Handheld meter robustness testing

[MiroS]

It can be exacerbated by holding the meter in your hands.

.. or by rotating multimeter with your body like 90 degrees left or right holding the meter in hand.
That looks like accelerometer/divergence sensor
That may be static charge effect (?). Nothing like that on FK289 , or F87V. 

It seems to be a real problem, handheld multimeter by design is 'moving/rotating' .

[MiroS]

https://youtu.be/_SD24ujP3vg
I found this on YT

[EEVblog]

Again, during the last test where the DC had been removed, the meter was set to VFD.  The bargraph will be disabled, the filter will be active causing a bit of an voltage difference for the higher frequency content waveforms and yes, it only displays two places past the decimal point.  We can see this on page 24. 

Is it worth repeating this test with the VFD disabled?   I could also just run two meter and remove the DSO, maybe slow it down a bit further to make it easier to track.   We could also increase the number of waveforms.  This test requires little effort to setup so if you feel it's worth it, let me know. 

I am still not sure what the comment was about UNI-T not having a problem.  I could toss the 87V into the mix as well if this is what you want to see.  It just gets more difficult for me anyway, to follow what is going on with so many devices active at once.

Hi Joe,

If it is not big effort and you're willing to do it, sure, it would be nice.
It would be nice to see limits of this, and maybe with what waveforms it will happen. I doubt autoranging will have problem with standard waveforms...
Maybe see it through full range (6V, 60V, 600V) ... Definitely check whether it is VFD related.

...

There was a comment on YT that BM869s has a problem with 560kOhm resistance measurement if it is performed close to power cord.  I would propose to add this to the list of tests if you will decide to make next tests with BM786.

I can confirm that this is a thing. It starts at around 520k and persists up to 660k or so. It can be exacerbated by holding the meter in your hands.

I ran cables from a resistance decade to my meter, with a few cm being parallel to my PC power lead.

I was unable to duplicate the effect with my 87V, 289, or UT139C.

I'm guessing it's some kind of low hysteresis between the autorange selection between 600k and 6M, as it basically starts flipping between them at a rate of two or three Hz.

Yep, confirmed. I didn't even need mains to do it. At one point I could just stand there holding the cables.
I goofed the Mohm range test, should have been 5.6M, but I tried that and it's fine, so definitely only 560k.
I'll start a new thread for this issue.

[EEVblog]

Please post about he BM869 issue here so we don't pollute this thread.
https://www.eevblog.com/forum/testgear/brymen-bm869-resistance-quirk/

[joeqsmith]

I also tried the resistance test on a few ranges and with the old and new firmware.  It seems to be this one range.  Hysteresis and switch points are similar to my Fluke 189.   I tried a few other Brymen products.  It seems to be related to the 869s.  The YT comment suggested all Brymen products suffer from this problem.  What other products exhibit this problem?  I assume the people who are posting here have contacted Brymen already.  What was their response?

[gnavigator1007]

Just to clarify, Dave's is the 869 non S model. The video linked by MiroS shows the same issue on the 869s

[MiroS]

The video linked by MiroS shows the same issue on the 869s

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.
From Dave message - resisatnce sims to work, but who knows which firmware is in Dave multimeter, that may be old one, Joe could be one of first external testers for corrected/imporved  firmware.

I think Brymen may be still not aware of issues, I do not think that anyone reported this to them for 869s.

[2N3055]

The video linked by MiroS shows the same issue on the 869s

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.
From Dave message - resisatnce sims to work, but who knows which firmware is in Dave multimeter, that may be old one, Joe could be one of first external testers for corrected/imporved  firmware.

I think Brymen may be still not aware of issues, I do not think that anyone reported this to them for 869s.

1. First problem is one with resistor. It has to do with instrument having high impedance and on the edge of autoranging ... It does seem a bit nervous, but any meter can be destabilised this way if you try hard enough. This is something I would like if Brymen could make better, but would not change the meter because of it.
2. Second problem I don't really understand what is he doing..
3. 3rd problem is not a problem at all and has nothing to to with VFD,  Crest mode has less resolution, in order to be fast enough. It's in a manual.

[Chalton_trc]

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.

It would be nice to try the new meter. I agree. Has it been tested with Fluke cables?

[Chalton_trc]

I don't know if Joe will have information from Brymen ... In Welectron they say that the BM786 will be available on 10.01.2021. If they are fixing the issues now, will there be a final version in stores by then? Does anyone know anything about this?

[2N3055]

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.

It would be nice to try the new meter. I agree. Has it been tested with Fluke cables?

That is not my quote, sorry..

WTF it has to do with the cables? Brymen cables are better quality than some Fluke cables, and none of them are shielded. They will all pick up same electric/magnetic field.

[Fungus]

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.

It would be nice to try the new meter. I agree. Has it been tested with Fluke cables?

What's wrong with Brymen cables? I've got some and they're awesome.

[Fungus]

1. First problem is one with resistor. It has to do with instrument having high impedance and on the edge of autoranging ... It does seem a bit nervous, but any meter can be destabilised this way if you try hard enough.

Yep. I bet it will disappear if you switch to manual ranging.

[2N3055]

1. First problem is one with resistor. It has to do with instrument having high impedance and on the edge of autoranging ... It does seem a bit nervous, but any meter can be destabilised this way if you try hard enough.

Yep. I bet it will disappear if you switch to manual ranging.

Of course it will, it is autoranging problem. You have value right at the  edge of range, very high impedance on input and you're injecting maybe a volt of interference. That combined with fast frontend, makes a jittery autoranging.
We are all annoyed when things are not perfect, but it is not a real problem.


Also Dave's test with BM786 is invalid .. Of course that 60000 count meter won't have problem with 560 kOhm. It might have problem with 650 kOhm, because that is where it will have switchover and similarily high impedance.....

Later, I will try this with some meters, right on their switchover point, and high resistance mode.. I think many of those will have same problem, unless they have very slow autoranging, and very slow front end and sampling...

[Fungus]

Also Dave's test with BM786 is invalid .. Of course that 60000 count meter won't have problem with 560 kOhm. It might have problem with 650 kOhm, because that is where it will have switchover and similarily high impedance.....

LOL! Yeah, good point.

[joeqsmith]

The video linked by MiroS shows the same issue on the 869s

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.
From Dave message - resisatnce sims to work, but who knows which firmware is in Dave multimeter, that may be old one, Joe could be one of first external testers for corrected/imporved  firmware.

I think Brymen may be still not aware of issues, I do not think that anyone reported this to them for 869s.

I did try the latest updates with the 786 (at its switch point) and saw no problems.  I also looked at a few other Brymen products as it was mentioned "There is a noise problem with brymen multimeters." which I read as ALL Brymen multimeters.  The only one I saw a problem with were the two 869s I have (two versions of firmware).   I had gone so far as to twist the test lead around a long line cord to my lamp.  Even then it's the only meter I saw a problem with.  Really odd as it seemed to be the only range effected.  The Fluke 189 has close to the same switch points, hysteresis, test currents as the 869s and even with the twisted wires, I saw no effect.   No doubt I could inject enough noise to screw them up. 

The OP has not responded on if they contacted Brymen or not.  I find it odd that anyone would post about it in social media but not contact the company who designs it.  I've found them to be very responsive but they can't address what they are not aware of.   Depending how far back it goes, it could have been possibly rolled into the version that extended back light.   I'll write them next week if no one else responds.

[joeqsmith]

1. First problem is one with resistor. It has to do with instrument having high impedance and on the edge of autoranging ... It does seem a bit nervous, but any meter can be destabilised this way if you try hard enough.

Yep. I bet it will disappear if you switch to manual ranging.

Of course it will, it is autoranging problem. You have value right at the  edge of range, very high impedance on input and you're injecting maybe a volt of interference. That combined with fast frontend, makes a jittery autoranging.  We are all annoyed when things are not perfect, but it is not a real problem.

It does seem like we should have a standard way to test it.   Maybe inject a signal (transformer couple in series with the test resistor)  and find out where the 869s exhibits the problem, then add a 50% margin to that?   Get the coupling and lead length out of the picture.   Always test on the highest range switch point? 

I didn't check the UNI-T 181A but that may be another one to look at.  Someone could also try it with their 121GW with the latest firmware.   Maybe that super fast HIOKI? 

[dcac]

The video linked by MiroS shows the same issue on the 869s

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.
From Dave message - resisatnce sims to work, but who knows which firmware is in Dave multimeter, that may be old one, Joe could be one of first external testers for corrected/imporved  firmware.

I think Brymen may be still not aware of issues, I do not think that anyone reported this to them for 869s.

1. First problem is one with resistor. It has to do with instrument having high impedance and on the edge of autoranging ... It does seem a bit nervous, but any meter can be destabilised this way if you try hard enough. This is something I would like if Brymen could make better, but would not change the meter because of it.
2. Second problem I don't really understand what is he doing..
3. 3rd problem is not a problem at all and has nothing to to with VFD,  Crest mode has less resolution, in order to be fast enough. It's in a manual.

My guess on the second “problem” -  he’s confused why the frequency reading shows zero Hz when he manually select higher ranges.

I don’t think it’s mentioned in the BM896 manual but minimum (RMS) AC volts required for the frequency readout at 50Hz seems to be:

Range:
500.00mV = 65mV
5.0000V    = 0.28V
50.000V    = 2.8V
500.00V    = 28V 
1000.0V    = 280V

So he’s first measuring 12.2V AC which is high enough for the 50.000V range, but not for the 500.00V or 1000.0V ranges.

Then he measures 116V AC which is high enough for the 500.00V range, but not for the 1000.0V range.


And just for comparison with the 121GW minimum (RMS) AC volts required for the frequency readout at 50Hz seems to be:

Range:
500.00mV = 15mV
5.0000V   = 0.15V
50.000V   = 1.5V
500.00V   = 15V 
600.0V     = 15V


So to an inexperienced user the BM896 might (possible) seem to be broken when it is in fact just different - or in this case less sensitive on the frequency display - than another multimeter.

[MiroS]

So to an inexperienced user the BM896 might (possible) seem to be broken when it is in fact just different - or in this case less sensitive on the frequency display - than another multimeter.

No , no ... look starting at 4:04 - 117AC on input ...

The same BUG with 235V AC 50Hz, no question for me , that is a real bug, easy to reproduce. I think that one is also unknown to Brymen and worth of testing with new Brymen multimeter.

[2N3055]

So to an inexperienced user the BM896 might (possible) seem to be broken when it is in fact just different - or in this case less sensitive on the frequency display - than another multimeter.

No , no ... look starting at 4:04 - 117AC on input ...

The same BUG with 235V AC 50Hz, no question for me , that is a real bug.

So he’s first measuring 12.2V AC which is high enough for the 50.000V range, but not for the 500.00V or 1000.0V ranges.

Then he measures 116V AC which is high enough for the 500.00V range, but not for the 1000.0V range.

And guess what, 235 V is still less than 280V needed to measure frequency in 1000V range..

So NOT a bug. Bug is unexpected behaviour. This is not unexpected behaviour. It is clearly documented in specifications. Fact is BM869S has frequency measurement that is not very sensitive, and you need to be in right range for it to work. It probably makes measurements stable and resilient to noise.

[MiroS]

So NOT a bug. Bug is unexpected behaviour. This is not unexpected behaviour. It is clearly documented in specifications. Fact is BM869S has frequency measurement that is not very sensitive, and you need to be in right range for it to work. It probably makes measurements stable and resilient to noise.

Right , I am idiot  ->  '1000.0V    = 280V', but really ...

[2N3055]

So NOT a bug. Bug is unexpected behaviour. This is not unexpected behaviour. It is clearly documented in specifications. Fact is BM869S has frequency measurement that is not very sensitive, and you need to be in right range for it to work. It probably makes measurements stable and resilient to noise.

Right , I am idiot  ->  '1000.0V    = 280V', but really ...

I don't know you so I will have to trust your word on that.

I will repeat: I took my BM869S, connected it to 237.00 V 50Hz mains, it autoranged to 500V range and shown 50Hz correctly. Then I manually ranged to 1000V range, it shown 0237.0 V and 0 Hz. It didn't measure frequency , because according to datasheet, you need 500 V RMS minimum in 1kV range for frequency measurement to work. Is it inconvenient that it doesn't measure it in that range ? It probably might have been made to work, but who cares. Just leave it in autorange, it will autorange to 500V range and measure frequency nicely

[EEVblog]

The video linked by MiroS shows the same issue on the 869s

Actually this video is mentioning three problems for 869s, I think it is worth of checking is new multimeter is affected by them.
From Dave message - resisatnce sims to work, but who knows which firmware is in Dave multimeter, that may be old one, Joe could be one of first external testers for corrected/imporved  firmware.
I think Brymen may be still not aware of issues, I do not think that anyone reported this to them for 869s.

I have reported it to Brymen and sent a link to my video.
Seem like everyone is confirming it on their 869 and 869S.
No issue on the BM786 which switches at 650k, it's smooth as silk.
Best to keep this in the dedicated thread for this issue.

[dcac]

So NOT a bug. Bug is unexpected behaviour. This is not unexpected behaviour. It is clearly documented in specifications. Fact is BM869S has frequency measurement that is not very sensitive, and you need to be in right range for it to work. It probably makes measurements stable and resilient to noise.

Right , I am idiot  ->  '1000.0V    = 280V', but really ...

I don't know you so I will have to trust your word on that.

I will repeat: I took my BM869S, connected it to 237.00 V 50Hz mains, it autoranged to 500V range and shown 50Hz correctly. Then I manually ranged to 1000V range, it shown 0237.0 V and 0 Hz. It didn't measure frequency , because according to datasheet, you need 500 V RMS minimum in 1kV range for frequency measurement to work. Is it inconvenient that it doesn't measure it in that range ? It probably might have been made to work, but who cares. Just leave it in autorange, it will autorange to 500V range and measure frequency nicely

Just to clarify - In my previous post the "280V" was more a derived number as my boosted signal gen maxed out at 50V and then I tested with 230V mains and that wasn't enough either in the 1000.0V range. But now I tested with a better booster and my BM869s needs 265V RMS at 50Hz to register the frequency. But higher frequencies will likely require higher voltages.
   

[MiroS]

I don't know you so I will have to trust your word on that.

oh do not take this literarly   50% of range seems to be really wired , especialy if you will look at lower ranges tresholds ... anyway it is what it is. I can not even imagine what real situation can require that so high. Anyway I made false statement, I can not blame Brymen for a bug, only for 'a feature'. I will tell that kind of differences are giving a shadow  on quite good multimeter. They should display OL or something like '------'  but not 0.000 Hz and than that is truly a bug to my eyes.