Products > Test Equipment
3457A Resistance Drift
(1/1)
TaylorD93:
Hi everyone,

I have HP 3457A on the shelf above my bench, it is rarely used to be honest and when it is, its either for finer DCV readings or as a glorified 4 wire Resistance meter. I have a 34401A which is at bench level as it takes up less space and does the job most of the time.

Recently i have noticed that the 3457A seems to be suddenly off by a few millohms, with the leads shorted they now read between -10 and -3 Milliohms. I know this isnt much, but i have swapped the Kelvin probe leads for another set and it has continued to show this odd offset when nulling the leads.

Ive never used the "auto zero" function nor the offset compensation and with the DMM being up out of the way, its not like ive knocked it and disabled any of these.

Also I believe it is on its original battery, and has had a bit less use in the past month than before. So im a bit concerned, could this be the signs of bad things to come? I have seen sudden step changes in calibration offsets suddenly reveal themselves as component failure.
Is it worth taking a punt on changing the battery now and just getting it recalibrated? I have had a quote for a calibration house to perform a full calibration on the instrument and it has come back at £175. So i dont really want to pay out and then find it fails later and i have to get it recalibrated (although that can be the case with older TE). If there is this little (but noticable) error, im skeptical i could fault find the circuit to find a "drifted" component, a hard fault would be much easier.

It has been a bit warm in recent weeks in the UK and i live in a top floor flat so ambient temps arent exactly constant, but i didnt think it would cause that much of an error/drift within the ohm scale? or has my Failure of GCSE Statistics bitten me in the backside again?

I hope someone can put my mind at rest, thanks in advance
Taylor
chickenHeadKnob:
Before doing anything rash make sure the insides are clean of dust bunnies and then leave it powered up for 3 weeks to see if it is a moisture problem. With sparing use and changing humidity you can get drift. A battery failure is a binary thing, the power on self-test either gets a valid checksum or it barfs with an error message.
Kleinstein:
With some care the battery can be changes without loosing the calibration. AFAIK the data can be read via GPIB and with some extra effort also written back.

If accuracy is relevant one should really use the auto zero mode. Even the non AZ mode would need an occasional zero. AFAIK the 4 wire ohms mode should kind of imply a kind of auto zero as the 2 sense lines are read one at a time.

An offset in the 4 wire ohms mode is a bit unusual. Even with normal component drift (e.g. from humidity) this should not happen. For a rarely used meter, relay contacts can be an issue. Switching between
front and rear for a few times could be a good idea.
I have some doubt that simply a recalibration would fix the problem. This is more like a problem that should be fixed before sending the meter to calibration, as this may indicate a defect of some kind.

srb1954:

--- Quote from: TaylorD93 on June 06, 2022, 10:18:31 pm ---Hi everyone,

I have HP 3457A on the shelf above my bench, it is rarely used to be honest and when it is, its either for finer DCV readings or as a glorified 4 wire Resistance meter. I have a 34401A which is at bench level as it takes up less space and does the job most of the time.

Recently i have noticed that the 3457A seems to be suddenly off by a few millohms, with the leads shorted they now read between -10 and -3 Milliohms. I know this isnt much, but i have swapped the Kelvin probe leads for another set and it has continued to show this odd offset when nulling the leads.

Ive never used the "auto zero" function nor the offset compensation and with the DMM being up out of the way, its not like ive knocked it and disabled any of these.

Also I believe it is on its original battery, and has had a bit less use in the past month than before. So im a bit concerned, could this be the signs of bad things to come? I have seen sudden step changes in calibration offsets suddenly reveal themselves as component failure.
Is it worth taking a punt on changing the battery now and just getting it recalibrated? I have had a quote for a calibration house to perform a full calibration on the instrument and it has come back at £175. So i dont really want to pay out and then find it fails later and i have to get it recalibrated (although that can be the case with older TE). If there is this little (but noticable) error, im skeptical i could fault find the circuit to find a "drifted" component, a hard fault would be much easier.

It has been a bit warm in recent weeks in the UK and i live in a top floor flat so ambient temps arent exactly constant, but i didnt think it would cause that much of an error/drift within the ohm scale? or has my Failure of GCSE Statistics bitten me in the backside again?

I hope someone can put my mind at rest, thanks in advance
Taylor

--- End quote ---
It may a thermal EMF problem although the error seems a little on the high side for this. It is unlikely that the lack of auto-zero is the cause of the problem as the power-on state of the 3457 is for the auto-zero to be enabled.

Reset the 3457A so it is in a known default state, take a resistance measurement and note the readings. Enable the offset compensation through the menu command OCOMP 1 and take another resistance reading to see if there is any significant difference. Any significant difference could indicate a thermal EMF is present and you need to check the condition of all connections throughout. Some oxidation or corrosion on a connector could result in the generation of significant thermal EMFs with heating of the leads through handling or even through rapid ambient temperature variations. After handling the leads always allow plenty of time for temperature differences between connections to equalise before taking a measurement.

To really verify the performance you need to have a proper 4W short made of low thermal EMF materials. Failing that a thick copper wire connecting all terminals - just be sure to thoroughly clean the the copper wire first as a copper oxide to copper junction has a very high thermal EMF potential.
TaylorD93:

--- Quote from: srb1954 on June 07, 2022, 06:48:53 am ---
--- Quote from: TaylorD93 on June 06, 2022, 10:18:31 pm ---Hi everyone,

I have HP 3457A on the shelf above my bench, it is rarely used to be honest and when it is, its either for finer DCV readings or as a glorified 4 wire Resistance meter. I have a 34401A which is at bench level as it takes up less space and does the job most of the time.

Recently i have noticed that the 3457A seems to be suddenly off by a few millohms, with the leads shorted they now read between -10 and -3 Milliohms. I know this isnt much, but i have swapped the Kelvin probe leads for another set and it has continued to show this odd offset when nulling the leads.

Ive never used the "auto zero" function nor the offset compensation and with the DMM being up out of the way, its not like ive knocked it and disabled any of these.

Also I believe it is on its original battery, and has had a bit less use in the past month than before. So im a bit concerned, could this be the signs of bad things to come? I have seen sudden step changes in calibration offsets suddenly reveal themselves as component failure.
Is it worth taking a punt on changing the battery now and just getting it recalibrated? I have had a quote for a calibration house to perform a full calibration on the instrument and it has come back at £175. So i dont really want to pay out and then find it fails later and i have to get it recalibrated (although that can be the case with older TE). If there is this little (but noticable) error, im skeptical i could fault find the circuit to find a "drifted" component, a hard fault would be much easier.

It has been a bit warm in recent weeks in the UK and i live in a top floor flat so ambient temps arent exactly constant, but i didnt think it would cause that much of an error/drift within the ohm scale? or has my Failure of GCSE Statistics bitten me in the backside again?

I hope someone can put my mind at rest, thanks in advance
Taylor

--- End quote ---
It may a thermal EMF problem although the error seems a little on the high side for this. It is unlikely that the lack of auto-zero is the cause of the problem as the power-on state of the 3457 is for the auto-zero to be enabled.

Reset the 3457A so it is in a known default state, take a resistance measurement and note the readings. Enable the offset compensation through the menu command OCOMP 1 and take another resistance reading to see if there is any significant difference. Any significant difference could indicate a thermal EMF is present and you need to check the condition of all connections throughout. Some oxidation or corrosion on a connector could result in the generation of significant thermal EMFs with heating of the leads through handling or even through rapid ambient temperature variations. After handling the leads always allow plenty of time for temperature differences between connections to equalise before taking a measurement.

To really verify the performance you need to have a proper 4W short made of low thermal EMF materials. Failing that a thick copper wire connecting all terminals - just be sure to thoroughly clean the the copper wire first as a copper oxide to copper junction has a very high thermal EMF potential.

--- End quote ---


Looks like I owe you a beer or two :)

Cleaned up the contacts,  they had some green grease on them, washed it off with IPA and a brush, now within +/-1mOhm.

OK it does still flick to negative ohms but on 10 NPLC doesn't stray any further than -1mOhm.

On 0.1 or 0.05 NPLC it is obviously picking up noise etc as it becomes noisy to the extent of +/- 30mOhm but this makes sense to me as it looses the resolution due to shorter sample duration etc

Thanks 😊
Navigation
Message Index
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod