[TCML] MOT Measurements

[bartb]

Hi Herwig,

Herwig Roscher wrote:
> So the assumption that the measured(!)
> overvoltage across the MOT's secondary winding was caused by EMF
> could be true. Remember: The MOT's no load output voltage was 2232
> Vrms and its voltage connected to the circuitry of Piranha II rose to
> 3675 Vrms.
Yes. The diodes and caps are in a Greinacher doubler configuration 
charging C1 and C2 on each rectified half cycle. Simulate 1 cap and 
diode then add in the second diode and cap, you'll find the voltage 
increases by about 1000 volts (you should see about 3500V with the two 
330nF caps). This is not due to transformer resonance but general 
multiplication. Of course, in a real world loaded circuit, the voltage 
will never be double. I think in cases of failed transformers in TC 
circuits, the common failure mode is when the secondary winding sees a 
transient voltage, arcs to a neighboring turn, and shorts. If this was 
the case on the original failed transformer, then you should see a 
burned area on the outer secondary winding, which points to transients 
as the catalyst.

> My conclusions are:
> * Use potted MOTs only
> * Use HV diodes with 15 kV PIV only (normally an overkill for 2 kV 
> tranny output)
> * Design the MMC voltage with enough "headroom"
>
> Do you know of other means of protection?
Building a low pass RC filter is about as good as one may get to 
preventing the general case transients, but there will of course always 
be some that get through, even with backup devices such as MOV's, etc.

Take care,
Bart