MEASURE RESISTANCE

[siliconchip]

hi all
I've been trying for months to read 3 resistances in the format I've attached i have used 1 analogue input to try to read in turn each resistance ie 1-2, 1-6 and 2-6 with the latter number in each being grounded I've used relays (relays are 5v 370 ohm coil non latching) driven by transistors both pnp and npn switching, to switch between them and although there is only 0.3 ohm between pole and NC of the relay the reading seems to be around 2 ohms higher or more some times, i have also used 4052 multiplexor but readings are off still, also used a ULN2003 relay driver but this makes the readings +20 ohms, if i measure just one resistor straight from the analogue input and reference resistor the reading is bang on compared to a fluke, my reference resistor is 267.3 ohms and 0.1% tolerance so i know my software is bang on the problem seems to be when running the analogue signal through the pole to N/O or N/C on the relay sometimes if the ground for the transistor is moved i get better results but not always. i really don't know how to proceed with this and run out of switching options i would appreciate any input to progress this further

bob

ps sorry if this is in the wrong section

[siliconchip]

hi
been messing with this again today and noticed that when putting the test resistor directly from the analogue input to ground it reads fine but if a jumper wire is added to the analogue input to the test resistor the readings are off, is there a way to shield the jumper wire from the input to test resistor or add additional components ??

bob

[medelec35]

Hi Bob.
Jumper wires will affect the readings because of the pin to breadboard strip resistance.
This is more so with smaller round pins than the larger square pins that are very common.
I have seen that effect when using an Analogue temperature sensor, smaller pin jumper wires and Sparkfun RedBoard.

[siliconchip]

hi martin
thanks for the reply thats a good point and would make sense as ive noticed different results moving jumpers around and im using small pin ones, where i have not done so so far i will replace all wiring with single core wire cut to length and check the results i will let you know, thanks

bob

[siliconchip]

hi martin,
you were bang on, it seems my jumper wires are poor, with single core cut to size i can now place the test resistor anywhere on the bread board away from the input and its bang on however when connecting to the pole of the relay from the input with the test resistor on the NO pin to ground when the relay throws the reading is out again, i applied power to the relay and checked the resistance between the pole and NC this comes back as 0.3 and when thrown it shows the same between pole and NO yet when connecting the test resistance through it its around 9 ohms higher than it should be i dont understand how as the relay shows 0,3 ohms powered up or not between these 2 points ??

bob