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

Re: [TCML] NST Measurements



Hi Gary,

Linearity of the voltage ratio from 10V to 120V:

Vin    Vratio
10.0   125.00
20.1   133.33
30.0   137.67
40.2   139.05
50.1   140.32
60.0   139.67
70.1   139.37
80.0   137.13
90.0   137.22
100.2  135.23
110.3  134.63
120.2  136.36

The voltage ratio is rather linear and it's not that high above 125X. However, I realize the leakage of the shunts will affect the voltage ratio, so I know the voltage ratio is not accurate (but close). I can see that at 10V Vin, I happened to match the NST rating. There was enough current in the core without affecting the shunts (or a balance that just occurred). However, on another Franceformer of the same model, it may be 8V or 12V where this occurs. On an Allison, Transco, or whatever, it will likely be very different. That's the problem of recommending a procedure.

In my equations of leakage inductance, self inductances, etc., I used the 125X based on NST specs. Everything simulated perfectly at 125X. The only issue is how to define the turns ratio in a shunted "unknown" case. For NST's, if we know the specs, no problem. MOT's are a different story. I would expect the voltage ratio to be much closer to the turns ratio than for NST's which are "huge" leakage transformers with power factors low in the 40 to 50 range, where MOT's have shunts but are not the huge leakage beasts that NST's are.

Regarding loading, 30mA would drop the voltage very low (not sure how low, but low). NST's are designed to provide overly high voltage (a bit more than the tube really needs to get started) up until the tube conducts, and then the tube load at conduction (volume and gas dependent) will provide a load and the NST will sit there at about 1/2 Voc due to the tube load (their normal application preventing the tube from burning out). Smart design but troubling here <grin>.

The large drop in voltage reduces the power requirements for load measurements. It would still take some hefty resistors, but far lower than 750W. That number is based on 15kV Voc and does not account for the drop in voltage. Loading will dramatically lower that value (thanks to the shunts that are causing all the measurement grief in the first place) and power requirements would lower dramatically as a result.

Best regards,
Bart

Lau, Gary wrote:
Bart wrote:
My measured output at 120.2V input was 16390 which is a voltage ratio of
136 (not 125 as spec'd).

I wonder what the secondary voltage would measure if it were loaded with a 30mA load?  Unfortunately that requires a 500K/750W resistor - not something I have in my junkbox.

How linear is the 136X forward ratio over the 0-120V input range?  Could a much lower wattage 500K load be used at a reduced input voltage to predict the full-power loaded voltage?

Regards, Gary Lau
MA, USA
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla