Brand new Bm869s calibration

[mrdave45]

Hi,
I've decided that I need to upgrade the twenty year old maplin dmm I bought as a student. It's starting to get a bit flakey and I'm also fairly sure it's quite inaccurate. This has generally not been too much of a problem. Mostly relative measurement have been good enough. I'm now needing to be able to measure various parameters to a decent accuracy as well as better precision.
I'll probably get my self a decent lcr meter too.

Im pretty much set on a Bm869s. This seems to be a good all round meter and should be leagues better than my current one.

I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

I've also read a few posts about people doing a quick calibration. I would have thought a fairly sophisticated pice of kit is needed to calibrate a meter properly beyond checking against a 5v source or whatever.

Do I need to get a calibrated one, to what level, or should I get one stock and put that money towards an ltz1000 reference.

Im mainly needing voltage and frequency to be super duper out of this thing. I expect to get really good capacitance (tightly matching pf caps) or resistance measurements I want an lcr meter. Probably a de 5000.

I think a high end bench meter is a bit beyond budget and probably overkill for what I need so just a good solid dmm should be just the ticket.

[Fungus]

I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

It has a warranty...

Buy it from a decent supplier, send it back for a replacement if it's out of spec on arrival.

[mrdave45]

Maybe that's answered my question. I'm not going to know of it put of spec unless it's calibrated.

[Fungus]

Maybe that's answered my question. I'm not going to know of it put of spec unless it's calibrated.

I guess if you have nothing to compare it with then you can't know. Do you know anybody else with a decent mater?

Welectron will sell you one with ISO calibration certificate for 70 Euros more.

https://www.welectron.com/Brymen-BM869s-Multimeter_1

That won't tell you if it's gone out of spec after 6 months though, or after an "incident". For that you need to learn about building voltage references and stuff.

Another option is to get yourself a second meter to compare it against. eg. The Aneng 870 has good accuracy specs for the basic measurements and would be a good sanity check that costs half as much as that calibration certificate. Get gold leads too, they're worth it.

(Plus you get a second meter, which you need  )

[mrdave45]

I still have my maplin wg020 as a second meter.
I guess I need the calibration and if I build a reference at least I can tell if the things drifting over time. Although that will presumably only tell me about voltage (maybe current) but won't say anything about frequency.

[Fungus]

I still have my maplin wg020 as a second meter.

Ok, third meter.

Now you'l only be one meter short of being able to measure the efficiency of a voltage converter.

I guess I need the calibration and if I build a reference at least I can tell if the things drifting over time. Although that will presumably only tell me about voltage (maybe current) but won't say anything about frequency.

You can also measure a few resistors the day it arrives then keep them in a safe place. 

The advantage of another meter is that it tells you about voltage, current, resistance, frequency, capacitance, everything.

AFAICT by measuring on various meters, my 0.05% 5V reference is somewhere between 5.001V and 5.002V. My meters show this:

[mrdave45]

https://www.welectron.com/DMMCheck-Plus-Multimeter-Calibration-Reference

What about plumping for one of these instead.
That's already calibrated. I'm guessing that the certificate for that is similar to the one I would get for the dmm but I do have something to double check against throughout the year and check all my other devices with, and presumably I could then send the refenece off every now and then the get calibrated.

[joeqsmith]

...
I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

Seems?  Please post the the links.  I would like to read them for myself.   
 

I've also read a few posts about people doing a quick calibration.
...

It's certainly possible that some people would have access to the level of equipment required to verify the meter.  I assume this is what you are suggesting, not that they are actually aligning them. 

I think a high end bench meter is a bit beyond budget and probably overkill for what I need so just a good solid dmm should be just the ticket.

Only you know what your needs are.  I recently saved an old bench meter from salvage but it required repairs and alignment.   I'm not one of the people you mention who has such equipment to calibrate, let alone align one so it's a bit of an unknown.  Still, for the limited functions it has, that meter will out perform any handheld meter I have looked at.   Most of my equipment is used, required repairs  and over 20 years old.  Doesn't hurt to look around.   Maybe check some local swap meets or salvage yards. 

https://www.eevblog.com/forum/testgear/some-old-school-instruments-showing-how-its-done-(hp-3325a-and-fluke-8506a)/

[mrdave45]

...
I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

Seems?  Please post the the links.  I would like to read them for myself.   

https://www.eevblog.com/forum/testgear/aim-tti-1908-5-12-digit-benchtop-multimeter-(preview)/msg3215950/#msg3215950

Maybe Ive just stumbled on the one post from the one person whos complaining about this thing.

[/quote]
Only you know what your needs are.  I recently saved an old bench meter from salvage but it required repairs and alignment.   I'm not one of the people you mention who has such equipment to calibrate, let alone align one so it's a bit of an unknown.  Still, for the limited functions it has, that meter will out perform any handheld meter I have looked at.   Most of my equipment is used, required repairs  and over 20 years old.  Doesn't hurt to look around.   Maybe check some local swap meets or salvage yards. 

https://www.eevblog.com/forum/testgear/some-old-school-instruments-showing-how-its-done-(hp-3325a-and-fluke-8506a)/
[/quote]

Well, at the moment theres definitely no swap meets going on.
I do scan ebay from time to time for bench meters but they always seem quite expensive. Unless they're in the states, then theres loads, quite cheap, but high postage plus imports plus major hassle if they turn out to be duff.

If someone knows a good source in the uk for such items that would be appreciated.

WHats the general consensus on something like that dmm check plus reference?

https://www.welectron.com/DMMCheck-Plus-Multimeter-Calibration-Reference

But that would be getting one of those over the getting the calibration cert for the bm869s

[joeqsmith]

...
I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

Seems?  Please post the the links.  I would like to read them for myself.   

https://www.eevblog.com/forum/testgear/aim-tti-1908-5-12-digit-benchtop-multimeter-(preview)/msg3215950/#msg3215950

Maybe Ive just stumbled on the one post from the one person whos complaining about this thing.

Only you know what your needs are.  I recently saved an old bench meter from salvage but it required repairs and alignment.   I'm not one of the people you mention who has such equipment to calibrate, let alone align one so it's a bit of an unknown.  Still, for the limited functions it has, that meter will out perform any handheld meter I have looked at.   Most of my equipment is used, required repairs  and over 20 years old.  Doesn't hurt to look around.   Maybe check some local swap meets or salvage yards. 

https://www.eevblog.com/forum/testgear/some-old-school-instruments-showing-how-its-done-(hp-3325a-and-fluke-8506a)/

Well, at the moment theres definitely no swap meets going on.
I do scan ebay from time to time for bench meters but they always seem quite expensive. Unless they're in the states, then theres loads, quite cheap, but high postage plus imports plus major hassle if they turn out to be duff.

If someone knows a good source in the uk for such items that would be appreciated.

WHats the general consensus on something like that dmm check plus reference?

https://www.welectron.com/DMMCheck-Plus-Multimeter-Calibration-Reference

But that would be getting one of those over the getting the calibration cert for the bm869s

When someone tells me "quite a few", I am expecting more than one example.  But as I age, other phases like  "it works" have taken on new meaning.  For example, if I see "refurbished", I know it was maybe cleaned.   "Tested" means it was plugged in.  "Works" means no smoke came out.   Now "quite a few" means one.     

In the context of that thread, they are talking about the 500,000 count mode.  I've mentioned "quite a few" times how I have used this mode to make relative measurements.  The meter is stable enough to be useful for me.   For absolute measurements, I normally use the old HP.    In your case, you may be trying to make absolute measurements with this sort of resolution.  My advice is make sure you understand your requirements before you buy anything!    The best handheld meter I have bought that was somewhat close to what I can achieve with my old HP is the Gossen Ultra.  About $1000 USD.   Still it was unusable out of the box.   It may have been improved but you would need to look into it.   

I have a box that's something like the DMMCheck you are asking about.   I have an APEX reference inside along with several foil resistors.  There are also a few mica, Polystyrene and Wet tant capacitors.  For AC, I have a simple quarts oscillator that I divide down and have direct or AC coupled.    I use that box to detect is a meter is starting to fail, not as a means to calibrate them.   Again, more of a relative comparison.  Still, if you checked that thread I linked on the old 80's Fluke I am working on, you get an idea on the price of these resistors and how tight they are. 

Depending on your own requirements, something like the DMMCheck may be a good sanity test.  I've never played with one.   

[Fungus]

https://www.eevblog.com/forum/testgear/aim-tti-1908-5-12-digit-benchtop-multimeter-(preview)/msg3215950/#msg3215950

Maybe Ive just stumbled on the one post from the one person whos complaining about this thing.

Those meters aren't "far off calibration", they're within spec, they're just outliers near the ends of the bell curve on the production line.

They'll be few and far between but if you're unlucky enough to receive one of those then a calibration certificate or fancy DMM check will only make you unhappy. Best not to get either of those.

[Fungus]

you get an idea on the price of these resistors and how tight they are. 

I bought a used 10k resistor for $7 on eBay. It can read a bit low on a cold day like today (see pic).

I use that box to detect is a meter is starting to fail, not as a means to calibrate them.

Yep. It's fun when a meter/resistor/whatever appears to be perfect but it's really a bonus, not a garantee.

The big problem is ambient temperature. Meters with lots of digits needs an internal heater for their references so they can be calibrated for a known temperature. You simply can't have that in a handheld (do you want a hot meter that chews through batteries?).

[joeqsmith]

you get an idea on the price of these resistors and how tight they are. 

I bought a used 10k resistor for $7 on eBay. It can read a bit low on a cold day like today (see pic).

I use that box to detect is a meter is starting to fail, not as a means to calibrate them.

Yep. It's fun when a meter/resistor/whatever appears to be perfect but it's really a bonus, not a garantee.

The big problem is ambient temperature. Meters with lots of digits needs an internal heater for their references so they can be calibrated for a known temperature. You simply can't have that in a handheld (do you want a hot meter that chews through batteries?).

I had started logging some temperature data while looking at the drift.  I think I figured about 3deg C change in the office but it's been a bit tighter than that over several hours.   I used my HP34401A as a reference to align the Fluke after repairs.   So again, absolute accuracy it a bit of an unknown but I would say in the ballpark.   I ran a 10K ohm on the 40 year old Fluke for about a half day.   You can find the datasheet for that resistor along with the data here:

https://www.eevblog.com/forum/testgear/some-old-school-instruments-showing-how-its-done-(hp-3325a-and-fluke-8506a)/msg3462878/#msg3462878

[IanB]

I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

It may vary from sample to sample, but I was pretty happy with the one I received. See below:

https://www.eevblog.com/forum/metrology/i-received-my-dmmcheck-plus-a-quick-test/msg1972838/#msg1972838

[IanB]

I can't guarantee you will get the same results, but here are a couple of pictures I captured in the past:

[radiolistener]

I don't have voltage standard, but I have ADR290E 2.048 V ± 2 mV reference.
And my BM867S measure it as 2.04828 V just out of the box.
Now, after 3 years it is measured as 2.04826 V.

I didn't performed calibration.

[bdunham7]

I've been searching round the forum about the 869s and it seems there's quite a few comments about these being far off calibration out the box.

I've also read a few posts about people doing a quick calibration. I would have thought a fairly sophisticated pice of kit is needed to calibrate a meter properly beyond checking against a 5v source or whatever.

Do I need to get a calibrated one, to what level, or should I get one stock and put that money towards an ltz1000 reference.

Im mainly needing voltage and frequency to be super duper out of this thing.

I doubt that Brymen is shipping out-of-spec meters, but you shouldn't expect a 500K count meter to be exact to the last digit--it doesn't work that way.  The few mentions I've seen of meters being 'off' is by people with unfounded expectations of some sort or another.

If you need a certain level of precision and accuracy, you need to look at the specs and mathematically understand what is called an 'uncertainty budget'.  "Super duper" is not going to cut it.  For any given measurement, you need to know how far off the actual value is likely to be.  This is actually a statistical issue, but it is commonly boiled down to a single value and a confidence level.  So you might conclude (these are just made up numbers)  that your 5.00000 volt reading actually tells you that the true value is between 4.99930 and 5.00070 with a 95% certainty.  If the confidence level isn't stated, you just have the specs and you can treat those as the limits, but you don't really know how certain you are.  I think of it in terms of counts--how many counts represents the limits of the meters specified uncertainty.  So when I look at a Fluke 8846A displaying 10.00000 volts, I know that the specified uncertainty is +/- 29 counts with a 99% confidence level, as long as I'm within the 1 year calibration period and between 18C and 28C ambient temperature. 

A calibration certificate is just paper, so unless you need paperwork, I doubt it will get you a more accurate meter.

As for self-calibration, you can check and verify it yourself in a limited fashion with a DMMCheck or random 'standards' like that, but this really only tells you when and if your meter is broken.  The actual calibration isn't difficult if you have the standards, but in most cases I think you would just make it worse.  I definitely wouldn't go trying to tweak it so that it reads exactly 5.00000 volts just because it makes you happy--that is a fools errand that will simply leave you thinking you have something that you don't.

[mrdave45]

That's brilliant, thanks all! That looks like its plenty good enough for what I need.
I think Ill just order one. Still need to get an lcr meter and a logic analyser. But i guess thats for another topic.

[mrdave45]

I just wanted to be sure that I could get a meter that I could be reasonably sure that would be accurate to the mV and it looks like this will be without having to resort to the calibration cert.

[joeqsmith]

Personally, I would hold off until you have a better understanding of what you actually need.   You would be amazed how many people write that they bought the wrong product and then want to modify it.      

I just wanted to be sure that I could get a meter that I could be reasonably sure that would be accurate to the mV and it looks like this will be without having to resort to the calibration cert.

I too would like to have a meter that could read accurate to 1mV on a 1000V signal.   

Something a little more realistic, my Fluke standard attached to three meters.  Note the mV reads 0.2, 1.1 and 3.4.   For me, this is good enough.  For you, maybe not.   

[mrdave45]

sorry, i should have specified mV accuracy from 0 to 4.096V range, preferably up to 10V
I would have no need to mV at 1000v range. Usually dealing with tube b+ if its not blown something up then 5% is probably good enough.

[bdunham7]

sorry, i should have specified mV accuracy from 0 to 4.096V range, preferably up to 10V
I would have no need to mV at 1000v range. Usually dealing with tube b+ if its not blown something up then 5% is probably good enough.

So look at the example above and consider the uncertainty budget.  The specs for the 869s on the 50V range are 0.03% of reading + 2 counts, so that comes out to 32 counts if I've mathed rightly.  That's +/- 3.2mV, so it won't be quite 'accurate to the mV'.  And in real life, the meters here are 23 counts apart, a disagreement of 2.3mV, let alone what the absolute error might be.  You should calculate this for yourself on the 5V range and see if that is good enough for you.

[joeqsmith]

Looking at a 1V signal off the Fluke standard.   Details are important.

Also, just a side comment that may not pertain to you personally.  I will get comments from people that will buy the BM869s as their first meter.   I think the last one, the person was sending their meter back because when they picked it up with no leads attached the meter would display a voltage.  ***They were comparing this where I was showing the Gossen shifting with the input terminated to 50ohms or shorted.  ***  This wasn't the first beginner who has sent me such a post.   I tried it with mind (being winter and dry) and I was able to throw up some numbers that were far worse.   I suggested I should perhaps return mine as well.   

I like the meter for type of electronics I play with but it may not be the best choice for a beginner.  I started out learning the basics with a cheap analog meter which I used for several years.  My taste for better equipment came from my hobby evolving.  Had someone given me the BM869s when I started out, it would have been a complete waste.   You may actually have a need for such a meter. 

Video showing my first meter compared with a free one.

[joeqsmith]

Fluke standard attached to my old HP, which was used to align the Fluke meter.   As I would expect, four meters yields four different answers.  I also have no idea what that standard is even good for.   

[mrdave45]

Ok, it looks as though, if i were to measure a 2.048v level, i would get between 2.0473 and 2.0486. Thats going to be much better than my current meter. Which is only 2000 count, much lower accuracy, and only 2 decimal places in the 20v range. Just looked at the specs, thats 1.5% +4d, I could expect to read between 1.97 and 2.12 or there abouts. In any case, a serious upgrade.

Im not (quite) a beginner. I did a degree in computer systems engineering 20 years ago. Although i followed a career in music, since leaving university I got into building hifi and analogue synthesisers.

Im currently developing a digital control system for an analogue poly synth and Im just finding myself needing to get a better idea of what value the components and voltages Im cobbling together actually is.

Id like a better scope, but that's on hold at the moment. But being able to read voltage and frequency in particular, reasonably accurately will go a long way to help me fine tune and figure whats what. Itll be using substantially less than 1v/oct internally, probably nearer 0.4v/oct which in tern means 33mV per semitone. My current scope and meter just aren't really up to this. I want this to work reasonably well before software correction.

I just haven't bought decent test equipment before and wanted some advice over getting one of these devices stock or whether i should get them calibrated. But having seen the photos of several of these next to higher end bench measurements, I fairly sure this will meet my needs.