Resolution measuring a resistor

I'm creating a procedure to measure a resistance decade with the 8508A.

I have some doubt on the resolution for the uncertainty calculation.

Let's assume that I use the following procedure:

9.004  8508         900Z                                        OC D6 N  4W
9.005  MATH         MEM = MEM - Zero
9.006  MEMC   100   900.00Z       0.5%
Where Zero is the zero residual of the decade.
I understand that in this case Metcal will not consider the resolution of the 8508A, but it will take 0.01 as resolution. In this case I will have a much larger uncertainty.
Do there is a method to automatically take the resolution of the 8508A? Or I really have to manually calculate for every step the correct resolution and write the appropriate zeros?


Date Votes
Michael Johnston

By using a resolution of 0.01 ohms in your MEMC nominal field, you've told MET/CAL that that is the resolution you want used in the uncertainty calculation and the reporting. You'll have to alter your Nominal field to be the correct resolution (or use TSET UUT_RES) for it to use the proper resolution.

Andrea Capone

Thank you Michael!
My goal would be to avoid to insert manually the resolution (we could do errors) but to have everything automated.

Do there is maybe another possibility to measure it?

I was thinking at this:

- Calculate the zero (5 measurements).

- Write the value in the certificate. Save the value in the registry.

- Perform a ZERO on the 8508A using the value saved in the registry (I don't know how to do it)

- Measure then the resistance directly with the command "8508 ..." instead of using MEMC.

This way the resolution is calculated automatically.

Is it possible to perform a ZERO (relative) in the 8508A with a variable?

Michael Johnston

If you use the 8508 FSC as your evaluation FSC, it's still going to use whatever resolution you use there as your reported and uncertainty resolution.

If you use registers for ALL of your nominal fields, it might make MET/CAL use the 8508 resolution.

So your code would be:

8508         Z                                        OC D6 N  4W

MATH         MEM = MEM - Zero

MATH         MEM1 = 900

MEMC   100   Z       0.5% 

That might do it.

Nickolas Short
"Good metrology practice should ensure that all high-precision measurements made with
the 8508A are preceded by an input zero of each measurement. However, a non-volatile
input-offset compensation is independently available for the front and rear inputs and will
be used whenever the input zero capability is not in use. The compensation is retained
after power down, is unaffected by clearing the volatile input zeros, and is unaffected by
range zero calibration adjustments. Compensation is available on all functions, and
includes separate compensation of 2-wire and 4-wire measurements in the resistance
Non-volatile input-offset compensation adjustment is available only when the rear panel
calibration keyswitch is in the ENABLE position and the CAL legend is not shown on the
main display.
Adjustment is performed by initiating an input zero operation with the 8508A in this
condition, using the INPUT menu as described in Chapter 3 or using the IEEE 488 input
zeroing commands described in Chapter 4. Appy a high-integrity low-thermal emf 4-wire
short to the input terminals for this operation. Following adjustment for a particular
combination of function, range, and input a compensation is also stored for the
corresponding volatile input zero, and the Zero legend is shown on the main display. The
Zero legend may be removed from the display without affecting the stored non-volatile
compensation by using one of the following operations:
1. Power the instrument down.
2. Press the CLEAR key and make any of the selections from the CLEAR menu.
3. Press the CAL key and entering CAL mode.
4. Use the corresponding IEEE 488 commands."
An excerpt from the manual. Apparently, you can use this to change the "default zero", but still be able to use the soft zero (volatile). If I read that correctly, there is a seperate "default zero" for each measurement function and each range within each measurement function.
Nickolas Short

I don't know why fancy editing isn't working. This are supposed to be three paragraphs with a numbered list inside.

Essentially, buried within that wall of text is a description of a "default zero". If you search the manual for "Non-volatile input offset compensation" you'll find it.

Nathan Ryder

VSET S2 = 0, Will assign no value to the resolution component of your uncertainy budget.

Michael Johnston

I imagine he still wants to calculate uncertainty, just doesn't want to find the full resolution for each test point.

Nathan Ryder

Using that line will still calculate uncertainty just no resolution contributor in the budget.......What is the resolution of a fixed resistor?




Please sign in to leave a comment.