# Homemade high-voltage probe not dividing the voltage by proper amount

I have tried to build a homemade high-voltage probe, as commercial ones are too expensive.

Firstly, I connected a 1 MΩ resistor to a ~1.4 V battery which gave an output voltage of ~0.2 V. By using this formula (where VR = output series voltage, R = resistance of resistor, and VB = voltage of battery):

$$R_M=\frac{V_R\cdot R}{V_B-V_R}$$

I managed to find the internal resistance (~1.4 MΩ) of the multimeter. I then multiplied it by 9 to get the resistance needed to divide the incoming voltage by 10.

Next, I connected the completed resistors in series with my multimeter and a 12 V power supply.

This was the result:

Instead of reading an expected voltage of 1.2 V, it read close to ~3 V (decreasing only by a factor of 4).

My first thought was that there was a bad connection when soldering the resistors, but it was indeed 1.4 MΩ when measured with a multimeter.

I am unsure as to how to move forward, and I am open to any suggestions.

• Did you change range? That type of meter does not have an 'input resistance', it has a 'current for full scale deflection', usually expressed as 'ohms per volt' (ie current). If you want to work with 'meter resistance', then you must keep the same range between calibration and measurement. On what range did you measure 0.2 V, and what range for 3 V? Dec 25, 2022 at 12:20
• Oh, I think that may be it sigh. I measured the 0.2V at the 10V range and the 12V at the 50V range. A further test on the 10V range did manage to divide the voltage correctly. What do you recommend I do, as I intend to be able to change the range for all measurements? Should I get a different type of meter? If so, what kind? Dec 25, 2022 at 12:24
• (What do you recommend I do Tell more about what you are trying to achieve. At more than 1kV, mind safety. (Are you just trying to substitute a/the (missing ?) 1kV probe?) Don't depend on the impedance of a multi-range meter: use an op-amp/instrumentation amp.) Dec 25, 2022 at 15:32
• From what I can see you do not have the voltage probe grounded to the HV source. You would then read the voltage at the junction of the two resistors. You also have a burden load from the meter that needs to be considered as well. Schematics help a a lott
– Gil
Dec 25, 2022 at 19:06
• Commercial high-voltage probes are expensive for a reason. Measuring kV-range voltages safely is not a trivial matter and the resistors are not really the most important and complex part. Dec 25, 2022 at 20:37

This meter claims to be $$\ 20k\Omega\ per\ volt\$$ when measuring DC voltage.
For AC voltages, ohms-per-volt is less (can't make out the value 3k/V? 5k/V?)

However, on its most-sensitive voltage range, it would seem that 0.25V corresponds to 100uA...so on that particular scale the meter would be $$\ 2.5k\Omega\ per\ volt\$$ instead - an anomaly.
I would expect higher DC voltage ranges would more-closely correspond to the 20k/V spec.

A 20k/V DC voltmeter would have an internal resistance that changes for each switch position:

• 0.5V scale: 10k
• 2.5V scale: 50k
• 10V scale : 200k
• 50V scale: 1MEG
• 250V scale: 5MEG
• 1000V: 20MEG (probe?)

You should really read the manual - it is unclear what "probe" means, written above the 1000V scale.
If you really wanted to extend this meter to 10kV for example, you might choose the 250V scale, and add a high-voltage-compatible series resistor of 195 MEG to the 5 MEG resistor already inside the meter. Full-scale would correspond to 10kV DC, so you'd actually read the scale marked "0 - 10" and interpret that as "thousands-of-volts".

Making high-voltage probes? This is unwise.
The probes shipped with the meter are somewhat safe at hundreds of volts on this meter. They should not be used at higher voltages.
You may think that a high-voltage probe merely requires a larger plastic shell to increase path length between your hand and high-voltage probe-end. The surface quality of that path needs attention too, as well as its cleanliness.
I have seen a 5kV probe break down in a user's hand. He barely survived with injuries only because a buddy ran to the breaker panel to shut down power.