3
\$\begingroup\$

Would like to have a simple milliohmeter for car wiring checking. Precision is not important, it's not about 3 vs 4 mΩ, it's about 3 vs 30 vs 300 mΩ.

Simple idea is a 1A floating current source (2 alkaline AA batteries and a very basic circuit) and using an ordinary multimeter to measure mΩ as mV in a 4-Wire connection.

An important point is that the circuit survives most possible abuses, specially being connected directly at +/- 12V on the RX terminals. I think it will survive having -12V at Q2 collector but I'm not really sure.

Will have to see if two AA are enough or it needs a third to go 4.5 V.

Do you think this would work? What would you change while keeping it cheap and simple?

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
  • 1
    \$\begingroup\$ Have you fed this through a simulator? If not then please do so. \$\endgroup\$ – Andy aka Apr 6 '18 at 10:33
  • \$\begingroup\$ @Andyaka I'm ashamed to say I've never used a simulator. They didn't exist when I studied electronics and it's something I've always wanted to learn. I see CircuitLab has a very simple one. Will try to see if I make it work \$\endgroup\$ – Toni Homedes i Saun Apr 6 '18 at 10:47
  • 3
    \$\begingroup\$ LTspice might be a better (more capable into the future) simulator, though it's a bit clunky to drive. Put resistors in series with your mV meter, and two antiparallel diodes across it, to limit fault voltage. \$\endgroup\$ – Neil_UK Apr 6 '18 at 10:58
  • \$\begingroup\$ @Neil_UK Thanks for the suggestion about protecting the multimeter too. Had not thought of it. As for LTspice I don't like using non open-source software, though at times you have to do it. I'll give it a look too. \$\endgroup\$ – Toni Homedes i Saun Apr 6 '18 at 11:03
  • 3
    \$\begingroup\$ Yes, LTSpice is free beer, rather than free speech. However it's very widespread and very capable, to the point where it's essentially the de facto standard. But any Spice will do. \$\endgroup\$ – Neil_UK Apr 6 '18 at 11:10
1
\$\begingroup\$

Maybe something like this:

schematic

simulate this circuit – Schematic created using CircuitLab

The fuse is there in case of a short failure in the semiconductors + connection to +/-12V so the battery doesn't have as much tendency to explode in your face.

\$\endgroup\$
  • 1
    \$\begingroup\$ Yes, this is simple and resistant to +-12 V on it’s test terminals! Thanks Spehro. \$\endgroup\$ – Toni Homedes i Saun Apr 6 '18 at 18:51
2
\$\begingroup\$
  1. You won't get a lot of run time from the batteries.
  2. 1 A seems excessive. If you really want to stick to that, I'd use bigger batteries, like D cells.
  3. It seems your overall circuit strategy is to keep the base of Q2 at some adjustable voltage. What you really want to do is keep the drop across R3,R4,R5 at a fixed voltage. Feedback to regulate that directly would require wasting less drive current and get better accuracy.
  4. It seems Q1, R2, and R9 is intended to be a crude shunt regulator, working against R1. That uses a lot of current just to produce the adjustable voltage, and isn't all that good a regulator anyway.

I'd probably start with a better voltage reference, like a zener diode or a 3-pin regulator, then use a opamp to regulate the voltage across the current sense resistance. The opamp would control a FET or a darlington to regulate the current.

\$\endgroup\$
  • \$\begingroup\$ Thanks, Olin. I don't worry about batteries because usage will be 1 or 2 seconds at a time. As for the circuit, I wanted to make it as simple as possible. No problem with 10 or 20 % error. I could use a zener instead of q1/r2/r9 but then I couldn't calibrate it. Adding an opamp is overengineering (for this usage). In fact, I would use only a potentiometer if I could find a 4 Ω / 3 W one but I guess this does not exist or is way too expensive. \$\endgroup\$ – Toni Homedes i Saun Apr 6 '18 at 12:04
2
\$\begingroup\$

Precision is not important, it's not about 3 vs 4 mΩ, it's about 3 vs 30 vs 300 mΩ.

Why not go low-tech then?

Using a multimeter with 200mV range we do need 1A to measure 3-4 mOhm.

I'd use AC like in this answer:

What is the most accurate way to measure internal resistance of a consumer battery?

Basically, grab a transformer from the junk bin, put a resistor and a capacitor in series, plug the primary into mains...

schematic

simulate this circuit – Schematic created using CircuitLab

Now you have a 50-60 Hz AC current injector. Adjust parts values to taste, the cap has to be big enough and withstand the max DC voltage... for example two polarized 2200µF 16V (or more) caps in series back to back. "DUT" is the wiring to be tested. Any decent digital multimeter on 200mV range will not be harmed by 12V on the input.

This hack is thus safe against any abuse, and the capacitor blocks any DC current so nothing can burn. There isn't a lot to burn anyway.

If you have a more sensitive voltmeter you can also lower the test current substantially. Keeping with the spirit of a junk bin parts hack, the sensitivity of a multimeter can be boosted in AC mode by using... another transformer.

If you want it battery-powered, you can use any AC source, like a 555 with output booster transistors to get enough current.

\$\endgroup\$
  • \$\begingroup\$ Thanks, the idea is good, but where I want' to use it has two problems: - I'd don't have an AC outlet where I park my car and: - I don't like the idea of getting mains AC anywhere near a petrol engine. Your circuit would surely be useful for checking home wiring though! \$\endgroup\$ – Toni Homedes i Saun Apr 6 '18 at 11:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.