Re: High Voltage Probe Instability

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Sparktron01@xxxxxxxxxxx 

Hi David

My bet is you're seeing a +TCR effect (heating) of the divider 
resistors.  As an example, 1.0 Megohm 1W Yageo metal-oxide resistors at 
Digikey are listed as 300ppm (0.03%) per deg C.  If you have a 100 deg C 
rise on the divider string (a lot I realize) will equal ~ 3% TCR.  If the 
TCR stacking is + on one side
of the divider and - on the other, a 5% delta would be relatively easy to 
achieve.

On my HV monitoring system, the divider resistors are mounted on a PCB 
immersed in oil as an insulating (and heat sink cooling) medium.  Worse 
case error term seen (even with 50kV thermal soak) was << 1% (at 80% rated 
power dissapation).

Regards
Dave Sharpe, TCBOR/HEAS
Chesterfield, VA. USA


> Original poster: "David Rieben" <drieben@xxxxxxxxxxxxxxx>
>
> Hi all,
>
>
> I have a question that I was hoping that some of you would have
> some insight on. I have been trying to measure the output voltage
> of my x-ray transformers with 1000:1 HV probe that I have. I mea-
> sure the input voltage while carefully controlling it with a variac
> and, at the same time, measure the output of one of the HV term-
> inals to ground and thereby obtain a ratio for the output/input
> voltage so I can have a pretty close approximaton of the output
> of the transformer. Sounds simple enough but the problem that I
> am having is that the high output voltage measured through the
> HV probe keeps slowly but steadily climbing while the input re-
> mains relatively constant. For instance, with an 80 volt input,
> the HV output initially reads around 28,400 volts but after around
> 10 minutes of constant measurement, the HV output has crept
> up past 30,000 volts, at least according to my HV probe/DVM
> combo. Of course, all measurements are taking with no load
> on the transformer. Is my measurement equipment defective or
> is this a normal phenomena? Should I wait for the output mea-
> surement to reach as high as it will go and go with that as an
> accurate measure or should I take the first initial measurement
> as the true measure of the voltage? Could the resistor of the
> 1000:1 probe be heating up and causing the instabiltiy of the
> reading? Any comments are welcome.
>
> David Rieben
>
>