Why does resistor measure different every time?

[paul8f]

Trying to measure the value of a SMD resistor on two identical boards. (The boards contain a mixture of power, analogue and digital circuitry.)

The resistor in question is a 10k. On the 1st board is measures as expected (10K), but on the 2nd board it measures about 6k. Sometimes the polarity of the DMM probes make a difference, and sometimes not.

The reading on the DMM doesn't ramp, so I'm guessing there are no caps getting charged from the probes...

Any ideas?

[Kleinstein]

An in circuit resistor measurement is not always reliable. There are sometimes other parts (resistors - semiconductors) in parallel that can effect the measurement. It is a principle problem - not a fault of the meter. One may get slightly better results when using a lower test current, so that the test voltages is low (e.g. < 200 mV), as than semiconductor tend to have less effect.
The only reliable way is when you know the circuit and know there is a capacitor /  diode that is not conducting.
The other reliable result is when one find a resistor reading too high - the circuit can only shift the resistance down.

[tszaboo]

You have an active circuit, like a diode, or an input of an amplifier, or something like that, that provides secondary path for the current.
Look at the schematic, and look for any active circuit that is connected to the same node as your 10K resistor.
The orientation matters because there is
Also (dont take offense, I am not aware of your knowledge) you should only measure circuits that are not energized. So unplug it before measuring and take out batteries.

[vk6zgo]

Trying to measure the value of a SMD resistor on two identical boards. (The boards contain a mixture of power, analogue and digital circuitry.)

The resistor in question is a 10k. On the 1st board is measures as expected (10K), but on the 2nd board it measures about 6k. Sometimes the polarity of the DMM probes make a difference, and sometimes not.

The reading on the DMM doesn't ramp, so I'm guessing there are no caps getting charged from the probes...

Any ideas?

It is not uncommon to measure a resistor in circuit & find the reading is less than the marked value of the component, as they often have other parallel resistors, etc.

Comparing identical boards is a commonly used means of fault finding.
If the resistance measured is not identical, it sounds like one board is faulty.

[Miti]

ICT machines use something called "Ohm guard" to reduce the effect of parallel components when measuring resistances in circuits.
A quick search found this
https://www.edn.com/ict-resistance-testing/

[Gyro]

Some bench DMMs too.

[Zero999]

There are two things which can be done to minimise the effect of PN junctions:

1) Use as lower test voltage as possible, preferably lower than a diode drop. Lots of modern multimeters do this. I've just tested my Amprobe 37XR with a cheap meter with a 1M input impedance and it reads 40mV. I think the higher resistance ranges use a lower test current/open circuit voltage because a known good 1N4148 reads open circuit.

2) Measure the resistance in both directions and record the higher value, which will be the one when the diode reverse biased.

If there's a small voltage on the circuit, which doesn't cause the meter to read open circuit, then measure the resistance with the probes connected in opposite polarities and take the average of the two readings. Another option, assuming it can be safely short circuited, is to measure the open circuit voltage and the current, when short circuited, and calcualte the resistance using Ohm's law: R = V/R. Obviously it's normally a bad idea to measure the resistance when there's a voltage present, but sometimes it's a necessary evil.

[coppercone2]

a bad side of SMD, with through hole you can pull all the 2 leaded components out (one lead only) with only a soldering iron to quickly diagnose the circuit without disturbing it, taking pictures or doing anything difficult, just heat one side and yank with tweezers, or if you are measuring a component that has stuff around it, you can disconnect the circuit elements which cause a problem for the in circuit measurement. It even works on nasty PCB that is screwed to a buncha shit with bus bars and all sorts of problems. It stays in place so you can even put it back on the shelf and you are left in the same place as you were before you started without taking any notes, pictures or other logs. And someone else could go to the job immediately after with zero turn instruction.

when you are done you just pull the solder out with a sucker, bend it back into place and resolder and your time to repair becomes significantly reduced.

of coarse because we are gluing rice to everything this suddenly becomes a major fucking operation

[Conrad Hoffman]

Any residual voltage will confuse the meter. You may have to short out any larger value caps.

[Kleinstein]

a bad side of SMD, with through hole you can pull all the 2 leaded components out (one lead only) with only a soldering iron to quickly diagnose the circuit without disturbing it, taking pictures or doing anything difficult, just heat one side and yank with tweezers, or if you are measuring a component that has stuff around it, you can disconnect the circuit elements which cause a problem for the in circuit measurement. It even works on nasty PCB that is screwed to a buncha shit with bus bars and all sorts of problems. It stays in place so you can even put it back on the shelf and you are left in the same place as you were before you started without taking any notes, pictures or other logs. And someone else could go to the job immediately after with zero turn instruction.

when you are done you just pull the solder out with a sucker, bend it back into place and resolder and your time to repair becomes significantly reduced.

of coarse because we are gluing rice to everything this suddenly becomes a major fucking operation

Unsoldering THT part is still quite some stress to the parts and PCB. Though it helps that THT was usually still SnPb solder.
With the wires bend before soldering it can get quite hard to unsolder.

Small SMD parts are easy to remove with hot tweezers / double irons or a special tip.  Just don't drop the little hot caps (resistors are not that bad as one often can just use a new one if they are labled.
Even just a single iron is not so hard with leaded solder. Still less stress to the parts and PCB than with THT parts.

The main problem is if part are glued    - than it gets really hard.

[Miti]

Some bench DMMs too.

Really? Which one?

[coppercone2]

a bad side of SMD, with through hole you can pull all the 2 leaded components out (one lead only) with only a soldering iron to quickly diagnose the circuit without disturbing it, taking pictures or doing anything difficult, just heat one side and yank with tweezers, or if you are measuring a component that has stuff around it, you can disconnect the circuit elements which cause a problem for the in circuit measurement. It even works on nasty PCB that is screwed to a buncha shit with bus bars and all sorts of problems. It stays in place so you can even put it back on the shelf and you are left in the same place as you were before you started without taking any notes, pictures or other logs. And someone else could go to the job immediately after with zero turn instruction.

when you are done you just pull the solder out with a sucker, bend it back into place and resolder and your time to repair becomes significantly reduced.

of coarse because we are gluing rice to everything this suddenly becomes a major fucking operation

Unsoldering THT part is still quite some stress to the parts and PCB. Though it helps that THT was usually still SnPb solder.
With the wires bend before soldering it can get quite hard to unsolder.

Small SMD parts are easy to remove with hot tweezers / double irons or a special tip.  Just don't drop the little hot caps (resistors are not that bad as one often can just use a new one if they are labled.
Even just a single iron is not so hard with leaded solder. Still less stress to the parts and PCB than with THT parts.

The main problem is if part are glued    - than it gets really hard.

hot tweezers things can get lost.. put backwards, etc. its alot better then hot air though

of course with glue its a cave man chizzle job, get to pick components for summary execution

repair procedure: get obsolete A5 PCB with 15 revision choices