes non-magnetic??? Good God it doesnt take much current with a winding of 9000 turns to exceed this so called saturation effect, but this example may be the other side of the coin. Nevertheless with 9000 turns of 23 gauge wire on a  3 inch ID, one can turn the current up via variac, but this does not decrease the magnetic attraction of the tuning fork to the core with alternating magnetic fields, instead it increases the attraction; whereby we might suspect that it if the amp turns are predominant for that consideration, the magnetic attraction should go weaker. But this is precisely what occurs when the frequency is increased to 465 hz.

 Could this have been the culprit ? 
> The small MOT got hot pretty quickly and the welder is only
> a cheap 150 amp 
> which also has a rather inadequate core that got hot.
> Whereas the ballast I 
> wound only gets slightly warm to touch after 5 mins on a
> Jacobs ladder. 
> With the new ballast I am still getting around 60 volts
> across the tranny 
> when the secondary is short circuit, so the tranny is still
> partially 
> behaving like an NST 
> 
> As my caps are woefully STR anyway I'm ordering some more
> CD 942's to take 
> me up to 75nF
Today I began more extensive investigations into a secondary power factor correction scheme using 75 nf with air core coils@ 60 hz. This required quite a large amount of air core inductance, about half of the supply on hand. It required two 70 lb 8 mile coils @ ~ 60 H and four of the 12 lb ~2.4 H coils. I have shown these smaller coils before on flicker URL's where the series resonant Q factors are 10, but the larger coils of 70 lb supply a 60 hz Q of 15. Disappointingly it appears that the same combinations of L and C made to be balanced by observations of independent reactive amperage consumptions rather then by formula considerations show that the values deliver quite a miserable performance factor in a tank circuit, with only Q factors in the 4-5 range at 60 hz. Some non-linear factors seem involved in that a perfectly tuned system tuned with 16VAC variac input showed a Q of 5.3, but when 150 volts was applied misbalanced amperages on the branches
 were noted, and the Q was reduced to 4. What this means in practicality is that four times less amperage enters the LC loop then exists in the loop itself. I have not yet tested this with a variac controlled 15K 30 ma NST. Four of these are available to be paralleled for impedance matchings of supply and demand.


(currently I have 42nF). This increase,
> though still STR, will 
> definitely move the resonance point and is something that
> the coil needs 
> anyway. If I still get safety firings with the new cap size
> I know then that 
> my stubbiness to stick with my heavy duty single static gap
> is the problem 
> and I can then complete my RSG (I milled the motor's rotor
> months ago). 
> When I get a DVM I will use Bart's spreadsheet and post the
> actual results. 
> 
> Regards 
> 
> Phil 
> www.follytowers.co.uk/tesla 
> 
> 
> 
> 
> 
> 
> -----Original Message----- 
> From: Bert Hickman [mailto:bert.hickman@xxxxxxxxxx]
> 
> Sent: 11 May 2009 04:09 
> To: Tesla Coil Mailing List 
> Subject: Re: [TCML] The Dreadful Task of Ballasting
> (longish) 
> 
> Phil Tuck wrote: 
> > Bert & Bart, et al 
> > 
> > As of today my DVM has died so until that's changed
> next week I will hold 
> > fire on the practical side of this. I will get some
> more Caps though in 
> the 
> > meantime for either a total Cp of 64nf or maybe 80nF 
> > As Bart mentioned to me my DVM will most likely not be
> giving an accurate 
> > reading anyway on my tranny (core = 8 sq ins). 
> > I think that was proved today before the DVM died. The
> small MOT was 
> ~52mH, 
> > which with its small core is probably accurate. So I
> built a usable 
> inductor 
> > with some identical big cores as those that the
> transformer uses. 
> > This measured to around 54 mH when finished, so
> therefore I was expecting 
> > the same current draw of approx' 16 amps that I got
> yesterday with the 52 
> mH 
> > MOT. However I only got ~6 amps, so I had to remove
> some turns plus some 
> > packing shims, as obviously the new inductor was more
> than the 54mH that 
> the 
> > DVM measured. 
> > So it is not up to the job with measuring inductance
> with big cores. Same 
> > goes for the welder reading I gave as well possibly,
> meaning yesterday's 
> > data may not be reliable enough. I have Barts
> spreadsheet to use when I 
> get 
> > a new DVM. The spreadsheet will be more reliable as it
> gives the results 
> > based on current and voltage measurements. 
> > 
> > Bert wrote>> 
> >>>>> If I understand your system correctly,
> your new transformer outputs 
> > about 11,500 volts with 240 volts input, making the
> approximate turns 
> ratio 
> > about 48:1. The turns ratio squared would be about
> 2296, making your tank 
> > cap look >>>>2296 times larger on the LV
> side of the transformer. 
> > ___________________________________ 
> > 
> > The turns came out to 42:1 Bert. I think with the way
> I have chosen to 
> build 
> > the tranny (centre tap with the two secondary's on the
> outside legs, each 
> > only getting half the core size / flux) that using the
> voltage output to 
> > determine the turns ratio is misleading. This method
> always assumes that 
> > your getting sufficient flux to start with.
> Theoretical voltage would 
> > actually be 240 * 42 = 10080v on each leg! I only
> actually get ~5750v 
> though 
> > on each side (57% - not too bad for half the core
> size) Interestingly when 
> > the secondary is shorted, the voltage drop across the
> ballast is around 
> > 180v, while the tranny still has 60 volts across it,
> so it seems to be 
> > acting a bit like an NST, because the core arrangement
> is affecting the 
> > coupling I suppose. Possibly even an added bonus. 
> 
> Very interesting! Now I understand your system much better.
> Your HV 
> transformer is actually behaving more like an an NST than a
> plate or 
> distribution transformer. This is apparently because of the
> combination 
> of the core and winding configuration and core mating gaps.
> Distribution 
> transformers have very low leakage inductance - the %Z
> rating typically 
> only 2-3%. If your 240 volt 4 KVA 50 Hz transformer had a
> comparable %Z 
> of 3%, its LV side short circuit impedance would be 0.0829
> ohms, 
> reflecting about 0.264 mH of leakage inductance. Based on
> your ballast 
> and transformer voltage drop measurements, the short
> circuit impedance 
> of your transformer is about 33% of the impedance of your
> ballast. The 
> %Z of your transformer appears to be 10-20X that of a
> "stiff" 
> distribution transformer, and more typical of the behavior
> of a 
> "self-ballasted" NST. 
> 
> > 
> > 
> > Bert also wrote>>>> 
> >>>>>> The turns ratio squared would be
> about 2296, making your tank cap 
> look 
> > 2296 times larger on the LV side of the transformer.
> This makes your 43.2 
> nF 
> > tank cap "look" like about 98 uF on the LV side of
> your transformer. A 54 
> >>>>>> mH ballast inductor in series with
> this will result in a resonance 
> peak 
> > of about 59 Hz. With this ballast, your system is STR.
> 
> > ____________________________________ 
> > 
> > I see you are only taking the ballast's inductance
> into account. I was 
> > unsure when I originally posted, but thinking about it
> since, shouldn't 
> you 
> > also be including the tranny's inductance as well
> during the charging 
> cycle. 
> > Admittedly when the SG fires the only inductance in
> the primary side 
> would 
> > then just be the ballast, as the SG's short get
> reflected to the primary, 
> > but when the cap is charging won't it have the primary
> inductance as well? 
> 
> 
> You're correct. My spreadsheet implicitly "assumed" a pole
> transformer 
> where the LV side leakage inductance was a small fraction
> of the 
> external ballast inductance, and it's contribution was
> virtually 
> insignificant. The sum of the inductances should be used in
> either case, 
> but it makes a big difference for your transformer. 
> 
> > 
> > 
> > 
> > Regards 
> > 
> > Phil 
> > www.follytowers.co.uk/tesla 
> > 
> > 
> 
> Best wishes, 
> 
> Bert 
> 
> 
> _______________________________________________ 
> 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