Re: OLTC update - Poor seconadry Q

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>

Hi Bert,

At 11:08 PM 8/31/2002 -0500, you wrote:
>Hi Terry,
>
>Yeah, I figured that was too easy! :^)
>
>After reading some of the later posts, I "plugged in" your secondary's
>parameters into a spreadsheet I use for coil design. The impact of skin
>effect and coil resistance drops the loaded Q to about 90. Assuming that
>proximity effects for your closewound coil are of the same magnitude as
>for skin effect, the Q would drop down to the mid 40's (even assuming a
>"perfect" ground to the base of the resonator). Throw in a bit ground
>impedance and add some more losses from the Sonotube, and the measured
>value is certainly in the right ballpark. Bummer! 

Paul's program cam up with a Q of 45.  I sent him a scope trace in digital
form and he got:

--------------
""Couldn't resist a quick look, your TEK00001.CSV gives

PK  FREQ kHz (Error +/-)    Q FACTOR (Error +/-)   LEVEL
 1   33.569 (0.01%,5Hz)       18.09 (3.95%, 0.7)  -3.3dB  1/4 wave
 2   43.335 (0.01%,6Hz)       20.84 (3.32%, 0.7)  -2.6dB  1/4 wave
 3  119.109 (0.01%,16Hz)      22.17 (6.64%, 1.5)  -19.7dB 3/4 wave
 4  181.202 (0.01%,25Hz)      15.10 (31.58%, 4.8) -24.4dB 5/4 wave

The measured Q of the secondary ringing alone is 45, which is
reasonable. But when the primary is involved, the overall Q is
being halved.""
--------------

I also sent him some primary only data.  The Q of the primary looks like
about 35 at low power.  It gets better at higher power.

--------------
""I ran the primary alone as shown:

	http://hot-streamer-dot-com/temp/OLTC08-31-06.jpg

I ran at 5 input voltages of 20,40,60,80, and 100 volts as read on the
meter.  On 120 VAC that can go up to 150 volts and at 240 VAC in it goes up
to 300 volts.  So I still have 3x the voltage to go.  I didn't go higher
here since I am not sure where the "power" is going?

The files are at:

	http://hot-streamer-dot-com/temp/OLTC/8-31PriD-1/

These are basically the voltage across the coil.  I don't have a
trustworthy way of measuring primary current.  I was thinking just a loop
at the bottom of the coil (like on the base) that went across a resistor
and was calibrated...  But never have been able to get that going (time).

I would expect the Q to rise as the input voltages goes up in keeping with
the way the IGBTs should handle loss.  I would be interested in how the Q
changes with increasing voltage.""
---------------

So it looks like I am bleeding Q all over the place.  I plugged the Q
values into MicroSim and got 11mARMS into the arc for an un-astounding
streamer power of 26 watts With 10 inch streamers according to John's
formula.  That is just about what I am seeing at this level.  I do have 10X
the power to go, but I am holding off till I better understand things.  If
I pump 1000 watts into the coil and 50 watts comes out...  I 'really' want
to know 'where' the rest is going!

I expected this OLTC thing to be at the "bleeding edge" of technology :o))
However, the electronics and all went so smooth and it looks like the
"interesting" part is in much more conventional areas of Tesla coil design
that we do not understand very well.  Primary Q, secondary Q, streamer
impedance, power transfer to streamers, losses...  All juicy subjects :o)))

The OLTC is great for work in these areas since it seems to have hit a nice
'bad spot' where are current knowledge is weak.  However, with it's simple
low voltage primary circuits, it is very easy to probe and measure things!
The areas were it raises question were 'next on the list' anyway ;-))

So it appears that the low Q SonoTube is a minor factor given the proximity
effects and all.  No use in trying a higher Q (like polypropylyne)
secondary form.  Probably needs a low Q secondary designed from the ground
up for low Q losses.  I expected low Q in the primary at these lower
powers.  Yeasterday's experiment that showed I could run 'big' current
throught the IGBTs allows me a whole lot of freedom to 'crank this baby up'!! 

I am still amazed that as much goofing around and fiddling with this coil
that I have done that nothing has failed.  The basic machine seems to be
solid and tough!!  It seems very capable of delivering 5000 peak amp long
ringdowns to a well coupled primary, even in it's 'first' form.  I bet the
Q of the primary will go up to like 75 at full power.  The basic figures
would then be...  k=0.25, Ipeak=5000 amps, PriQ=75...  You just 'have' to
be able to make a Tesla coil from there ;O)))

Cheers,

	Terry


>
>Sounds like you'll need a significantly larger diameter coilform wound
>with larger diameter magnet wire to get Q up to a decent level. For
>example, using a 30" x 6.5" toroid atop a 20" coilform wound with 35" of
>#22 AWG (~1230 turns relatively closewound) should give you about the
>same Fo of 38 kHz. However, Q should be closer to ~70 - 80. Like the car
>ad says, wider is [probably] better.. :^)
>
>However, a 30 pound system it ain't (any more...).
>
>Best regards,
>
>-- Bert --  
>-- 
>Bert Hickman
>Stoneridge Engineering
>"Electromagically" (TM) Shrunken Coins!
>http://www.teslamania-dot-com
>
>