From: jl@glen--canyon.com   
      
   On Sun, 9 Nov 2025 18:34:12 +1100, Bill Sloman    
   wrote:   
      
   >On 9/11/2025 5:35 am, john larkin wrote:   
   >> On Sun, 9 Nov 2025 04:26:06 +1100, Bill Sloman    
   >> wrote:   
   >>   
   >>> On 9/11/2025 2:25 am, john larkin wrote:   
   >>>> On Fri, 07 Nov 2025 19:23:31 GMT, Jan Panteltje    
   >>>> wrote:   
   >>>>   
   >>>>>> liz@poppyrecords.invalid.invalid (Liz Tuddenham)wrote:   
   >>>>>>> Bill Sloman  wrote:   
   >>>>>>   
   >>>>>>> On 7/11/2025 10:41 am, john larkin wrote:   
   >>>>>> [...]>   
   >>>>>>>> A real inductor is a nightmare. Especially a long solenoid. Every turn   
   >>>>>>>> inductively couples to every other turn with all possible coupling   
   >>>>>>>> coefficients. Distributed capacitances will be similarly complex.   
   >>>>>>>   
   >>>>>>> This is ignorant nonsense.   
   >>>>>>   
   >>>>>> I suspect the ignorance is yours.   
   >>>>>   
   >>>>> John is largely right   
   >>>>> I have wound big RF coils (MHz) for my shortwave transmitter for high   
   power,   
   >>>>> several 100 W, on just a simple ceramic former, since the sixties.   
   >>>>> The mil sets do the same, that is where I got the idea.,   
   >>>>>   
   >>>>> Same for small cores,   
   >>>>>   
   >>>>>   
   >>>>>> I worked for a company that built its reputation on the R.F. inductors   
   >>>>>> it designed; these factors were among the many problems they tackled.   
   >>>>>   
   >>>>> I wind almost everything I need myself.   
   >>>>   
   >>>> I designed and wound this myself!   
   >>>   
   >>> As much as you design anything.   
   >>   
   >> I don't think that you approve of engineering.   
   >   
   >Oh, I do, but I don't think that what you do qualifies as engineering.   
   >Desperate improvisation comes closer to the mark.   
      
   I wasn't desperate, it was fun.   
      
   >   
   >>> The thermal contact pad under the coil   
   >>> that keep it from getting too hot was presumably optimised by the same   
   >>> kind of well-thought-out rigorous process.   
   >>   
   >> Thermal conductivity of the pad and an estimate of contact area. I   
   >> wrote a program that helps with such thermal calculations.   
   >   
   >You need to write a program to calculate that?   
      
   It speeds things up and handles mixed units. Do you still use pencil   
   and paper and a slide rule?   
      
   >   
   >> There a lot of vias through the board to pipe heat from the inductor   
   >> down into the baseplate. That adds theta too.   
   >   
   >They just conduct heat. Heat pipes rely on evaporation and vapour   
   >transport, and can have a much lower thermal resistance, if you have the   
   >space to squeeze one in.   
   >   
   >> One of those gigantic multiphysics programs would be better, but I   
   >> don't have one. It wouldn't be worth the effort to estimate coil   
   >> temperature a few degrees better.   
   >>   
   >> The gap-pads are at best around 7 w/m-K, and one can doubt their data   
   >> sheets too. Pad theta is a strong function of compression, even more   
   >> 3D fun. One could do math for months, or test things in an hour.   
   >   
   >Then the manufacturer makes a slight change in the way they make their   
   >thermal pads, and you have to do it all again.   
   >   
   >> But the Pockels Cell driver works and is a few per cent of the volume   
   >> of coventional MHz-KV designs.   
   >>   
   >> How would you design such an inductor? Or such a pulser?   
   >   
   >I'd find out what the Pockels cell needed, and go from there. Sometimes   
   >you can reorganise the parts that are being driven to make them easier   
   >to drive.   
      
   Of course I had requrements. It's a cavity dumper of sorts.   
      
   >   
   >>>> https://highlandtechnology.com/Product/T850   
   >>>>   
   >>>> We have a carefully selected red Sharpie pen that is the official coil   
   >>>> former.   
   >>>   
   >>> Measuring the internal diameter with vernier calipers could give you a   
   >>> more easily archived reference.   
   >>>   
   >>>> That thing takes 48V from a wall wart and makes 1400 volt pulses at   
   >>>> MHz rates into a Pockels Cell.   
   >>>   
   >>> Somebody more image-conscious might have machined (or 3D printed) a   
   >>> metal helix to more precisely specified dimensions. Something a bit   
   >>> bigger might have been easier to cool.   
   >>   
   >> Size affects inductance!   
   >   
   >So does number of turns.   
   >   
   >> The customer hasn't complained.   
   >   
   >So it is good enough. Not necessarily as good as it might have been.   
      
   Design a better kilovolt pulser. Show us.   
      
   >Engineering  has been defined as doing for one dollar what any damned   
   >fool could do for two.   
   >   
   >Low volume niche markets rarely justify that kind of optimisation.   
      
   My main object is to amuse myself. Often enough, people want to buy   
   the results.   
      
      
   John Larkin   
   Highland Tech Glen Canyon Design Center   
   Lunatic Fringe Electronics   
      
   --- SoupGate-DOS v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)