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

Re: Coupling



Original poster: "Malcolm Watts" <m.j.watts@xxxxxxxxxxxx>

Hi Bart, all,
              I'd like to put a plug in for Dr Rzesotarski's MandK
program which was the first I knew of for calculating k. I checked
its outputs against k measurements of several coils including a scale
replica of one variation of Tesla's CS machine. Both measurement and
the program gave a pri-sec k of around 0.6 for that machine, a figure
I seem to recall Tesla actually aiming for. It is of course one of
the "magic" k values.

Malcolm

On 16 Nov 2005, at 11:02, Tesla list wrote:

> Original poster: "Barton B. Anderson" <bartb@xxxxxxxxxxxxxxxx>
>
> Hi Mike,
>
> Yes, I really meant that the primary would be 1" above the bottom of
> the secondary winding. Racing sparks are caused by poor tuning and by
> a coupling coefficient overly tight (energy transfer is too fast
> causing a voltage stressed secondary winding). The vertical position
> of the secondary affects coupling and also affects the proximity of
> primary to secondary voltage standoff (we don't want to arc from
> primary to secondary, which is a different mechanism than coupling).
> You have a 1.25" gap between the secondary and primary when the sec
> bottom turn is even with the pri inner turn. For your coil, that's
> about perfect in my mind. As the primary inner turn is moved
> vertically up and beyond the bottom sec turn, this 1.25" gap is
> maintained. You should not have arc over. However, coupling will
> change as a result of adjusting the primary to secondary vertical
> position.
>
> The reason your coupling was so low, is because the primary was 2.5
> inches below. This is a large distance. As you decrease the inductive
> proximity (bring the coils inductively closer together, coupling will
> increase.
>
> I used Javatc to calc the coupling coefficient. It is more accurate
> than you can measure. There are a few programs out there which perform
> this degree of accuracy. Javatc, Fantc, ACMI, Paul's TSSP software,
> and I believe Antonio's INCA program. ACMI was the first to do it and
> was developed by Paul Nicholson. There were numerous measurement vs.
> program projects to make this happen. Eventually, ACMI could predict
> better than we could measure (due to error within our instruments).
> Javatc includes this portion of ACMI (among other things). What is
> different about Javatc is that it is a "Tesla Coil Design" program.
> It's not just a program for inductance's and frequency's. When coilers
> have questions such as yours, it's really the perfect program to
> answer questions.
>
> I use it because I can input nearly all parameters which I know are
> taken into account. For example, you just mentioned you tap that 10
> turn primary at 8 turns. That information makes a difference, not big
> difference, but there is a change. So now I can easily recalc and see
> what changes.
>
> This tap position changes your current k to 0.081 (slightly lower). I
> can then just enter 0.15 into Javatc's "Desired Coupling" input box
> and re-run the program. Ok, now I see 0.15 k in the output. I can then
> go up to the primary and look at the primary height input values and
> compare to the secondary bottom height value. The difference is 0.935
> or nearly 1". I used a base height value of 20" off the floor for the
> bottom turn of the secondary, thus, your primary was 2.5" below this
> at a height of 17.5". The program now shows the height at 20.935",
> indicating that for a coupling of 0.15, your primary must be 0.935"
> above the secondary bottom height.
>
> It's kind of strange for me. For a long time, we could only measure
> coupling and guess during the design process. Since ACMI and the work
> performed by Paul, Terry, myself, and others, coupling is now one of
> the outputs which carry's the highest accuracy (thanks Paul :-) ).
>
> When Javatc adjust the "desired coupling", it actually equates
> coupling at the original values. It then takes half steps and re-runs
> the program, over and over until a desired degree of accuracy is found
> (which is something like 0.1%). The half steps are changes to the
> primary height positions. The reason is raises the primary instead of
> lowering the secondary is to prevent super-excessive computation time.
> There's no reason you couldn't simply lower the secondary to achieve
> the same result. For the program, it's just faster to compute the
> primary positions.
>
> Hope this helps,
> Take care,
> Bart
>
> Tesla list wrote:
>
> >Original poster: "MIKE HARDY" <MHARDY@xxxxxxxxxx>
> >
> >I'm tapped at about 8 turns. I plan on a bigger topload once I get
> >the bugs out. Do you really mean to put the primary above the bottom
> >turn of sec? Won't this cause racing sparks? How did you calculate
> >the coupling from my measurements? ----- Original Message ----- From:
> >"Tesla list" <tesla@xxxxxxxxxx> To: <tesla@xxxxxxxxxx> Sent: Monday,
> >November 14, 2005 10:36 PM Subject: Re: Coupling
> >
> >
> > > Original poster: "Barton B. Anderson" <bartb@xxxxxxxxxxxxxxxx>
> > >
> > > Hi Mike,
> > >
> > > Your coupling coefficient is currently at 0.083 and could
> > > certainly be adjusted. I usually like to start out at 0.15 k. If
> > > you position your primary 1" above the bottom secondary, this will
> > > get you to nearly 0.15. This is assuming the coil is tapped at 10
> > > turns which indicates a pretty good size topload.
> > >
> > > Take care,
> > > Bart
> > >
> > > Tesla list wrote:
> > >
> > > >Original poster: "MIKE HARDY" <MHARDY@xxxxxxxxxx>
> > > >
> > > >I know that tweeking these coils is how we figure this stuff out,
> > > >but getting educated opinions doesn't hurt. I finally got to fire
> > > >up my new coil.
> > > >
> > > >6" O.D. sec., wound for 32", about 1000 turns
> > > >3/8" cu tube, spaced 3/8" for 10 turns
> > > >sucker gap
> > > >right now 30 nF 35 KV maxwell
> > > >14400V PT running @ 1800 W -125 MA
> > > >
> > > >I'm not getting the performance I expect. Right now about 52"
> > > >strikes. I suspect the coupling may be a little loose. The first
> > > >turn of pri.is 1-1/4" from sec, and first turn of sec. is 2-1/2"
> > > >above the pri. Tomorrow I may try to find some longer nylon
> > > >screws, and raise the primary. So what do you think, am I too
> > > >loose? Should I bother doinng this?
> > > >
> > > >
> > > >
> > > >
> >
> >
> >
> >
> >
>
>
>
>