Help!

Reading a service manual of HP E34401A Digital multimeter and i came across these terms. These are the accuracy required for a standard for calibrating this equipment.

DC Voltage <1/5 dmm 24 hour spec +/- 1 pmm liearity
DC Current <1/5 dmm 24 hour spec
Resistance <1/5 dmm 24 hour spec
AC Voltage <1/5 dmm 24 hour spec
AC Current <1/5 dmm 24 hour spec
Frequency <1/5 dmm 24 hour spec

I would like to know what these means. What is dmm? Can anyone translate these in, like, so and so volts plus or minus so and so percent or something?

Thanks in advance.

cheers!

I haven't heard of the "E34401A". I am familiar with the "34401A". I visited the Agilent website and couldn't find anything on the E34401A.

I will assume that we are speaking of the 34401A. If that is the case, "dmm" is a standard acronym for "Digital Multimeter". That being the case, the DMM spoken of in those specs is likely referring to the 34401A (or possibly E34401A).

Many HP (Agilent) instruments have 24 hour, 90 day and 1 year specs. This gives some versatility to users. The 24 hour specs are often quite a bit better than the 1 year.

I couldn't find the PDF file version of the 34401A service manual. I am not in the lab today, but at a remote office, so I don't have access to the library.

But based on what you wrote in your post, I think they recommend the specs of what ever standard is selected for the calibration be 1/5 that of the 34401A (i.e.: 5 times more accurate). More specifically, 5 times more accurate than the tightest specs of the 34401A.

I know Agilent has some "E" series instruments, so I grant it quite conceivable there is an E34401A I haven't yet heard of. I apologize for my assumptions above if that is the case.

Regardless which is the correct model, the answer above should apply.

Well, lets break it down. First off, DMM stands for Digital Multimeter (the 34401A itself). As for the standard requirements:

The 24 hr specification (what Agilent believes the 34401A should hold within 24 hours of calibration to nominal) for DC Voltage in the 10 VDC range is 0.0015% of reading + 0.0004% of range. Therefore, if my test voltage will be 10 VDC on the 10 VDC range, then:

0.0015% of 10 VDC (input voltage) is 0.00015 VDC
0.0004% of 10 VDC (range) is 0.00004 VDC

Add these together to get a total specification of 0.00019 VDC @ 10 VDC on the 10 VDC range. Therefore, its accuracy @ 24 hours would be 19 PPM (parts per million).

What the standard must supply is 1/5 of 19 PPM, or 3.8 PPM @ 10 VDC. The Fluke 5700A 1 year specification is 7.0 PPM of setting, which is not quite good enough for the 24 hour specification on the 34401A. The Fluke 5720A 1 year specification is 3.5 PPM of setting, which is good enough for the 24 hour specification on the 34401A. But...

Why are we worrying about the 24 hour specification of the 34401A? Will it be calibrated (and nominalized) daily? Not very likely. What you really want is to use 1/5 (or 1/4 even) of the 34401A 1 year specification, in which case the 5700A is fine for the job.

Ryan

Guys,

Thanks.

Oops! Yes, I am referring to 34401A.

I dont have Fluke 5700A but I do have a Wavetek 9100 calibrator and Wavetek 1281 DMM. The Wavetek 9100 is, on most cases, less accurate than the 34401A. Can i use the 1281 DMM in parallel to 9100 so i can use the accuracy of the 1281 instead?

Ok, another one; on the 1281 manual, it says, (i just labeled it from a to c)

(a) range: 100mV
(b) 1 year enhanced accuracy relative to calibration standards: +/- [6ppmR + 0.5ppmFS]
(c) calibration accuracy: 4.5ppm

In computing for the ratio of accuracy of 34401A and 1281, which will i use as 1281's accuracy? b or c? (I think its c but it won't hurt to be sure) What is the diffrerence between range and full scale? Aren't they always equal?

Ryan, no offense meant but i cant seem to find the accuracy of 34401A in the 10 VDC range being 0.0015% of reading + 0.0004% of range. All it says is +/- 190 microV on 24 hour (I am looking at Service Guide of HP34401A Multimeter June 1992 edition 3). Am i somehow wrong in understanding the manual? Please help.

Thanks again!

Cheers!

Lord Ituralde

Lord Ituralde said:

Guys,

Thanks.

Oops! Yes, I am referring to 34401A.

I dont have Fluke 5700A but I do have a Wavetek 9100 calibrator and Wavetek 1281 DMM. The Wavetek 9100 is, on most cases, less accurate than the 34401A. Can i use the 1281 DMM in parallel to 9100 so i can use the accuracy of the 1281 instead?

Actually, the 1281 can run circles around the Fluke 5700A, so go for it.


Ok, another one; on the 1281 manual, it says, (i just labeled it from a to c)

(a) range: 100mV
(b) 1 year enhanced accuracy relative to calibration standards: +/- [6ppmR + 0.5ppmFS]
(c) calibration accuracy: 4.5ppm

In computing for the ratio of accuracy of 34401A and 1281, which will i use as 1281's accuracy? b or c? (I think its c but it won't hurt to be
sure) What is the diffrerence between range and full scale? Aren't they always equal?

First off, ppmR is not Range, it is Reading. 1 ppm of 100 mV is 0.0001 mV, while 1 ppm of 50 mV (measured on the same range) is 0.00005 mV.

Now, when computing the 1281 uncertainty, (b) and (c) are additive. The basic accuracy of the 1281 is 6 ppmR + 0.5 ppmFS. But, you must add to that the possible error at the time of calibration that cannot be accounted for due to standards, etc, aka calibration accuracy. Therefore, your accuracy would be:

± [6 ppmR + 0.5 ppmFS + 4.5 ppm], or in the case of 100 mV input on the 100 mV range, (Here is the tricky part on the 1281) 11.5 ppm.

WHAT??? 6+0.5+4.5 is NOT 11.5!!!!

But, if you read the 1281 uncertainty fine print, you find out that FS (full scale) is actually the published full scale (100 mV) times 2 (the 1281 will overrange to 200 mV, and well, they use their overrange in the FS calculation, because 0.5 ppmFS looks much better than 1 ppmFS).

Ryan, no offense meant but i cant seem to find the accuracy of 34401A in the 10 VDC range being 0.0015% of reading + 0.0004% of range. All it says is +/- 190 microV on 24 hour (I am looking at Service Guide of HP34401A Multimeter June 1992 edition 3). Am i somehow wrong in understanding the manual? Please help.

Thanks again!

Cheers!

Lord Ituralde

Well, you know that fancy thing in all of the ISO standards that says your manuals must be up to date? Well, the latest dates on the 34401A manuals seems to be in the 1998 and later timeframe. Luckily, Agilent is pretty good about posting their manuals online. For the simple spec sheet, try http://cp.literature.agilent.com/litweb/pdf/5968-0162EN.pdf . For the latest User's Manual, try http://cp.literature.agilent.com/litweb/pdf/34401-90004.pdf . That should help you out a bit. Oh, and for the 1281 specs, along with the tricky fine print, try http://www.fluke.com/Download/Calibrators/1281_spex.pdf .

Ryan

Lord Ituralde said:
I dont have Fluke 5700A but I do have a Wavetek 9100 calibrator and Wavetek 1281 DMM. The Wavetek 9100 is, on most cases, less accurate than the 34401A. Can i use the 1281 DMM in parallel to 9100 so i can use the accuracy of the 1281 instead?


IF you have software that supports it, then you can do exactly this. This is a built-in function of Wavetek's Portocal-II software, which unfortunately vanished when the company was acquired by Fluke. (Does anyone know if MetCal will do this?) If you have other measurement instrument control software you may be able to create a routine that has the same effect.

The basic process is

Generate correction values by characterising the 9100 using the 1281. This process is the same as calibrating the 9100, but no adjustments are made. The 9100 is the UUT and the 1281 is the measurement standard. The software stores the measurement results and uncertainty for future use.
Use the 9100 to calibrate the 34401A (or other instrument) using the corrections. The 9100 reverts to being a measurement standard. The uncertainty of the result is calculated from
[/list=a] [list]
1281 accuracy specification
1281 calibration uncertainty
measurement uncertainty from the characterization process (a)
9100 short-term stability
measurement uncertainty of the calibration measurement now being made (b)

(If you have the Portocal-II software, look in chapter 14 if the manual.)

As usual, you will have to validate the process before using it. We have not had to actually use this function in my lab because we also have a Wavetek 4808 standard as well as a couple of 9100's. That means I cannot give you an objective idea of the amount of improvement to expect -- does anyone else have that information?


... cant seem to find the accuracy of 34401A in the 10 VDC range being 0.0015% of reading + 0.0004% of range. All it says is +/- 190 microV on 24 hour (I am looking at Service Guide of HP34401A Multimeter June 1992 edition 3).

I have found that in a lot of cases, the SERVICE manual does not always say things in the same way as the specification published in the USER manual. This is such a case. If you do the math, at a 10.0000 V input on the 10 V range, the 24-hour accuracy is ... 190 microvolts! BTW, I am looking at the specification sheet Ryan gave a link to - it is dated 23 January 2001.

thanks guys!

we have a portocal II software but unfortunately, installing it in our pc is beyond me. :( I dont even think it would work because its slot wont fit in P3 and newer motherboards.

thanks a bunch!

cheers!

This is a built-in function of Wavetek's Portocal-II software, which unfortunately vanished when the company was acquired by Fluke. (Does anyone know if MetCal will do this?)


Yes, met/cal can do this... There is an example procedure that fluke provides that characterises a 5500A calibrator with an 8508A multimeter, then uses it to calibrate a 34401A....