From: jl@glen--canyon.com   
      
   On Sat, 24 Jan 2026 01:56:46 +1100, Bill Sloman    
   wrote:   
      
   >On 23/01/2026 6:44 am, john larkin wrote:   
   >> On Thu, 22 Jan 2026 22:45:56 +1100, Bill Sloman    
   >> wrote:   
   >>   
   >>> On 22/01/2026 5:48 am, john larkin wrote:   
   >>>> On Sun, 18 Jan 2026 15:29:11 +1100, Bill Sloman    
   >>>> wrote:   
   >>>>   
   >>>>> On 18/01/2026 10:33 am, john larkin wrote:   
   >>>>>> On Sat, 17 Jan 2026 15:58:01 +1100, Bill Sloman    
   >>>>>> wrote:   
   >>>>>>   
   >>>>>>> On 17/01/2026 4:19 am, john larkin wrote:   
   >>>>>>>> On Sat, 17 Jan 2026 03:59:00 +1100, Bill Sloman    
   >>>>>>>> wrote:   
   >>>>>>>>   
   >>>>>>>>> On 16/01/2026 11:01 am, john larkin wrote:   
   >>>>>>>>>> On Thu, 15 Jan 2026 23:01:38 +0000, John R Walliker   
   >>>>>>>>>>  wrote:   
   >>>>>>>>>>   
   >>>>>>>>>>> On 15/01/2026 18:15, john larkin wrote:   
   >>>>>>>>>>>> On Thu, 15 Jan 2026 17:51:59 +0000, liz@poppyrecord   
   .invalid.invalid   
   >>>>>>>>>>>> (Liz Tuddenham) wrote:   
   >>>>>>>>>>>>   
   >>>>>>>>>>>>> john larkin  wrote:   
   >>>>>>>>>>>>>   
   >>>>>>>>>>>>>> On Thu, 15 Jan 2026 15:18:31 +0000, liz@poppyreco   
   ds.invalid.invalid   
   >>>>>>>>>>>>>> (Liz Tuddenham) wrote:   
   >>>>>>>>>>>>>>   
   >>>>>>>>>>>>>>> john larkin  wrote:   
   >>>>>   
   >>>>>    
   >>>>>   
   >>>>>> I'll reveal the secret mathematics:   
   >>>>>   
   >>>>> The secret you reveal isn't in the mathematics. It's the thermal   
   >>>>> resistance from the dissipating element to ambient   
   >>>>   
   >>>> No; average power is low.   
   >>>>   
   >>>> What matters is microsecond-range heat capacity in the resistive   
   >>>> element. The ceramic substrate may help a little.   
   >>>   
   >>> The impulse load ratings for the Vishay thin films that I posted   
   >>> suggested that for them the thermal mass of the substrate stops   
   >>> mattering below about 300usec.   
   >>>   
   >>> You may have to worry about the skin effect as well, which restricts   
   >>> very high frequency components to the edge of any track. For higher   
   >>> resistance metal film resistors the tracks tend to be pretty narrow anyway.   
   >>>   
   >>>>>> 180 watts at 0.1% duty cycle is 0.180 watts.   
   >>>>>>   
   >>>>>> 15c divided by 0.18 is 83 watts per degree C.   
   >>>>>>   
   >>>>>> Please keep this confidential.   
   >>>>>   
   >>>>> Why? It's on the data sheet of whatever resistor you happen to be using,   
   >>>>> which you haven't specified. You may not realise this.   
   >>>>   
   >>>> The data sheet of a cheap thickfilm resistor does not characterize it   
   >>>> for pulsing at 1000x rated power.   
   >>>   
   >>> So you probably shouldn't use them for that job.   
   >>   
   >> Should be fine at 250 volts per resistor.   
   >>   
   >>>>   
   >>>> Or 2000x, which I'm running now.   
   >>>   
   >>> And where you are starting to see changes in resistance with time.   
   >>   
   >> After a day of pulsing at 500v, 500 watts, the resistance has dropped   
   >> about 0.1%.   
   >>   
   >> It takes me about a minute a day to run this experiment, so I'll let   
   >> it run some more.   
   >   
   >0.1% in a day is worryingly high.   
      
   It's within the tempco range of a cheap thickfilm.   
      
   If the resistor is being damaged, we may have fused a few weak spots,   
   and things may stabilize. So I'll let it keep running.   
      
   >   
   >> I might try the thinfilm 1206 just for fun.   
   >   
   >A prudent engineer would try a part that was specified for some kind of   
   >short pulse use, rather than one which wasn't at all.   
      
   A sensible engineer will test parts that are available. Maybe you can   
   find a small surface-mount 499 ohm resistor that is specified to   
   tolerate kilowatt pulses.   
      
   Being prudent isn't fun, or a good way to invent things. It's good to   
   get crazy and reckless once in a while, destroy things.   
      
   The final design stages of a sellable product should be very prudent   
   and reviewed by careful and prissy people.   
      
   >   
   >> I should have bought a higher voltage version of this power supply.   
   >> The mosfet is good for 1500 volts.   
   >   
   >All sorts of nasty stuff starts happening at high voltages. Cambridge   
   >Instruments didn't like photomultipliers where the photocathode was more   
   >that 1kV away from ground. The glass widow of the photomultiplier tube   
   >and the glass window of the sample chamber both started conducting   
   >current by ion migration at that sort of voltage, with a little   
   >electroluminescence, which was an inconvenient source of noise.   
      
   Numbers under a kv or so aren't too scary on a surface-mount PCB. Much   
   higher ones are.   
      
      
   John Larkin   
   Highland Tech Glen Canyon Design Center   
   Lunatic Fringe Electronics   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)