The most accurate portable multimeter on the market

[acinho]

I'm calibrating various lab instruments in the field. The procedure that I follow recommends 5 1/2 digit multimeter with a >10 MOhm input impedance.
Gossen Metrahit 30M satisfies these requirements but it's no longer available so I use Fluke 8846A. It's a bench multimeter and cumbersome to carry around and so I'm looking for something more portable.
500000 count Brymen 869s caught my eye but I'm not sure it's accurate enough, 0.02% + 20d in the 500000 count mode.
Are there more accurate alternatives that are for sale? It doesn't have to be a handheld multimeter, just small enough that I can carry it in a backpack.

[Fungus]

500000 count Brymen 869s caught my eye but I'm not sure it's accurate enough, 0.02% + 20d in the 500000 count mode.

That's the outer limit accuracy across the full temperature range of the meter. You should find it's a lot better than that at standard room temperature.

People have tested this and found the error to be more like 2-3 digits:

https://www.eevblog.com/forum/testgear/bymen-bm869-measurement-of-dcv-linearity/msg384188/


https://www.eevblog.com/forum/testgear/brymen-bm869-and-bm867-batch-calibration-check/

The real problem is that that for more digits you need a temperature controlled voltage reference (ie. a small oven inside the meter) to get more digits. You won't find it one in a battery powered meter because it would drain the batteries in no time.

If you're calibrating at certain fixed voltages you could also compare the 500,000 count Brymen to your Fluke and know/compensate the error (if there is one).

[acinho]

I plan on doing side by side comparison in the field.

[alm]

Is characterizing the error yourself over the range of environmental conditions and time yourself acceptable in your process? Or do you need to rely on manufacturer specifications?

Could you measure relative to a more accurate standard? Like make a differential voltmeter out of a PDVS2 + a multimeter as null meter? Of course this would be more cumbersome.

[Fungus]

I plan on doing side by side comparison in the field.

Good idea. Take a Brymen and your Fluke for a while, see if they agree.

(and be sure to post about it here!)

[Grandchuck]

Is characterizing the error yourself over the range of environmental conditions and time yourself acceptable in your process? Or do you need to rely on manufacturer specifications?

Could you measure relative to a more accurate standard? Like make a differential voltmeter out of a PDVS2 + a multimeter as null meter? Of course this would be more cumbersome.

alm's suggestion is a good one.  The PDVS2 could be an excellent solution.

[EEVblog]

I wonder what the lowest power 6.5 digit bench meter is and if it can be converted to battery power?

[coromonadalix]

Maybe  some agilent keysight i think  usb meter   U2741A , NI usb modular meter model 4065,  damn the prices have gone  crazy

You have some NI  pcia meter cards too, not sure they go 6.5 digits ??

I would not use any "fake" 500000 counts  dmm

you could use some metex / chauvin arnoux mtx3292-3293  100,000 real count dmm's ? or some Gossen meters at 300,000 counts    not the obscure 30m   you do have some 28 - 29s  floating around ?? but the acquisition time is slower like 2 meas / seconds


the easiest convertible bench meter  would be the L4411A, if i recall the psu in it is very simple, but to make it work   you have to control it thru usb, lan  gpib

[Fungus]

the easiest convertible bench meter  would be the L4411A

I think OPs problem is that bench meters are "cumbersome to carry around", not the lack of mains plugs.

[coromonadalix]

yep   

The Gossen METRA HIT ULTRA BT   is 300k count, but sold at least 1k$$$   

cumbersomeness  ... mmmm

But if you build a good road case or a metallic suit case  ?? for the meter and put your stuff in it, it could be a very good compromise, unless you carry a Fluke calibrator

I think you'll have to do some eventually ... you can not have both worlds easily ....  you have to be realistic ...  calibrating stuff at 6.5 digits, what about the ambient temperatures, humidity, warm up time  ...

my 2 cents

[Fungus]

The Gossen METRA HIT ULTRA BT   is 300k count, but sold at least 1k$$$   

Brymen is 500k count and four times less money. 

[J-R]

The harsh reality is that yes, the BM869s is mostly a fake 500,000 count.  It can only be used for DCV/DCmV.  On top of that, unless you calibrate it, even the last digit in 50,000 count mode can be rough.  It's quite a bit more sensitive to time from power on as well as temperature compared to the higher-end handheld Fluke DMMs I have.

Some other notes:
- If you calibrate it yourself in the conditions in which you'll use it, then 500,000 count mode can be somewhat more usable.
- Since the calibration point is at the top of the range, if you only use it up to about 50% of the range then it's quite a bit more accurate.
- It's a pain to try to get this DMM to a stable point in daily use due to the short battery life.  The rubber case has to be removed to reach the 9V battery.

(So if you needed to measure 1V or 10V for example, and could do the manual calibration using a bench meter before taking the field measurements, you may have some hope.)

Quick glance at my collection and the old-school Keithley 197A seems like a possible solution here.