HP 3468B resistance measurements

[johngalt47]

I have a weird issue with this meter. I leave it on all the time, hoping for some consistency. I usually measure a batch of around 15 resistors at a time. Let’s say that it has been a couple of hours since using the meter. The first couple of resistors in a batch display stable readings immediately. The one after that require several seconds to display a stable reading. For example, a 2400 ohm resistor might read 2406.45 as soon as it is connected with my Kelvin leads. Then it starts counting down and may reach 2406.37 before it stops. So why would the first couple in the batch be display a value that does not change.

Since I leave the meter on I would think that the temperatures have stabilized.

Bythe way, I usually measure Vishay/Dale RN55 or RN60 metal film, high quality ones.

[kj7e]

So your talking about the measured value dropping 0.08 on the 3K ohm scale?  The 24 hour spec for the 3K range is 0.0035%, which comes to 0.084 Ohms of 2400.  You may be asking for more than stability the 3468 is capable of providing.

Slight drift can be normal while taking a measurement, the constant current source could drift slightly, perhaps the 3k range current sense resistor is warming and the tempco being seen, maybe the DUT is warming (or cooling for that matter) and your seeing the DUT's tempco?  The 3468 is a good meter but not a particularly high accuracy meter and you seem to be within spec.

[johngalt47]

My problem is that the first two of any batch DON’T drift while the remaining ones do. I could understand it if I left the meter off and did not let it warm up when measuring the first two. I could understand thermal drift, but not inconsistency.

[kj7e]

Is it always the same exact resistors that show drift?  Or does the drift show after so much time making measurements, that is, regardless of the resistor, but after say 1 min of testing the drift shows.

[2N3055]

Maybe cables / clips heating from hands.. ??

[johngalt47]

I only use the clips to pick up the resistors. I will have to do some more testing, but it seems that maybe some values may be more solid than others. I was thinking along the lines of maybe a slightly cold solder joint.