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

Re: Semi-Rotten Spark Length



Hey Reinhard,

Hope your feeling about back to normal now. I don't reply much on the list
these
days as I'm really deep into moving as you know, but I couldn't resist your
post.
Thanks for the details of your setup. I wish all were as detailed as you are.

I ran the numbers through my own program (yep, a bit different than the one you
have). Interesting query. I've been working on my own spark length calculation
and have been running numbers as people post their designs. I am not satisfied
with the results, but I am getting closer all the time. My latest spark length
equation is a bit complicated without full representation of what I'm
doing. In a
nutshell, my equation is SparkLength = sqrt((sqrt(2eJ / Ct) * Ctop) * 0.6

I am still taking Ctop into consideration, however, I am now using the
effective
top terminal voltage which is based on several circumstances: In the equation
above, (sqrt(2eJ / Ct) is the effective output voltage. I use this value
Ctop to
express a ratio of voltage and storage capacity to derive inches of spark
length.
However, in many of my expressions, "effective" is the key. This word always
denotes that I am deriving a value based on the variation of something
else. With
the terminal voltage, eJ is the effective joules. In order to come up with
eJ,  I
must define  .5 * Cp * eVp^2. Notice that eVp is "effective Vp^2". Here I
define
eVp as  Vp * Cp Charge. Basically, I use the charge time as derived by your cap
size and the charge time given. Cp Charge is of coures a percent based on the
expotential curve of capacitor charging current. And further, charge time is
"enhance" with your spark gap presentation time, power input, etc...

The best way to see what I do is of course to take a look at my latest file
(I'll
send it - just plug in your values).

What I saw after putting in your numbers was "NOT" a big arc length. John
Freau's
formula's vary only slightly with my own. John Corture's calcs are of course
controlled lengths and will not apply to your situation. J.F's. equation
defines
a  52.8 inch spark length (not far off from what you are seeing!) based on
power
input of his own experiments with his coils with proper top load sizing. My
calculation using your latest coil defined a 51.1 inch spark length (yes, close
to John's and close to your own real life measurement). However, your
toroid size
is small for the power input.

Calculating out your toroid size at 3:1 of Cself (common for those with
long arc
lengths), your toroid should then be about 42pF, which is about a 10" x 38"
toroid or just under 50% larger than your 2-toroid stack (primary tapped at
7.15t
for calc tuning). With this taken into account, my personal sparklength
equation
then comes up with about 62 inch arc length. The other equations will not take
into account the Ctop change. The differences from John's equation is only
slight. My difference is an example of a 3:1 max Ctop to Cself ratio (who's to
say this is the limit though - no one, only your coil).

So, dispite your arc lengths achieved, the calculations are telling you
that your
real life observations are very close to ball park equations. I would never
assume  .5 * Cp * Vp^2 will dictate your joules at the output, as there are
many
variations between this and the end result. But, take into account the
spark gap
( I used a 100 bps , 50Hz gap), charge time of the cap, and presentaton time of
the gap, the effective energy gets relatively close and begins to show a
picture
of high or low speed gaps, high or low voltages, etc... and their effect on arc
length based on standard equatons taking into account the effected voltages,
joules, etc... My equation is not perfect, but it is getting there - I think.
There is a lot to do and as always, the file is in the middle of some thought
process. It is never completely satisfying and is why I've never put it out
except to a couple people. However, as I am real close to moving to CA and
almost
all of my time is spent taking care of that, I may throw it up on my web page
until I can get back into the Tesla mode again.

My bottom line is to use the file (even if just for a few minutes) to find out
the best top load and power input for you coil. Works for me (so far). You
might
be surprised - good and bad.

Take care,
Bart
(BTW, decided to take your advice and hall all 32 caps to CA with me. Their
already crated. Also, keeping the Dodge. Throwing a new cam in this week - for
the torque of it!)




Tesla List wrote:

> Original Poster: "Reinhard Walter Buchner" <rw.buchner-at-verbund-dot-net>
>
> Hi All,
>
> Iīve got a problem. Maybe someone has some ideas.
> First of all, here is my complete coil data:
>
> Xformer:
> -----------
> 7500V modified NST (with a movable shunt).
> (originally 75mA)
> I shortend the shunt.
> Now: 170mA (measured)
> Input voltage: 230-260V
> (depending on setup)
>
> Primary Cap:
> -----------------
> MMC: 67nF
> Caps per String = 12
> Strings = 13
> Single Cap= 62nF-at-1000Vdc
> Input Joules = 0.5*7500V*sqrt(2)*(67*10^-9) = 3.77J
> (for 15kV comparison this would be a 16.75nF cap)
>
> Toroids:
> -----------
> 4.33" x 20" = 21.70pF
> 5.31" x 23" = 25.19pF
>
> The toroids are stacked directly on top of each other.
>
> {Ct= k1*C(T1) + k2*C(T2) where K1 ~0.7-0.8 (for the
> toroid closest to the top of the secondary coil) and k2
> is 0.4 - 0.6 depending upon the spacing between the
> two toroids.}
> so:
> Ct1 = 0.75*21.70pf= 16.27pF
> Ct2= .5*25.19 = 12.59
> => CtTotal = 28.86pF
>
> Primary Coil:
> ----------------
> Flat primary
> Total dia: 27.17"
> Inner dia: 11.8"
> Width of one side = 18cm = 7.09"
> Center to center = 24.2cm = 9.5276"
> Tapped at turn 5.8-6.0 (6.3 would be optimum, claims my homebrew proggy)
> Pri induc = 23.44ĩH
>
> Pri current = V*SQRT(C/L) = 10606*SQRT(67*10^-9/23.44*10^-6) =
> 567A (43A per cap string).
>
> My MMC stays ICE cold, even though I ran the coil for a total of 1.75 hrs
> (a few minutes at a time, cool down time was low, as the coil was off
> only during adjustments).
>
> My secondary coil specs:
> -------------------------------
> Wire spool weight before/after winding: 22lbs/12lbs
> Wire specs:
> i.d:0.85mm = 0.033465"
> o.d.:0.908mm = 0.035748"
> AWG 19-20 (19.5)
>
> Secondary former:
> ------------------------
> Diameter: 20 cm = 7.874"
> Total length:106 cm
> Chosen winding length:83.65 cm = 32.93"
> h/d ratio: 4.1825
> Turns:920 (+1.25 turns space wound over a length of 12cm = 4.72")
> 12 cm = heigth of toroid above actual coil winding
> Aprox wire length: (579m = 1898 ft)
> DC resistance= 17.0 ohms
> Calcīd inductance: 36.047 mH
> Measured 35.6 mH
> Cself=14.75 pF
> CTotal ~43.55pF
> FresFree=218khz
> Fresloaded = 127khz
> Bare secondary Q (measured) = 196
>
> Spark length data from start of project till now:
> -----------------------------------------------------------
>
> 1.) 15nF MMC & 562VA unmodified NST. This got me around 30" sparks
> 2.) 25nF MMC & 562VA unmodified NST. This got me around 41" sparks
>
> 3.) 31nF MMC & 900VA unmodified NST (I just turned the shunt out for 120mA).
> This got me around 50" sparks.
>
> Up to then I was really happy with my coil and itīs performance, but
> letīs look at (being the latest setup) case #:
>
> 4.) 67nF MMC & 1257VA modified NST
> (I shortend the shunt for 170mA output current).
> This setup only gets me around 57" sparks. Only once did I get a 63" spark.
>
> NOTE: Spark length is measured "point to point". I do NOT "straighten out"
> the spark. Doing so, would give me longer spark lengths, but the way I
> measure the length is more realistic, I think.
>
> Now, why the **!-at-##-at- is my spark output in case number 4 so rotten?? I
> almost doubled the input Joules compared to setup number 3. I increased
> the xformer current by a factor of 1.4. I would have expected to get at
> least around 65-70" continuous sparks, but no go. The only thing I could
> finally get the coil to do, was to break out a single streamer in case #4,
> using a blunt nail. Without the nail output spark length was only slightly (I
> didnīt actually measure it, I just eyeballed it) less than the 57", but the
> sparks swirled all around the toroid, making measurement difficult. I
> tried various additional "toploads" as a discharge electrode. I also tried
> increasing / decreasing coupling, but it didnīt help in making the sparks
> any longer (making them shorter was easy enough though :oÞ). Best
> spark length was achieved, when the primary (!) was ~4.8" above the
> lower most secondary winding. Going up to 5" didnīt change it too much.
> I didnīt go above 5" as I have seen racing sparks in earlier setups (i.e: in
> case 3) appear above 5". Decreasing the primary heigth to 4" started
> reducing spark length. Primary tapping is semi-critical. Moving the tap
> more than (+/-) 1/4 turn will let the output drop like a rock. None of the
> above MMC are mains reso caps, although case #3 does come pretty
> close to it.
>
> Any clues, tips, or comments are more than welcome, because I am
> totally STUMPED on this one.
>
> Coiler greets from germany,
> Reinhard