DIY multimeter adjustment based on calibration report

[NaxFM]

I need to do accurate measurements with my 34401A so i sent it for calibration. The only calibration laboratory that would do calibration for hobbysts (at a reasonable price) does not perform any adjustment, so it is up to me to make my dmm more accurate. I will only recieve a calibration report stating the read value with respect to the applied reference, and I already know that in the 10V range (the range I care the most) it reads about 200uV high.
I already checked the zero value of every range with a proper low thermal emf short before sending it for calibration, and everything reads zero as expected, this should take care for offset error, which is basially nothing.
My question is: let's say that i recieve back my dmm with a cal certificate stating that it reads 200uV high at full range, can I enter in the adjustment menu and adjust the internel coefficient so that it reads 200uV lower than before at full scale? Is it a valid method to adjust a DMM or do I risk to make the accuracy worse?

For example: I have a 10V reference (assume it is exactly 10V) which reads 10.00020V and knowing from the calibration that the DMM reads 200uV high i can enter in the adjustment menu to tell it it should read exactly 10V instead.
The idea is that since the zero value is accurate (checked with a short) I only need to adjust the full scale reading, and i know from the calibration report by how much i should adjust it.

[Kleinstein]

Doing the adjustment has a chance to reduce the accuracy, as the DMM itself shows more or less random variations. To the initial and adjusted reading could have an additional shift of a few ppms. Rounding on the numbers entered also adds a little to the error and uncertainty.
Similar to the DMM the reference voltage can also show slight variations - even of not very likely there is no really good way to check to better and about 1 ppm.

If the error is not too large (200 µV on the 10 V range are still only 20 ppm) one could keep the DMM the way it is and correct the readings numerically for those few cases where it really matters.

[2N3055]

This is second time you asked this question.
Do you expect different answer this time ?

[NaxFM]

This is second time you asked this question.
Do you expect different answer this time ?

I didn't really get a clear answer last time, only tangents.
This time i made my question easier to understand and immediately Kleinstein gave me a precise and clear answer.

So, I guess asking twice sometimes might help

It's not that I didn't like the other answers, I always appreciate when someone answer my question.
But last time it was my fault and I wasn't very clear regarding the calibration, and since this is my only option (No lab will perform adjustment) I wanted to be very clear this time and remove all my remaining doubts.

[Caliaxy]

Not sure about DIY calibration, but a 10.00020V reading on the 10V range seems within the specs of your meter: +/-(0.002% of measured value + 0.0005% of range)@90 days = +/- 0.00025V (>0.0002V).

[2N3055]

OK fair enough.
In addition to as always very competent Kleinstein post:

I wouldn't mess with the meter in such a way.

If I had another well calibrated meter (preferably better one) and stable voltage source I might be tempted to adjust meter that is out of whack to show same as the good one.
But only if readings are really out there.

As Kleinstein said,  for instance, 200uV on 10 V range is only 20 ppm off and well in specs.

For your meter, there is a performance verification procedure in Service manual. There is a table where all limits are listed in plain language. If meter comes from calibration with reading that are inside those limits , do not touch it.

Also, on calibration paper, they will state measurement uncertainties of their lab/calibration process. Those will have to be taken into account. So they say your meter was 50 ppm high. Part of that includes their error, so your meter might be better or worse than this.

That is why you didn't get clear answer last time. It is not that simple. And that is why people don't do what you proposed. You could as well make it better or worse. And you would need to send it for another calibration to be sure..

[NaxFM]

Not sure about DIY calibration, but a 10.00020V reading on the 10V range seems within the specs of your meter: +/-(0.002% of measured value + 0.0005% of range)@90 days = +/- 0.00025V (>0.0002V).

You are right, but unfortunaly a 34401a is all I have at home, and don't have the money to buy anything better. I wanted to get the best accuracy possible.
When you are a poor student hobbyst you try everything to get the best possible results from whast you have

[NaxFM]

OK fair enough.
In addition to as always very competent Kleinstein post:

Thank you, really.
Unfortunaly I'm passionate about high accuracy measurements but I don't have the means or the budget to do it properly.
I'm a voltnut with no money or proper equipment, and yet I'm trying to work with voltage references and precision resistor decade boxes.
Buying my 34401a and getting it calibrated was a big stretch for my finances, that is why I'm desperately trying to get the best out of it.

[bdunham7]

You are right, but unfortunaly a 34401a is all I have at home, and don't have the money to buy anything better. I wanted to get the best accuracy possible.

You should start by understanding what is possible and what is really not practical.  There is a reason the tolerances in meters are often referred to as 'uncertainties'.  You can dial your meter in all you want relative to a calibration report or some standard you may have, but you're probably not be improving anything and you just won't realize it because you have no way of knowing.  And without knowing the details of your economy calibration lab, deciding how to improve on things would be sheer speculation.

First, a properly working 34401A is a pretty accurate meter just by being within its stated specifications.  Precision and accuracy become insanely expensive beyond that point.  Trying to do better than what it is specified for is not a casual endeavor that is likely to turn out well for someone with limited resources.  I'd be curious to know exactly why you want to do this.

Second, in my limited experience with them the primary driver of uncertainty in these seems to be temperature.  You can plot out the readings over a 4 hour warmup period in a room with only moderate temperature control (a heated house room, for example) and you'll typically see 10-15ppm change over that period and 2-3 ppm changes with room temperature cycling.  Low TC systems are difficult and expensive to make and 35ppm basic accuracy is actually quite an achievement.  Simply dialing your meter to some exact number under one specific condition--even if you had a perfect standard--isn't going to result in a perfectly accurate meter going forward as conditions change.

Last, before you do anything you need to know what the capabilities and uncertainties of the calibration lab are.  If this is a budget calibration, there's a good chance that the results you get may not be all that helpful.

[Grandchuck]

Try to find someone you can work with that owns a trusted voltage source or meter or both.  Here, we have a calibration club that rotates two good voltage references (and other goodies) through the members on its list.  Good luck, have fun, and try not to get too frustrated (warning: voltnuttery without frustration is not feasible).

[free_electron]

You would have to recreate the exact environment under which it was measured. room temperature to begin with... and then all the equipment used with exact same specs. (which is an impossibility. you would need the actual machinery the lab uses. not just the same model ... the same physical machine they have)

As for adjusting a 34401 : that is done through software. there are no adjustment points inside the machine. you unlock the cal menu ( code is most likely 034401 unless it was changed) and follow the procedure. you need the correct standards to do it.

[J-R]

What 10V reference do you have?

Obtaining calibration data for your 34401A is a good idea, I would go for it.

[coromonadalix]

Same  answer,  i know you're a hobbyist,  BUT dont think like that if you want accuracy and repetable results

Save money and send it to a proper facility to see with a cal report what the meter is and what it will became after the calibration

my 2 cents

[Fungus]

My question is: let's say that i recieve back my dmm with a cal certificate stating that it reads 200uV high at full range, can I enter in the adjustment menu and adjust the internel coefficient so that it reads 200uV lower than before at full scale?

No. It doesn't work that way.

Read the service manual to see how calibration/adjustment is done. You need all sorts of special equipment.