Interchange / ratio Error Issue on Precision Bridge /. resistance bridge

[email_removed_by_admin]

Hi everyone,

I work with precision resistance measurement instruments, specifically the Measurements International 6020T Bridge, which is based on a DC comparator (DCC) bridge. The 6020T is used for precise measurement of resistance ratios by comparing the unknown resistor to a known standard through a turn ratio measurement.

I’ve encountered an issue where the interchange error in quick mode is around 0.1 ppm in one unit. I tested another similar unit, and its interchange error is around -0.07 ppm, which is also higher than expected.

Here are my questions:

What could cause interchange error to consistently remain high in the 6020T units, specifically in quick mode?
Could this be related to internal components (e.g., magnetic core, capacitors) or external factors (e.g., lead connections, environmental conditions)?
What are the best troubleshooting steps to minimize interchange error and achieve more consistent results with the 6020T Bridge?
Any insights would be greatly appreciated

[Kleinstein]

An error of 0.1 ppm is not really that large of an error. If the resistors are not very high grade it could be just a temperature change on swapping them.

There are several possible errors. It depends a bot on the resistance range which effects are more relevant. With small resistors it is more about thermal EMF and maybe magnetic coupling or an issue with the wiring towards the DCC core. With relatively large resistors there could be issues from leakage, capacitance and maybe amplifier input currents.
So at what resistors is the measurement and does it chance with resistance.

[email_removed_by_admin]

the place i work at they make very precise units and 0.1 ppm is not big but their standard is 0.02 or around that, and what do you mean by wiring around DCC? can you explain that in detail please. it would be very helpful for me

thank you for your reply reallly appericiate man

[Kleinstein]

With low resistance / relatively high current the DCC may use just a few turns. With only a few turns it gets difficult to distribute the turns evenly around the core and this can make the DCC somewhat sensitive to external magnetic fields.  The wires going towards the core also produce some magnetic field and the exact position of the wires makes a difference. The difference between units could also be from the transformer part (e.g. how even wth turns are or mechanical stress at the core). Different from a DC background field the field direction from the wires would change with a polarity swap. So the turn over part may not suppress this error. If the polarity is swapped with the help of relays one could also get interference from there windings, also with latching relays. So good magnetic shielding may be needed there.

An issue with the DCC is only one possible problem, there could also be an issue in the voltage measurement (e.g. thermal EMF at relays, residual ripple from the DCC and different inductance for the resistors).

To find such a tiny error in a rather complex instrument is really difficult. Even the developers have a hard time finding them. Hunting tiny non-ideal effects can take a long time and detailed knowledge of the instrument and is more work in progress, never really finished.
It may help to have a few more points, like a test with different resistor values, test currents and maybe ratios / turns settings. Another point is how much is a real systematic error and how much is more random (e.g. thermal effect, cables used, EMI). Knowing the error under different conditions could be a first step for corrections or knowing the limits of the instrument, even if one does not find the actual cause.

[email_removed_by_admin]

i get it what you are trying to say and it fair but the thing is these units are build to measure from 1ohm to 10k, and we measure ratios for each one 1:1 and 10:10 like that,few days ago i tried different firmware which only uses 1024 turns for both normal and quick mode before it was using 1024 for quick and 4096 for normal and error dropped from 0.1 ppm to around o.o3/4 ppm consistent throught all the values untll it reaches to measure 1k and 10k it moves back to 0.1 on 1k and 2pppm around 10k. so im trying to firgure why does that now ? the unit is pretty well designed and company dont wanna sell the unit with 0.1ppm error. thedcc is covered inside the sell and there is not residue on pcb boards either.

also when we measure both 1k and 10k around 3MA and 0.3mA respectively, so  when i measured 1k with 0.3mA error dropped from 0.1 to 0.03ppm same when i reduce the current from 10k Error dropped

also i have question about another units its simple its temparature controlled air bath for resistor another unit my compant makes, i have to check the stability of the unit by measureing temparature at certain point from 15 to 40 and it i take around 400 measurement to check the stabliltiy to check that units is staying at the same temprature and not moving, the limit for stablitiy is 0.05c and its pretty good the stablity untll it reached the 35 and 40c the stablity is 0.1 at both 35 and 40c, i tried to increase and decrease voltage of heating ciruit  the which is 49V i tried to move 0.2V both side but it still stablity at both that point is around 0.1,  do you have any idea about air baths ?


thank you for the answer man, can we connect on insta or somehwere ?

[Overspeed]

Hello

An interesting document on the subject oil bath air bath and metrology setup

Regards
OS