HP3455A Ohms?

[flittle]

    
So last week the ohms measurement in 4 wire ohms function was spot on but this week it's 10% off.  I hadn't done 4 wire stuff since EE classes but I feel fairly certain I am doing it exactly the same as I was last week.   The photo shows last week this week it shows 9K.   

Do you have any ideas why that might be?

[pelule]

I assume your DUT is a 10k resistor.
It is hard to guess the used wiring out of the picture - you may like to make a rough schematice (using Dave@CAD) to help people to understand the setup.
/PeLuLe

[flittle]

Thanks for the Reply.
Yes the DUT is 10K Standard Resistor.  I was so happy when it came up 10,000.00K last week.  I even took a picture of it.  But this week I was measuring some other resistors and they seemed too far off.  So I got the 10K Standard out again and boom it is 9039.35K or something like that.  I tried some other resistors I had measured last week and they were similary off.

Here is the diagram that I used to measure from the 3455A manual.

[Mr. Scram]

Thanks for the Reply.
Yes the DUT is 10K Standard Resistor.  I was so happy when it came up 10,000.00K last week.  I even took a picture of it.  But this week I was measuring some other resistors and they seemed too far off.  So I got the 10K Standard out again and boom it is 9039.35K or something like that.  I tried some other resistors I had measured last week and they were similary off.

Here is the diagram that I used to measure from the 3455A manual.

Has the temperature shifted significantly?

[flittle]

Has the temperature shifted significantly?

I was wondering about that as it has been significantly warmer outside.  In my basement lair it hasn't changed more than 1 or 2 degrees C though.  I recently built an environmental logger and I have been monitoring the temperature pretty closely.

Would temp change all standard (pictured) and foil resistors 10% though?

[Vgkid]

The old l&n 10k should be around 10ppm/c , thevfoils should be around 5. The 3455 has a pretty low tempco , now if something is wrong it will be worse.

[Cerebus]

I have a vague memory that more than one person has reported problems with the current sources dying on 3455As.

Try measuring the resistance of another meter's current shunts so that you can get a figure for the actual measurement current being used. Check this against the 3455A's nominal test current. The nominal driving circuit for each range is shown on page 1-3 of the service manual.

[flittle]

I have a vague memory that more than one person has reported problems with the current sources dying on 3455As.

Try measuring the resistance of another meter's current shunts so that you can get a figure for the actual measurement current being used. Check this against the 3455A's nominal test current. The nominal driving circuit for each range is shown on page 1-3 of the service manual.

That sounds like a good plan!  I will try that when I get home.

[flittle]

 
It works again.  I was measuring the shunts of my meters and all of the sudden it started working again.
I couldn't make it go back out again. The L&N 10K resistor started measuring 10.000047K which might account for the temp change.

When measuring the current of the shunts I got .7ma for the .1/1/10 scales and 5ua for the 100,1K, 10K scales which agrees with the table 8.0.1 on page 8-66 of the manual I have.

I feel like it had something to do with the GPIB but thanks for all the help.  I bet it comes back soon.

[kj7e]

Dirty relay maybe.  Just needed a cycle to clean up perhaps.

[pigrew]

It could also be cold solder joints.

I just tonight found one on the power switch of a HP DMM. The DMM was working (I opened it to proactively replace capacitors, though all of 1990-era capacitors seem perfectly fine and even had epoxy end-seals....). Note that you don't want to automatically reflow all the resistors since that'd heat them up and likely disturb the calibration of the instrument. I've also found bad joints in 33120A signal generators (output power transistors, failure causing the output to be clipped). I mostly find bad solder joints on large-gauge leads (switches, transformers, power transistors, etc).

Dirty relays does sound likely. In addition to cycling ranges/modes, you may also want to cycle it from front to rear terminals a few times.

[flittle]

I had previously cycled front to back and I feel like it started working because I hit the 10K(K) button which I haven't been using.  Might be a coincidence of course.

Thanks to everyone for your help.  It's a great community.