Brymen 869s EMC Issue

[Trader]

I'm not sure if someone has already reported this problem:

[AVGresponding]

Yes, couldn't find the threads, but there's a couple of them on BM869S ohms measurement problems; there's the one your vid shows, and there's one where if you try to measure I think around 350k or so, it oscillates between ranges.

[joeqsmith]

Apparently I tricked data, what ever that means.

A few links talking about various problems:
https://www.eevblog.com/forum/testgear/brymen-bm869-resistance-quirk/
https://www.eevblog.com/forum/testgear/some-problems-with-my-brymen-bm869s/
https://www.eevblog.com/forum/testgear/brymen-869s-dc-offset-issue-when-reading-acmv/msg3961952/#msg3961952

***
I can believe what they are showing in the video but I can't believe someone is dumb enough to be moving around when making such a measurement.   Then again, maybe it is my lack of moving around that constitutes my tricking of the data.  I call it good practice.   

To repeat their test, I used the attached resistor plugged directly into the meters to avoid effects from the leads.   The meters were placed flat on the ESD mat.  I set each meter to record the min/max and then waved my hand around the meter until the meter stopped detecting a change. 

The worse meter I have by far is my Gossen Ultra.  Note however that it is second best next to the UNI-T UT181A.   The worse meter was one of my Fluke 189. 

Watching the video that was linked, he shows about a 1Meg change which isn't too bad.  My guess is had he used the min/max and tried to build up some more charge, maybe use some cloth, he could have made it seem much worse.   

So is it a problem?  For me, not at all.  For our friend in the video, yes. 
****
Also, I want to be very clear that while the GMW Ultra has a lot of custom made shielding to help mitigate this problem, the UT181A is also modified to help it survive ESD transients.  This modification could very well have improved its immunity.    Some of the other meters tested have been damaged but repaired using OEM parts.   
****
I have added a few other high res meters.   The UT61E+ has been fitted with different high speed clamp transistors than the original after I damaged it.  However, I do not believe this would have any effect on the results.    The OWON was also damaged but again, I don't believe it would effect the results. 

We have a new winner.....   The EEVBLOG/121GW I tested was one of the last two I had procured.  This meter was never stressed in any way.  Because they changed the filter constants so often, I have included the firmware revision.

Keep in mind again, it takes little intelligence to buy an expensive meter.  Doing so doesn't make you any smarter.  This is a dumb test.  I am showing an average delta of 3.59, or about 4X more than shown in the video.   I'm fine with it. 

[Trader]

Yes, couldn't find the threads, but there's a couple of them on BM869S ohms measurement problems; there's the one your vid shows, and there's one where if you try to measure I think around 350k or so, it oscillates between ranges.

I remember this, but is a problem related to a resistance value, I'm not sure if was an EMC issue.

[Martin72]

I'm not sure if someone has already reported this problem:

What problem ?
In the video you only see the displays, he is doing something (you can´t see there) to get the appx 22Mohm.
Then he test if someone of the meters are a theremin...

[Trader]

I'm not sure if someone has already reported this problem:

What problem ?
In the video you only see the displays, he is doing something (you can´t see there) to get the appx 22Mohm.
Then he test if someone of the meters are a theremin...

yes, the video won't show all setup, as many youtubers do.

the problem is when he moves his hand around the DMM, it's pretty clear that this interferes with the reading.

[bdunham7]

yes, the video won't show all setup, as many youtubers do.

the problem is when he moves his hand around the DMM, it's pretty clear that this interferes with the reading.

So instead of all the mystery, perhaps someone with an 869S can connect a ~22M resistor and wave at it to see if they can replicate the issue.

[Martin72]

yes, the video won't show all setup, as many youtubers do.

They should do it better..

perhaps someone with an 869S can connect a ~22M resistor and wave at it to see if they can replicate the issue.

I got one, do it "later".

[Martin72]

Hi,

There is something...

But I want to make sure if it´s not a contact problem, so I´ll do further tests later on when I got the resitor(-s) proper fixed to the brymen.

[Trader]

The worse meter I have by far is my Gossen Ultra.  Note however that it is second best next to the UNI-T UT181A.   The worse meter was one of my Fluke 189. 

Watching the video that was linked, he shows about a 1Meg change which isn't too bad.  My guess is had he used the min/max and tried to build up some more charge, maybe use some cloth, he could have made it seem much worse.

What's this "Dumb20MegTest.PNG" table means?  All the Top High-End DMMs are very imprecise to measure 20 MOhms? Should I always measure for a while and take the average? (and cross fingers?)

[joeqsmith]

The worse meter I have by far is my Gossen Ultra.  Note however that it is second best next to the UNI-T UT181A.   The worse meter was one of my Fluke 189. 

Watching the video that was linked, he shows about a 1Meg change which isn't too bad.  My guess is had he used the min/max and tried to build up some more charge, maybe use some cloth, he could have made it seem much worse.

What's this "Dumb20MegTest.PNG" table means?  All the Top High-End DMMs are very imprecise to measure 20 MOhms? Should I always measure for a while and take the average? (and cross fingers?)

From the post:

To repeat their test, I used the attached resistor plugged directly into the meters to avoid effects from the leads.   The meters were placed flat on the ESD mat.  I set each meter to record the min/max and then waved my hand around the meter until the meter stopped detecting a change. 

The min/max function captures the peak which is what the OP in the video you linked is bitching about.     If you want to make precise measurements, start by not changing the setup (waving your hands around, moving the leads, running the test next to something that emits magnetic radiation).    For accuracy,  I would turn to the documentation.

Last time I ran this same test I did it with and without the leads.  You can see what a difference it made.  It was also in the winter when the air is dry and ESD becomes much more problematic.   It was a dumb test then and still is today.   Notice how he links the same video as you. 

https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg4003000/#msg4003000

[bdunham7]

What's this "Dumb20MegTest.PNG" table means?  All the Top High-End DMMs are very imprecise to measure 20 MOhms? Should I always measure for a while and take the average? (and cross fingers?)

The current through the resistor is very small when measuring high resistance with these handheld DMMs so any motion or interference can change them a bit.  The way to measure high resistance is typically to set up the leads and then back away (and turn off the ceiling fan or close the windows as needed) to let the reading stabilize.  The Brymen is known to use somewhat smaller test currents for resistance, diode and continuity checks, so perhaps that makes it a bit more susceptible to this sort of interference.  Also, most or all of these meters will be operating in a parallel mode when measuring resistance this high and that involves smaller voltage differences than the lower ranges.

Edit:  I couldn't resist 'trying' a few meters.  Waving around the meter does get you more action than waving around the resistor or leads.  I got results similar to the Dumb20MegTest, but the most stable was the Fluke 27 with 19.72 and 20.48 min and max.  

[joeqsmith]

Our friend who produced the video you linked had written me a few times about the resistance and diode problems.    The last comment I think hits the nail on the head.  He's a typical Fluke fan boy trying to justify to the world why he bought that 87V.   

[Martin72]

Hi,

Now I´ve taken alligator clips and relatively short wires, measuring 7 3Mohm resistors in series.
I´ve only drilled them together, tomorrow I´ll solder them for best contact.
But it won´t make much difference because yes, the value varies a little bit.
But we´re talking about 21.080 between 21.100 Mohm...You can judge by yourself what that means.
And yes the second, it varies when you put your hand near to the cable, but in the ranges mentioned before.
Tomorrow last test but I wouldn´t care much about this thing. 

[Trader]

Our friend who produced the video you linked had written me a few times about the resistance and diode problems.    The last comment I think hits the nail on the head.  He's a typical Fluke fan boy trying to justify to the world why he bought that 87V.

I'm not blaming people to prefer some brands over others, but I read some complaints about Brymen (Taiwan), and my own experience about the Gossen Metrawatt (Germany) support is very bad.

In the USA, (MY opinion, OK?), by far, the best support comes from Fluke and Keysight (almost all American companies give excellent support).

Remember a while ago, the Keysight U1272A EMC issue? What they did do? They GAVE a FREE New U1272A or a new U1282A for All the affected customers (keeping the old one)!!!  Fluke also replaces the Fluke 289 super-capacitor free (I confirmed).

Both issues were already fixed, but the most important is the Reliable support, maybe this should be the decisive criteria when choosing a professional product.

I remember your videos about the Gossen Metrawatt Metrahit Ultra M248B, AFAIK, Gossen NEVER addressed that issue, a typical "sweep it under the rug" behavior against the customers.

[bdunham7]

In the USA, (MY opinion, OK?), by far, the best support comes from Fluke and Keysight (almost all American companies give excellent support).

Remember a while ago, the Keysight U1272A EMC issue? What they did do? They GAVE a FREE New U1272A or a new U1282A for All the affected customers (keeping the old one)!!!  Fluke also replaces the Fluke 289 super-capacitor free (I confirmed).

I'm guessing your opinion is based on a limited amount of second hand information.  IMO it is not reasonable to judge a company's support based on experiences in markets they don't officially serve and thus do not provide direct support to.  A limited selection of Brymen products are sold here in the US by Greenlee (and Fluke, but that's another story) and I've not heard any complaints about them.  I suspect that they rely more on unit replacement than having an in-house repair center, but I've no first-hand knowledge.

As for the support from Fluke and Keysight (and "almost all American companies") it can be a mixed bag.  Fluke's typical policies for their industrial line (the ones with the modified lifetime warranty) are pretty good, but they're struggling a bit with communication and turn-around time at the moment.  Civil unrest, 3% unemployment and locust-like consumer demand are probably their main issues.  Keysight, well it depends on who you are and what you have.  There was a whole batch of their handhelds with unobtanium failed displays that just got chucked in the trash at a fairly young age. 

[joeqsmith]

Remember a while ago, the Keysight U1272A EMC issue? What they did do? They GAVE a FREE New U1272A or a new U1282A for All the affected customers (keeping the old one)!!! 

In Brymen's case, they were willing to work with me on their 78x series Before it was released.   I couldn't have asked for better support than that.   In the video you linked, you have a clueless person blaming the product.  Its pretty typical.   

[Trader]

Remember a while ago, the Keysight U1272A EMC issue? What they did do? They GAVE a FREE New U1272A or a new U1282A for All the affected customers (keeping the old one)!!! 

In Brymen's case, they were willing to work with me on their 78x series Before it was released.   I couldn't have asked for better support than that.   In the video you linked, you have a clueless person blaming the product.  Its pretty typical.   

In Summary:

1) Gossen Metrawatt didn't reply to you about your tests, neither care about the customers.

2) Brymen is Using you as a Volunteer Electronics Test Engineer*, saving a lot of $$$, but still won't care about the customers with issues.

3) Keysight & Fluke still with, by far, the best support.

4) Is better to buy with local support (@bdunham7).

5) Non-experts can't blame products, nor be a fanboy?

* In fact, I really appreciate so much all your tests, they help a lot all the EE community, thanks.

[joeqsmith]

1) Gossen responded early on.  The USA distributor told the main division lies about what I was asking.  We went over all of that. 

2) As far as Brymen, yes indeed, they took me up on my offer to help out with the new design.   Keep in mind it was not the first time we discussed their products and it was Brymen who first offered to repair or replace my BM869s if I were to damage it during the initial tests.   They wanted to see it ran.   Call it what you want but I don't see that level of willingness to stand by your product from any other company.   I provided Dave the same opportunity with the 121GW.   The prototype 121GW had many problems still seen with the released meter.

3) While Fluke responded to my emails asking about their in-house testing,  Keysight has never responded to any of my emails.   In the 80s, I bought parts from Fluke to repair my 8000A DMM.   A few years ago, I spoke with their legal department to allow me to release a training video for one of their old DSOs.   Keysight, well they did show us that using a glass filled material for a detent spring is a very bad idea. 

4) I tried to get a replacement IC for the prototype 121GW and for the TPI Summit.   Beyond this, I haven't had a need to try and contact any of the suppliers for support.  Most of the products I have bought worked out of the box.   I did return a Mastech that was DOA but I think that's the only one.   

5) Anyone who watches a 5 minute video is now an expert on the subject.   On the EEVBLOG, if you call yourself an engineer, you are.   So sure, experts and engineers can certainly blame the products for all sorts of problems.  I would say they are also some of the most entrenched fan boys (pick a brand).     

[rob77]

tested my 869S in a noisy environment - computer without a case , LCD monitor , power strip with few phone chargers and led lights near the meter... , attached 3 x 6M8 resistors in series, hit REC and tried to "play theremin" on the meter itself and also on the test leads...
MAX 22.970Meg
MIN 20.549Meg
AVG 21.211Meg

without "playing theremin"
MAX 21.360
MIN 21.060
AVG 21.230

absolutely normal behavior i would say 

[rsjsouza]

The dude in the video is indeed showing how a 10x more sensitive instrument is more susceptible to environmental variations when compared to other two. Nothing to see here, unless for the untrained eye. Heck, one of Dave's first videos that sort of catapulted him into the spotlight describes a similar scenario with a TDS220 and an oscilloscope probe. Conclusion? When operating sensitive equipment doing sensitive measurements leave the surrounding environment alone!

*In one of his comments indicates he has a beef with joeqsmith... Get in line, I guess? 

Dave's original video:


I did a video about it (in portuguese):

[joeqsmith]

Yours are all pretty stable compared witch my >Meg shifts.  You use a different technique.    Using some material that you could charge with your hair or a cloth may improve the results.    I used some old IC shipping tubes and packing foam with the Gossen.  I wonder how that meter would handle a weak RF field now that it has all that shielding.   



>beef with joeqsmith... Get in line,
It's a very long line.

[Doctorandus_P]

the problem is when he moves his hand around the DMM, it's pretty clear that this interferes with the reading.

So, what is he doing with his other hand when he's waving at the DMM?

Without having any sort of background of this youtube guy I default to not trustworty.
He's got a few other videos on youtube, all 2 years old and the're vaguely related to transportation.
For all I know this guy is paid by fluke to give Brymen a bad name.

[Trader]

the problem is when he moves his hand around the DMM, it's pretty clear that this interferes with the reading.

So, what is he doing with his other hand when he's waving at the DMM?

Without having any sort of background of this youtube guy I default to not trustworty.
He's got a few other videos on youtube, all 2 years old and the're vaguely related to transportation.
For all I know this guy is paid by fluke to give Brymen a bad name.

There are many reasons for that result:  Is his Brymen 869s defective? Does he have an excessive field environment (noise, carpet, wool, etc)?

I won't say he faked the tests, I guess it's correct.

Did you see @joeqsmith "Dumb20MegTest2" table?  Some of his DMMs can vary above 15%.

And the @rsjsouza video? All of his DMMs are varying just 1% roughly.

My idea with this post is just to bring this behavior to the discussion, I really appreciate everyone that contributes with some tests.

[Martin72]

A few weeks ago after buying the 869s, I gave it away for external calibration.
It comes back succesfully calibrated with documents enclosured, telling me everything is alright, including the resistance measure.
And the tests I´ve done a few days ago with high resistance were in the specified range, although they can vary.
As long as they did it in the range, there is no need to worry.
High precision/sensitive measure devices interacting with the enviroment, you have to care about when taking measures.
In some of our test procedures there are notes to take the measure equipment away for a minimum distance of the DUT, because of EMV.
Therefore I didn´t care much about of those "look when I do this, something crazy will happen" videos.
Especially when no further details are present.
Or to keep it short, when your multimeter will be disturbed, take care for non-disturbing.

[Caliaxy]

Not that I don’t have anything better to do, but got curious and checked mine, on my bench. It does it, too. Waving the hands around the LCD area seems to change the reading with a few MΩ up and down, briefly. Once you stop the motion, the reading gets back to normal, slowly. Holding the hands still or moving them very slowly does not seem to bias the readings. Waving the hands around the resistor/wires does not change the reading, at least not by much.

Other high-end handhelds I happen have around (a Keysight U1282A) do it to, but in a smaller extent (a few tens or hundreds of kΩ). Lower-end Brymens (235, 257), Uni-T 61E, Agillent U1252A seem pretty immune and so are my Flukes. Sanwa PC7000 (same chipset as BM869S, according to some fellow eevblog forum members) does it too, in the same extent as BM 869S.

My data point… Of course, as already pointed out by others, there is no reason to wave your hands like that while measuring a 20MΩ resistor. We can all agree on this

[rsjsouza]

Yours are all pretty stable compared witch my >Meg shifts.  You use a different technique.    Using some material that you could charge with your hair or a cloth may improve the results.    I used some old IC shipping tubes and packing foam with the Gossen.  I wonder how that meter would handle a weak RF field now that it has all that shielding.

Indeed I did not put much effort to maximize the interference. I was going on the premise of how an occasional "hand waving" will disturb the measurements, which is terribly imprecise and will be influenced even by the season of the year - my lab is now hovering around 50%RH but in the winter the influence would be much higher when it reaches 10 or 15%RH. Heck, if I bring a party balloon and its 6~10kV of surface static voltage (measured in the winter), I can even risk damaging something.

If I am bothered enough, I might break out my ionizer and field meter to get some more controlled measurements...

Other high-end handhelds I happen have around (a Keysight U1282A) do it to, but in a smaller extent (a few tens or hundreds of kΩ). Lower-end Brymens (235, 257), Uni-T 61E, Agillent U1252A seem pretty immune and so are my Flukes. Sanwa PC7000 (same chipset as BM869S, according to some fellow eevblog forum members) does it too, in the same extent as BM 869S.

My data point… Of course, as already pointed out by others, there is no reason to wave your hands like that while measuring a 20MΩ resistor. We can all agree on this

Indeed this is an interesting observation. While our Keysights agree, my UT61E was pretty sensitive when compared to yours and Joe's. Not only the internal build, but the bandwidth of the meter will also influence this.

All in all, it boils down to procedure but it is an interesting discussion indeed.