I am new to this forum. I have been out of the electronics field for quite some time and with down sizing in the company I work for I have had a project thrown in my lap. It is a simple light tester that will test the resistance of an object. The window of a successful test is to be .1 ohm to 6 ohms. If the object is dead short or above the 6 ohms resistance, then a Bi-color LED will light red. If the objects resistance falls within the .1 ohm to 6 ohm window, then the same Bi-Color LED will light green. This circuit needs to be accurate and will be used in very cold temperature ranges of around 0 degrees F. I am pressed for time on this project and have been looking over comparator circuits. I would appreciate any input or help in this circuit design.

Thank you,
Scott

 

I was thinking "window comparator" before I got to your 2nd to last sentence. http://www.aldinc.com/pdf/cd_23002.0.pdf or http://www.ecelab.com/circuit-window-comp.htm

 

Hi Scott,

More info needed...

1) What is the temperature range of operation?

2) How will you connect to the device being tested? Aligator clips?

3) How to be powered... batteries or 120vac?

4) What accuracy is needed? 1%? 5%? 10%?



Wayne

 

A simple constant current souce coupled to a comparitor would do it. The constant current source will create a voltage, if it is below a certain value then it would read green.

You could even calibrate the constant current source to create a specific range of voltages.

Do you need it to show red if it is below 0.1Ω resistance?

 

You'll need to tell us what the "object" is, will current passing through it damage it?

Also detection settings cannot be exact.

There should also be a grey area range on each end of the detection range where the detection will come up either true or false or indeterminate. e.g. 6.2Ω 6Ω and 5.8Ω

You can specify at 6.2Ω the test must fail and 5.8Ω it must pass, something like that.

 

I suggest to use an op amp as an amplifier and put the resistance you want to measure in the feedback loop. This will output a voltage proportional to the resistance in the feedback loop and thus the measured resistance.

Note that if you want a very high accuracy you need to use a precision op amp.

 

This tester will be used in temperatures ranging from 0 degrees F to 80 degrees F. It will be a handheld device testing metal probes.
It will be 9 volt battery powered. The probes being tested will be directly connected by a harness. It needs to be very accurate. + or - .2 ohms.

I really appreciate all your help!

 

For that temperature range, I'd suggest using Lithium 9V batteries, they are spendy, but stand up to low temperatures far better than alkaline.

I'm confused on the accuracy, +/- 0.2 Ohms is only for the maximum, or is a short only considered < 0.5 ohms in the worse case?

 

I am sorry. +/- .2 on the maximum side. Anything less than .1 would be a fail or a short.

 

Scott,

I've adapted a circuit of a device, I used a long time ago to test lamp filaments, to your needs. It uses the window comparator thingmaker3 refered to. Simulates okay, but you can change (fix) to suit your requirements.

Wayne

 

ifixit you rock! Thank you very much. I will bread board it this weekend and start working on it. I really appreciate your help! I will let you know how it goes.

Thanks again!
Scott

 

Wayne,
Didn't you mean for R5a/R5b to go between R2 and ground, like the attached? (Too bad that LTSpice doesn't come with a model for a pot )

Oh, I threw in R8 as well; a pot that allows for some adjustment of the R low threshold.

Some quick notes:
1) I suggest that the pots be 10-turn (or more) trimpots.
2) In the simulation, LED current was 10-12mA; fine for most LEDs.
3) I took a shortcut in my schematic; the current path to the DUT should be as direct as possible like Wayne shows.
4) The LM311 comparator can sink up to 50mA, while an LM339 may only sink 6mA; so if you're tempted to make substitutions some resistor values may have to be changed.

 

Scott,

You are welcome... good luck and have fun while you build.

SgtWookie,

Correct, the R5s go to ground. I was just trying to show Scott how the sense lines should be kept separate all the way out to the DUT test connector.

I was thinking pots wouldn't be needed if Scott can find 1% resistors, but pots might be easier for the bread board phase.

I have a pot for LTspice, which I made myself. You can set the value with a parameter and control the wiper position with a voltage source (0 to 1V). I can post the details if you need one.

Wayne

 

Hello,

How does the circuit react on a lower getting battery voltage ?
Is there a compensation for the voltage changes ?

Greetings,
Bertus

 

Bertus,

The circuit acts like a bridge and therefore rejects VCC variations.

 

Hello,

OK. So the DUT is one leg of the bridge
and the two legs to the comparators are the other side of the bridge.
(the counterpart is double for low and high).
Thanks for pointing it out.

Greetings,
Bertus

 

Correct, the R5s go to ground. I was just trying to show Scott how the sense lines should be kept separate all the way out to the DUT test connector.

I agree with ifixit regarding the connections to the DUT. The circuit must be connected like the way ifixit have shown in order for the 0.1Ω detection to work properly. If the schematic have been drawn as what SgtWookie proposed, many people will miss those crucial connections.

Do not try to save the extra two wires or else circuit won't work reliably.

I was thinking pots wouldn't be needed if Scott can find 1% resistors, but pots might be easier for the bread board phase.

But here is a factor of x10 in the detection voltage setting. Some selections might be needed in choosing the resistors to fit the correct range even with 1% ones. I don't like the idea of using POT too.