Handheld meter robustness testing

[2N3055]

Dave is right, we should be discussing this in a topic he created for that...

[dcac]

I'm not sure if Dave meant that for all BM869 discussion or just for the 'resistance quirk'...

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]

But higher frequencies will likely require higher voltages.

True. 

IMO, it all has to do with meter robustness so no big deal to put it here or where ever. 

Good to hear it's been reported. 

[joeqsmith]

How sensitive is the BM869s in its upper resistance range compared with other meters?

Started by using the resistive decade box to determine the thresholds for the upper range.  Use the center as the test resistance.   Using my terminated  HP34401A signal generator and placing it in series with the DMM and the decade box.  Using a 50Hz  sinewave, increase the voltage until the DMM is unstable.  Reduce the voltage until it recovers. 

The newer Brymen BM869s with the latest firmware has thresholds of 540K and 470K ohms.  Setting the decade box to 505K, the meter is unstable at as low as 27mVRMS.  Test current was around 0.85uA.

The Fluke 189 that my friend gave to me, has thresholds of   560K and 470K oms.   Using a test resistance of 515Kohms, the meter was unstable as low as 740mVRMS.   Test currents are around 1uA

My UNI-T UT181A after being damaged, repaired and modified has thresholds of 600K and 550K.  Using 575Kohms, the meter was unstable as low as 680mVRMS.  It has an odd behavior where the longer it sits with the AC wave applied, the chances are it will have a problem.  For example, I can apply 1V and the meter will read the correct value for a short time, then bounce between ranges.   

The Yokogawa TY720, again damaged and repaired but no mods to meter.   Thresholds were 560K & 490K.  Testing at 535Kohms, I could test up to 3.52 Volts RMS and the readings were stable.  It of course effects the reading but it never hunts.   

The Brymen BM786 with the last firmware I received has thresholds at 660K and 600K.  Using a test resistance of 630K, the was unstable as low as 380mVRMS. 

How good is good enough?
The longest standard leads I have came from Keysight.  They are 1.4 meters long.   If I attach these to the decade box and run them aside a 120V 60 lamp cord for a few feet,  I get about 20mV.    It I use their entire length, about 30mV.    If I twist them the full length with the line cord, 10 turns total, about 43mV.     I'm not suggesting I would use a meter this way but it puts some numbers on the board.   

*****************
I was curious about the Gossen Ultra or what ever Gossen's marketing now calls it (PN# M248B).   Never damaged but a lot of Netic and copper foil added for shielding.  This meter has some pretty major problems but can throw up some very impressive data after I added all that shielding to it.     

I tried two tests with it.  First threshold was 310K and 270K.  The next was 3.1Meg and 2.7Meg.   I ran it at 290K and 2.9Meg.    In both cases, I could go to 7.1VRMS and saw no effect.   It's stable down to the last digit!  That's 290.771Kohms BTW.    I doubt the shielding I added helped at all with this testing other than I can get near the meter without it going unstable. 

Clip of the Gossen with up to 35V superimposed. 


The thresholds are correct but the data was not!   Repeating the test with the proper connections, at 270K the range was still stable up to 3.5VACRMS.  At 2.7M with 3.5V applied, again the range was stable but the values are off by a fair amount and the reading vary by about 40Kohms.  .

*****************
So for fun, using the prototype 121GW which was damaged, repaired, modified, aligned using hand picked parts.  Version 1.57 firmware.  Thresholds are 550K and 383K.  Using a test resistance of 467Kohms, the meters range is mostly stable up to 1VRMS.  The displayed resistance varies a lot.   Even with 18mV applied, the reading is not very stable.     

I tried it with the Production #1  121GW FW 2.02 as well and it seems about the same. 

When I first ran this test, I realized I had connected it up wrong when I ran the Gossen.  No wonder it was so stable!!  Should have rechecked the obvious.  Corrected above. 

[MiroS]

How sensitive is the BM869s in its upper resistance range compared with other meters?

Old Fluke 25/27  can  measure 2kOhms with 1V AC, 1MOhm with up to 2V AC noise.
That noise is visible as an oscillating  bar graph.

[joeqsmith]

How sensitive is the BM869s in its upper resistance range compared with other meters?

Old Fluke 25/27  can  measure 2kOhms with 1V AC, 1MOhm with up to 2V AC noise.
That noise is visible as an oscillating  bar graph.

Of course, randomly picking some points isn't going to tell us much.  If for example, I choose to test at 260Kohms with the BM869s (mid wayish point of the two ranges to allow for the highest immunity).  Remove the terminator from the 34401A and transformer couple it into the series chain.   I can now apply 36VRMS and the meter is still stable.   That's all fine but it's not helpful to this discussion.     

If you have some pots, you could try to find the thresholds, then measure the pots values.  Then set the pot to the mid point and try to see what voltage level upsets it. 

****
Just for fun, changed the decade box to 2Meg, similar to what I show in the video. Same 50Hz, 36VRMS being applied.  The BM869s is stable.  Explains why it was stable during my tests with a 1.9M part.      Again, same wrong setup as I mentioned with the Gossen.  The meter will begin to hunt at 750mV.  Still a huge improvement.  Makes more sense now..   

[MiroS]

If you have some pots, you could try to find the thresholds, then measure the pots values.  Then set the pot to the mid point and try to see what voltage level upsets it. 

I took this data from Fluke manual, not by testing, no idea why Fluke published this for  2k and 1M, and no idea if this could be a trend that lower resistance is more sensitive (1V) than higher resistance (2V)

[dcac]

The newer Brymen BM869s with the latest firmware has thresholds of 540K and 470K ohms.  Setting the decade box to 505K, the meter is unstable at as low as 27mVRMS.  Test current was around 0.85uA.

That correspond very well with what I'm experience when measuring 10 x 100K SMD resistors in series on a small PCB. Using the Brymen test leads and holding one probe in each hand, no mains wire or power supplies are close by.

Everything is fine until 500K (5 x 100K) and my BM896s starts hunting. Then switching it to AC mV and repeat with the probes in same position I get 28mV RMS. I noticed by just changing the angle on the probes slightly I could get the voltage down to 22-24mV and this also caused the hunting to stop in Resistance mode.

   

[MiroS]

If you have some pots, you could try to find the thresholds, then measure the pots values.  Then set the pot to the mid point and try to see what voltage level upsets it. 

I took this data from Fluke manual, not by testing, no idea why Fluke published this for  2k and 1M, and no idea if this could be a trend that lower resistance is more sensitive (1V) than higher resistance (2V)

I have  compared for my curiosity BM869s, FK87V and FK289  - 2kOhm with 0.3V AC 50Hz:

BM869s - not able to show resistance at all
FK87V - no any problem at all, shows what is should , even in HiRes mode, nicely singaling noise on bargraph
FK289 - no any problem at all, shows what is should  and singaling noise on bargraph

FK28II - the same as FK87V

[dcac]

Here's two more for reference. I used one of those cheap KKmoon signal gens as you can easily run them from a 5V power bank and much exclude any mains interference. I verified I got the same results that BM869s starts hunting at about 28mV RMS - I also tested at 60Hz and here it was perhaps slightly more sensitive at 26mV RMS.

Then tested the 121GW - thresholds seems to be at 400K and 550K so tested at 470K and it could take 0.81V RMS before starting to hunt.

And then Agilent U1252A which seems to have thresholds at 450K and 510K so tested at 480K and it could take more than 2.0V RMS - of course the resistance value got noticeably distorted with even higher voltage - but it still would not hunt.

Edit: just noticed Joe already did the 121GW - but mine had FW 2.04 - I don't think that would make any difference though.

 

[dcac]

I tried some measurements with my Picoscope + laptop on batteries, trying to minimize mains coupled interference as much as possible.

So this is BM869 measuring 525K resistance with a 10X 10M scope probe in parallel.

Two captures at 200mS and 500mS/div.

Cyan trace = 500K range.
Magenta trace = 5M range.
Blue trace = Hunting auto range.

You can see the faint overlayed 50Hz hum on all traces and then there’s about 10% overshoot each time the 500K range is selected. But as I do not have the schematics it’s difficult to speculate exactly what’s causing it. 

And then the overall hunting ‘frequency’ at about 6-7 SPS - doesn’t really make sense as BM869's nominal update rate is 5 SPS. So it actually seems to be hunting faster than that and that doesn’t seem to leave much time to settle after each range change.

[joeqsmith]

Quote from: joeqsmith on November 19, 2020, 02:02:50 am

    I have repeated the DCV test to see if I could replicate the problem with the bargraph.  It doesn't look like it locks up but rather that they calculate it differently than the readout.  It's really strange but easy to reproduce.  I have asked Brymen if they can fill us in on what it's doing.

See Brymen's response below:

The digital reading and analog bargraph of BM780 DCV function come from different algorithm designs. Analog bargraph algorithm uses faster converter to get the peak averaging RMS (not True RMS) of the input signal. That is the reason why bargraph display speed is more faster than digital reading update speed. That is also the reason why bargraph algorithm can be with the mechanism to judge if the input DCV is with additional high ACV and if meter should auto switch to higher measuring range in case of being with high ACV. Nevertheless, its peak averaging RMS algorithm design is with a nature. While the DCV level is less than ACV peak, the bargraph will display ACV peak averaging RMS only. The bargraph will start to reflect DCV component only in case DCV level is higher than ACV peak. That is the reason why the bargraph appeared to lockup against your tests.     

[/i]


I haven't forgotten about this meter.  There was another question that arose when working on Part 4 and I wanted to get an answer to that before moving forward.   

[joeqsmith]

Someone had posted this link, asking me if the new meter was going to be available January. 

https://www.welectron.com/Brymen-BM786-Multimeter-EEVBlog-Edition_1

To be clear, I don't have any inside knowledge as to when the meter is going to be available, what the cost will be or who will be distributing it.   Ultimately it's up to Brymen.  What I can tell you is that while working on Part 4, I had another question.  Brymen has written me and are looking into it. 

With this meter not being released yet, I would rather provide them with the opportunity to address anything that comes up.  They continue to be very responsive but again, making changes requires time.  Personally, I am glad that they are willing to take the time to investigate questions I have.   

Just a quick reminder.  I block posts with links to prevent using the channel for spamming ads.

[EEVblog]

Someone had posted this link, asking me if the new meter was going to be available January. 
https://www.welectron.com/Brymen-BM786-Multimeter-EEVBlog-Edition_1
To be clear, I don't have any inside knowledge as to when the meter is going to be available, what the cost will be or who will be distributing it.   Ultimately it's up to Brymen.

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

[joeqsmith]

Someone had posted this link, asking me if the new meter was going to be available January. 
https://www.welectron.com/Brymen-BM786-Multimeter-EEVBlog-Edition_1
To be clear, I don't have any inside knowledge as to when the meter is going to be available, what the cost will be or who will be distributing it.   Ultimately it's up to Brymen.

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

What about price and release date?  Anything solid yet?

[EEVblog]

Someone had posted this link, asking me if the new meter was going to be available January. 
https://www.welectron.com/Brymen-BM786-Multimeter-EEVBlog-Edition_1
To be clear, I don't have any inside knowledge as to when the meter is going to be available, what the cost will be or who will be distributing it.   Ultimately it's up to Brymen.

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

What about price and release date?  Anything solid yet?

No release date yet, they are still working on it.
Like I said, I can sell it significantly under US$150. Reseller prices and amazon prices may vary.
For those who want it cheap sign up for my newsletter on the website and you'll get notified of a discount when it goes on sale.

[bdunham7]

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

I'm curious as to why neither you nor Brymen have a US distributor?

[joeqsmith]

Someone had posted this link, asking me if the new meter was going to be available January. 
https://www.welectron.com/Brymen-BM786-Multimeter-EEVBlog-Edition_1
To be clear, I don't have any inside knowledge as to when the meter is going to be available, what the cost will be or who will be distributing it.   Ultimately it's up to Brymen.

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

What about price and release date?  Anything solid yet?

No release date yet, they are still working on it.
Like I said, I can sell it significantly under US$150. Reseller prices and amazon prices may vary.
For those who want it cheap sign up for my newsletter on the website and you'll get notified of a discount when it goes on sale.

Are you waiting on a final cost from Brymen?  Just curious as when I asked them about pricing, they also gave me a vague sort of answer.   A range based on other products.   

[joeqsmith]

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

I'm curious as to why neither you nor Brymen have a US distributor?

Greenlee and AMPROBE have rebranded some of their products.  Maybe there are others.   

[Cliff Matthews]

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

I'm curious as to why neither you nor Brymen have a US distributor?

Greenlee and AMPROBE have rebranded some of their products.  Maybe there are others.

or maybe it's because he's a one man shop with 2 sons looking up to him? If he extends trust to a US rep to carry a $25K stock level on all his merch and a cash float of $10K to manage flow, one false move and someone may miss-out on college..

[EEVblog]

The BM786 is exclusive to the EEVblog, it will not be sold under any other brand or dealer.
Welectron and Spark Labs/Simon's Electronics are the only dealers who will carry it.
You'll have to buy the BM789 or the lower model if you want the Brymen brand.

I'm curious as to why neither you nor Brymen have a US distributor?

Because there is no hassle selling into the US from Australia like there is to the EU.
If you are absolutely desperate and want it tomorrow in the US then I sell via Amazon, with Amazon US holding the stock. In fact I sell tons via Amazon US, so much so that I have a hard time keeping stock  levels up.

[EEVblog]

Are you waiting on a final cost from Brymen?  Just curious as when I asked them about pricing, they also gave me a vague sort of answer.   A range based on other products.   

They gave you a vuage answer because they don't set retail prices, and various dealers would likely have different FOB cost deals and different margins.
I've just been lazy and haven't worked out a final retail cost yet, is that ok?

[Fungus]

Like I said, I can sell it significantly under US$150.

Makes me wonder how much Fluke makes on every meter.

[joeqsmith]

Are you waiting on a final cost from Brymen?  Just curious as when I asked them about pricing, they also gave me a vague sort of answer.   A range based on other products.   

They gave you a vuage answer because they don't set retail prices, and various dealers would likely have different FOB cost deals and different margins.
I've just been lazy and haven't worked out a final retail cost yet, is that ok?

Of course that's fine. 

***
I assume my asking the question was ok as well?

It sounds like we will be back in business soon.   

[bdunham7]

Makes me wonder how much Fluke makes on every meter.

Hopefully enough to pay their employees well, cover their warranty and support obligations, and make every other DMM maker jealous of their profits.