high frequency voltage dividers and measurement / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

high frequency voltage dividers and measurement

previous 1 2

Move Thread

LAN_403

Dr. Slack

Sun Apr 15 2007, 09:37PM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

Of course, there is no such thing as a pure resistive or pure capactive divider, only things that approximate to them (OK, I suppose in the capacitive case, the apprximation is pretty damned good, but then they don't go down to DC). But, you don't have to settle for resistive OR capacitaive.

The problem with a resistive divider is the stray capacitance on the instrument input. Say a typical 'scope is 1Mohm, 15pF. If you have a series 100Mohm resistor for a nominal 100:1, then the bandwidth will be clobbered by the 15pF stray. However, if you put a 100:1 capacitive divider in parallel, by adding 0.15pF in parallel with the 100M resistor, then the resulting divider will be flat from DC to HF, being 100:1 resisitive at DC, crossing over seemlessly to 100:1 at HF. However, getting a 0.15pF capacitor, to stand xkV, adjustable, is pretty tricky, the 100M resistor might well have that amount of stray across it, or more, if so the divder will be over-compoensated before you start. A good plan is to increase the instrument input capacitance to something like 100pF to 1nF, to bring it well out of the stray level, and then the compensating series capacitance is a bit more manageable at several pFs.

Steve Conner

Mon Apr 16 2007, 09:14AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

wrote ...
100:1 resisitive at DC, crossing over seemlessly to 100:1 at HF

Seamlessly, you think. In practice it doesn't and that is how probe manufacturers make their money.

The easiest way to explain it is that the high value resistor used in the "100" arm is of quite a large physical size and has stray capacitance to nearby objects all along its length, as well as the stray capacitance from end to end that you already noted.

So a slightly less naive model of this resistor is a T-network: two resistors in series with a small shunt capacitor to ground. No amount of twiddling of a single capacitor will compensate this. So all sorts of crazy RC networks are needed in the compensation box as the voltage rating of the probe goes up and the divider resistor gets physically bigger.

I think Dan McCauley's home-made probe used a series chain of resistors and capacitors as the divider. I guess the intention was to make the end-to-end capacitance of the divider resistor so big that it swamped the distributed strays to ground I mentioned above. But this would lead to the probe having much higher input capacitance than a good HV probe should.

Read the Sam's RepairFAQ article that was linked for more details.

ragnar

Tue Apr 24 2007, 08:35PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

Well, stumbling for the Nth time through a DEI technical note, I found the following:

"C36, 37, 51 and 53 form an 11:1 capacitive voltage divider providing a
sample of the drain voltage to J4."

And considering the drain supply is 300VDC, they're looking at 950V+ peak, which is in the same ballpark for voltage as my divider. And the technical note was for a 13.56MHz 1kW inverter, so exactly same ballpark for frequency!

Argh, I wish my local store sold 1kV caps. I already have some high-voltage 250pF ceramics I can make use of for the bottom half of the divider.

Dr. Shark

Wed Apr 25 2007, 11:51AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I really think that you would be better of just buying a 1:100 probe. I don't know about the accuracy of the cheaper ones, but in any case there is less room for error than if you brew your own.
Besides, these probes are useful for all sort of things, since they don't load the circuit as much as a 1:10 probe. I end up using mine all the time, but of course I have no idea if what it shows me is really what is going on...

HV Enthusiast

Wed Apr 25 2007, 01:02PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

wrote ...

I think Dan McCauley's home-made probe used a series chain of resistors and capacitors as the divider. I guess the intention was to make the end-to-end capacitance of the divider resistor so big that it swamped the distributed strays to ground I mentioned above. But this would lead to the probe having much higher input capacitance than a good HV probe should.

Actually, the hv divider i designed is how almost all high precision industry high voltage, high bandwidth dividers are constructed. The compensation network is chosen to match the input of the desired measuring device.

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob