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

Re: Max Indutance of a Solenoid



Original poster: "Barton B. Anderson" <bartb@xxxxxxxxxxxxxxxx>

Hi All,

I have a correction to make. While contemplating the maxL data, I noticed values of Q were odd looking. I verified against low h/d measurements which Malcolm Watts had made. As I investigated the situation, I found an error that effected h/d's of less than 1 in Javatc. This problem did not affect h/d's greater than 1 (so likely, no one).

Anyway, findings haven't changed much but the data had to be completely rerun and updated. This time, I increased the h/d increments to 0.01 (meaning 100 different simulations now) so much better resolution. It took a couple days to do, but I wanted to get it run and done before I responded.

http://www.classictesla.com/download/max_L_table.pdf

Maximum inductance at resonance (Les) occurs at an h/d of 0.20.
Maximum inductance at low frequency (Ldc) occurs at an h/d of 0.40.

It will be interesting to check against other coils with similar simulation. Paul Nicholson back when TSSP was strong ran a virtual secondary simulation on over 13578 records (coils) which took 30 days using a 14 Beowulf Cluster, and that data is far higher resolution than my own. BTW, you all have access to it if ever you want to look at "anything". Likely the best database of simulations (anywhere).

http://www.abelian.demon.co.uk/tssp/vsd/

Well, my pdf is available for anyone to crosscheck. Javatc has also been updated to 10.N to correct the error in Q (and Rac) with h/d's less than 1. In Javatc, Q = sqrt(Les/Ces)/Rac, and Rac required an h/d factor (an extremely long calculation). Because the h/d factor was off for low h/d coils, these two outputs (Q and Rac) were also incorrect. All other outputs are fine and have no relation to this error for all h/d values.

In the pdf, I also included a chart of both Les and Ldc and also added a little more info for crosscheck.

Take care,
Bart



Tesla list wrote:

Original poster: "Barton B. Anderson" <bartb@xxxxxxxxxxxxxxxx>

Hi All,

After looking at Gary's coil data from 7 years ago, Mike asked offline to look at max inductance of a solenoid coil. Sounded interesting so I did. If you have been on the list long enough, you've likely heard max inductance h/d ratios. The most commonly stated is maxL is when H=0.9R or H=0.45D. I ran some numbers to look at that. BTW, for those particular max inductance statements, a fixed wire length and a fixed wire pitch must be maintained.

Here is a pdf of the data I ran:
http://www.classictesla.com/download/max_L_table.pdf

The first table is a solenoid coil where the wire length is maintained and the wire pitch is maintained. This requires the radius and height to be varied as h/d is incremented (and in my table, by 0.05 increments). What was found is that at low frequency (Ldc), H=0.45D is relatively true, but only for low frequency inductance. This has no bearing on a resonant transformer. At the resonant frequency of the transformer, Les is realized by the transformer. Les is the lumped value for the inductance at the transformers resonant frequency. It is easiest to view it as the "resonant frequency inductance". At high frequencies, the current is not distributed evenly along the coils length. The change in current within the windings causes the inductance. Because high frequencies cause different currents in one portion of the winding versus another portion of the winding, the total inductance at high frequency is different than the total inductance at low frequency.

In both cases of Ldc or Les, the actual geometry is ridiculous for a classic tesla coil application, but may be good for some other application. Just thought I'd share the table with the list.

The 2nd table shows what would occur if the wire length was fixed yet the pitch was allowed to increase with h/d. In this case, max inductance had an h/d of 2.5. Easier to look at with data than to actually build.

Take care,
Bart