Brymen BM869 Resistance Quirk

[EEVblog]

A thread specifically for this quirk.

[Martin72]

Yeah, as known as in my daily test practice, I rub the leads on my shirt or put them "hard" on the mains cable...
In private, I got a BM869s, will take it to work as we got resistance decades there and test it.

[gnavigator1007]

Tried it on the bm235 yet?

[EEVblog]

Tried it on the bm235 yet?

Yep, not a problem.

[gnavigator1007]

Was curious as Joeqsmith noted the original commenter on his video did seem to imply that this was common in Brymen meters, but was vague. Pretty interesting following along as issues like this are just being found on such a mature product that has been popular in the community.

[dcac]

I can confirm this too. With my BM896S I cannot measure 470K - 990K with handheld probes without it starting to alternate between 500K and 5M range. However if I manually select 5M range I can get a perfectly stable reading.

I did not have the test leads near any mains powered equipment - just holding the probes one in each hand.   

I would not be surprised if this problem is (much) more noticeable in 50Hz environment than in 60Hz. I'm in Sweden so here it's 50Hz but Joe is in the US 60Hz so perhaps that's why he hasn't noticed it that much yet.

[coromonadalix]

would be nice to try other meters with the same chipset

[dcac]

I've always thought Brymen did their own chipset - but perhaps they use the same on some models.

[coromonadalix]

oops  yes indeed  they are btc branded if i recall     my bad

I  tought some amprobe meters like the am140 am160  where using the same ic's  since they have 500,000 count mode ?

I have an Amprobe  am140, but no decade resistor box         out of curiosity

[Fungus]

Look at the bar graph, it's obvious where it happens - right around the point where a 50,000 count meter switches between the 500k and 5M ranges.

Seems to me like the threshold where it switches should be just a teeny bit lower.

[floobydust]

What is the test current? Handhelds seem to compromise due to their low supply voltage.

34401A Ohms 10MEG 0.5uA
34401A Ohms 1MEG  5uA
34401A Ohms 100k  10uA

AN8008 Ohms 999.9k 0.2uA
AN8008 Ohms 99.99k 1.2uA

TECH310 Ohms 20MEG 0.3uA
TECH310 Ohms 2MEG  0.5uA

Some of Dave's rubbin' is making large common-mode voltage due to static electricity. You can try it by touching one multimeter lead/banana jack (while on ohms with test resistor connected) while shuffling your feet.

[AVGresponding]

I measured the current at 1uA for the 500k range, and 0.1uA for the 5M.

Used my Keithley 2000 and Fluke 289, they agreed with minor LSD differences.

[radiolistener]

There is also another electrostatic issue with brymen. I have 867S, when I touch left bottom side of the LCD with some electrified stick, it affects measurements. And LCD segments starts to blinking. You can use some polyethylene film to make electrostatic charge on the display for the same effect.

[hammy]

A thread specifically for this quirk.

This only happens in autorange mode.

BTW we all know all devices have their own quirks. This is the reason it is important to KNOW your multimeter.
Induction from a mains wire to your test leads is something you know to avoid during a test.

Using some cheap copper-wire for measurements is allways a bad idea, especially with unknown isolation who is prone for static charge.
This is the reason technicians should do measurements and tests in the real world, and not egg-headed academics. You need to know the real world influences and the limits of your measuring device.

This is the perfect storm in a teacup!
  

[AVGresponding]

Hardly a storm in a teacup. As the thread title states; it's a quirk.

Exploring the conditions where this quirk manifests is important for knowing "the real world influences and the limits of your measuring device" as you put it.

I don't see anyone here as an "egg-headed academic", pretty much everyone I've conversed with on this forum has a great deal of real world, practical, test and measurement experience.

[Psi]

Antistatic mats are pretty good at coupling mains noise into your circuits.
I've seen that on more than one occasion, circuit gets all screwy when sitting on the workbench matt but works fine if lifted up.

It's probably why Dave had it fail once when there was no mains psu/cable nearby, the meter wires were probably resting on the antistatic mat.

[dcac]

Look at the bar graph, it's obvious where it happens - right around the point where a 50,000 count meter switches between the 500k and 5M ranges.

Seems to me like the threshold where it switches should be just a teeny bit lower.

Yeah or a slightly longer delay before the meter decides to switch range - BM896 is quite fast at auto ranging so quirks like this could be seen as a trade off for that speed. Still out of 10 or so different handheld multimeters the BM896 is the only one that has this problem on my work bench.

[Fungus]

Dave,

To test this on the new Brymen 786 you need to use something like 680k Ohms - it's a 60,000 count meter.

[Martin72]

In private, I got a BM869s, will take it to work as we got resistance decades there and test it.

Test it, same here....
But: Regardless if it´s in auto- or manual range.

[Martin72]

Further testing at home, a few minutes ago...

Instead a decade, I took a single metalfim resistor with 560k and put it in the shortest way to the sockets of my 869s, with wago cage-clamps:

https://www.reichelt.de/de/de/bananenstecker-4-mm-cage-clamp-orange-wago-215-211-p100778.html?PROVID=2788&gclid=CjwKCAiA7939BRBMEiwA-hX5J_1QEKpYP8unTK6uJHkBykHVb1sbwAPEngR5MCL3_3C-NwRtVNTPFhoCOmAQAvD_BwE&&r=1

Result: Stable displaying value, you can "knock" on it, nothing happens.
Then I took 1m 4mm cables, value becomes instable when moving the cables.

[EEVblog]

Dave,
To test this on the new Brymen 786 you need to use something like 680k Ohms - it's a 60,000 count meter.

Yep, did that. Ultimate range switching threshold is about 650k, less if you do big jumps. It's as smooth as silk, no sign of this issue at all.

[floobydust]

I would check if it's sensitive to the range switching time-constant, the firmware might not be waiting long enough and then just bounce back and forth between ranges.
Add capacitance (across) the resistor, say 100pF or 220pF and see if changes things for the worse. It should make no difference...

[cluca1969]

I did some testing on my BM869s and it seems that there are still some capacitive and inductive input problems. I used a resistance decades and I varied values between 510K-600K, it seems that these areas are the most affected. I used different cables, some shielded and others I shielded myself, in vain, the defect (anomaly) did not disappear. But to shorten the discussion I will tell you what I managed to get after a few hours of testing. I put a capacitor with a value of 47nF - 100nF at the input of BM869s and the device became very stable regardless of the position of the cables or the touch with my hand of its active parts or of the resistive decade. (First, I made measurements in Manual Mode and here it seems to be fine).

[wraper]

Tried it on the bm235 yet?

Yep, not a problem.

Appears to be autoranging issue. If you use a multimeter with different resolution, of course the magic value will be different even if autoranging algorithm behind it is exactly the same. You should have selected the range manually and see if quirk goes away.

[bdunham7]

I measured the current at 1uA for the 500k range, and 0.1uA for the 5M.

Used my Keithley 2000 and Fluke 289, they agreed with minor LSD differences.

Can you check with a range-appropriate resistor in  the circuit?  I think that the switch from 500K to 5M is going from a constant-current range to a proportional-current range where a 10M or so resistor is switched in across the output and the resistance is calculated.  These proportional resistance ranges are always less accurate and less stable, so I suppose they are more susceptible to interference as well.

[dcac]

It would be interesting to know if BM867s also has this quirk. It seems to be identical to BM869s except for only one Temp input, no VFD mode and some differences in the accuracy.

[Fungus]

It seems to be identical to BM869s except for only one Temp input, no VFD mode and some differences in the accuracy.

RMS measurements go to a much higher frequency in the 869 (100kHz vs 20kHz IIRC).

[Martin72]

Yepp.

All in it´s a interesting thing here, want to test our Flukes tomorrow the same way...

[AVGresponding]

I measured the current at 1uA for the 500k range, and 0.1uA for the 5M.

Used my Keithley 2000 and Fluke 289, they agreed with minor LSD differences.

Can you check with a range-appropriate resistor in  the circuit?  I think that the switch from 500K to 5M is going from a constant-current range to a proportional-current range where a 10M or so resistor is switched in across the output and the resistance is calculated.  These proportional resistance ranges are always less accurate and less stable, so I suppose they are more susceptible to interference as well.

I picked a 1M 1% 0.6W metal film resistor, measured it at 1M003.

The current stays at 0.1uA, but there was an interesting difference between using the Fluke 289 and the Keithley 2000 to measure it.
The Brymen really didn't like being connected in series with the Keithley, the bargraph was pumping like a VU meter at a rave, and the resistance reading went up by ~25k. It had to be done in manual range; autorange gave the same glitchy behaviour that started this thread.
The Fluke it didn't mind at all, and the resistance reading was thrown off by only 1.5k or so.

It's worth noting that my "lab" is a corner in my living room, and if I want precise readings, say using my 7075, not only do I have to warm the instrument up first, as is normal, I also have to switch several things off, such as my PC, TV, and LED lights.
I didn't bother to do that for this test, so there's likely a lot of noise for the poor 869s to deal with.

Still, none of my other meters struggle in the same way (that I've yet found, I haven't devoted any time to exploring the possibility).

[EEVblog]

I heard back from Brymen:

"There should be no body who would perform high-resistance measurements under the shown conditions. This should be a very minor issue. Anyway, to change C28 to use bigger capacitor may improve it drastically. Currently C28 is a 10nF MPE capacitor. You may try to add a 47nF MPE capacitor to C28 in parallel as attached photo shows to check."

So sounds like they don't consider this a problem that needs fixing.

[joeqsmith]

Thanks for taking the time to ask them.   Did they offer any clues to what the tradeoff is when changing this value?  I assume it degrades some other performance. 

While I haven't had a problem with the 120V 60Hz mains,  I ran a fairly extensive test on a few of the meters I have and it does appear the BM869s is much more sensitive than others.  I measured how much coupling I would get with the leads right next to the line cord but I suspect if you lived in an area that had higher voltage mains with more distortion, it would exasperate the problem.   

https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg3336556/#msg3336556

[AVGresponding]

I heard back from Brymen:

"There should be no body who would perform high-resistance measurements under the shown conditions. This should be a very minor issue. Anyway, to change C28 to use bigger capacitor may improve it drastically. Currently C28 is a 10nF MPE capacitor. You may try to add a 47nF MPE capacitor to C28 in parallel as attached photo shows to check."

So sounds like they don't consider this a problem that needs fixing.

I can't bring myself to see it as a "problem that needs fixing", it's just a quirk to be aware of.

[dcac]

I heard back from Brymen:

"There should be no body who would perform high-resistance measurements under the shown conditions. This should be a very minor issue. Anyway, to change C28 to use bigger capacitor may improve it drastically. Currently C28 is a 10nF MPE capacitor. You may try to add a 47nF MPE capacitor to C28 in parallel as attached photo shows to check."

So sounds like they don't consider this a problem that needs fixing.

When I find time I will test this and then re-run the measurement I did here: https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg3338880/#msg3338880


BTW, Dave do you have the BM869 schematics or perhaps you know where in the circuit C28 sits?

[Martin72]

I heard back from Brymen:

Mee too, from welectron (brymen distributor) and it´s the same plus they don´t plan to change the capacitor in the current series, also it´s not clear which effect to the accuracy the capacitor change will have.
So for me one thing is clear, I won´t change anything on my calibrated brymen.
Just taking suitable leads und don´t moving around. 

[2N3055]

I experimented a bit.
Signal has to be injected into circuit, common mode interference is well suppressed.
In order to provoke this, you have to ground one side (to protective earth), and touch other side with  finger... In this case, because of high impedance, enough voltage is injected to create problems.

Thing is, while I agree it's not optimal, I never had this problem, in few years of everyday work with it. I thought about it and here is why:

1. When measuring resistors that are more than few kOhms, you should use grabbers or crocodiles and not touch it with fingers (if resistor is in free air) anyways. Your finger resistance will kill accuracy right there. When resistor is mechanically attached, you use just probe tips, again no finger contact. So no voltage injection from you.
2. For every measurement it is a good thing to make sure that you don't make your probe wires an obvious secondary of a transformer. It will screw up your AC measurements, or true RMS, or AC+DC...  So don't wrap your probe cables around transformer or electric motor. Keep your signal and power cabling separate. Make sure to keep them away as much as possible. Make that constant rule, for every measurement.
3. Everybody speaks about ranging hunting as a problem. It is not. Problem is change in measured resistor value, that as a consequence makes meter hunt for range.  And it is a good thing it did, because when you see meter hunt, you know you have some interference and that you have to fix your measurement setup. It makes it obvious. Otherwise, maybe nobody would notice that your meter read 10 or 20% lower than it should.   I for one, if this sensitivity to injected AC cannot be made more robust, want to keep this rang hunting behavior on purpose, as warning indicator that I have unstable measurement. That is much better than instrument silently lying to me....

[dcac]

Measuring resistances above let’s say 100K or so and you want an accurate value it’s of course always a good idea to not touch the test leads at all and/or switch to shorter leads. But I cannot see a hunting auto range as something beneficial in any situation. Sure in (more) noisy environments the shown value might be jumping a bit - but I really want to see the value and not a display hunting between two ranges essentially telling me nothing.

If BM869 firmware could (easily) be updated by the user - Brymen probably would have fixed this that way. It will be interesting to see if it is fixed on newly produced BM869‘s.

I have yet to test the suggested C28 cap mod - but hoping it provide a more useful meter at least for me anyway.

[2N3055]

Measuring resistances above let’s say 100K or so and you want an accurate value it’s of course always a good idea to not touch the test leads at all and/or switch to shorter leads. But I cannot see a hunting auto range as something beneficial in any situation. Sure in (more) noisy environments the shown value might be jumping a bit - but I really want to see the value and not a display hunting between two ranges essentially telling me nothing.

If BM869 firmware could (easily) be updated by the user - Brymen probably would have fixed this that way. It will be interesting to see if it is fixed on newly produced BM869‘s.

I have yet to test the suggested C28 cap mod - but hoping it provide a more useful meter at least for me anyway.

That is what I'm saying. Hunting is indication that read value is bullshit. You don't want it. It is more than 10% in error and fluctuating. In which case you fix measurement situation and than get a reading.

Making a mod that will make instrument not hunting ranges, but will happily show, say, 480 kOhm value for the 565 kOhm resistor without you suspecting anything is not going to be more useful. It is going to be much worse...

If they cannot provide a "fix" that will make BM869S better reject (suppress) errors in measurement on that range with AC superimposed on input, I don't need more stable autoranging. Because this way I at least get some warning something is wrong. So, if nothing else I know I have to grab another meter and measure it with that one, instead of reading some resistance with 10-20 % error.

Let me explain what happens. 869S is very sensitive to injected noise in 5 MOhm range. So when you measure 500-560 kOhm, that is at the very bottom of that range, just about to range down to 500kOhm range.
So you inject some noise, that probably rectifies over some diode in the circuit (parasitic or maybe protection) and it makes meter think resistor is, say, 480 kOhm. Meter than ranges down to 500 kOhm range. Which is not so sensitive and has much lower impedance, and that range accurately says "oh that R is more than 500k" so it ranges back up.. Rinse and repeat. And all that flip-flopping happens quite fast, because 869S has quite fast autoranging.

There is nothing to fix in autoranging. It works perfectly. Thing that is not so robust (as it could be, frankly) is that in 5MOhm range 869 is sensitive to superimposed AC signal when measuring. If that cannot be made more robust, then there is nothing else to fix.

But with correct measurement practice, you shouldn't encounter problem anyways. If you do see hunting, you know what it is, and rearrange wiring and if you get stable reading you can be sure it's going to in spec too.

If all of that is a problem to you, and you work in environment that is so noisy that you get this problem all the time, you, then need to get some other meter.

[dcac]

Like I said earlier out of 10 or so handheld multimeters BM869s is the only one having this problem on my workbench.

And also it even seems to be hunting faster than it's nominal update rate of 5 SPS. I really cannot think of any scenario at all where range change needs to be that quick - so yeah I say that's not normal and needs fixing.

But if it anyone feels better thinking this could be considered a feature - that is of course entirely up to them.

[Fungus]

I say that's not normal and needs fixing.

I wouldn't say it desperately needs fixing, just stop rubbing the cables on your jumper when you measure resistances.

[dcac]

I say that's not normal and needs fixing.

I wouldn't say it desperately needs fixing, just stop rubbing the cables on your jumper when you measure resistances.

I'm wearing a t-shirt.

[2N3055]

Same here. If someone thinks that it is ok to have 30%  error in measurements as long as autoranging is slow, that is their prerogative.

Just to make sure, BM869 is hunting because it reads wrong value in presence of AC.
It's not problem in autoranging. Problem is that at that range, some other meters can reject 10-20X more interference without measurement being wrong. So their measurement is stable, and so is ranging.

Even if you make it range more slowly, it will still hunt, just slower. and autoranging will be slower (that is not what i want, and neither anybody else does), and it will still read wrong values. So you get meter that is both slow and wrong... What's not to like..

If C28 mod makes meter better reject AC superimposed on circuit while measuring in 500kOhm range, and doesn't interfere with measurement settling time, bandwidth of meter in other ranges or what not, than it would be something that I would be tempted to do. But autoranging doesn't seem to me to be the source of problem..

And I never had a problem manifest itself.. So I think this is the weird corner case, not a showstopper. It is annoying to know that something is not perfect, but nothing is. This is similar to spurs in your spectrum analyser, as long as you know what and where they are, you can work around  them.

Don't get me wrong. If this not annoying characteristic of otherwise superb BM869 can be made better I'm all for it. As AvE sys, "betterer is better"...

If not, I won't replace it or think less of it. It's not a real problem, most of the time. And it is not hidden flaw.
When it goes wrong it clearly signals alarm, and you'll know to do something about it. At that moment it might be annoying but I still trust it. It is same thing as when instrument say "Uncal", or "Out of range". There are some instruments that will say "Not enought data" or "Measurement unstable".

[bdunham7]

Just to make sure, BM869 is hunting because it reads wrong value in presence of AC.
It's not problem in autoranging. Problem is that at that range, some other meters can reject 10-20X more interference without measurement being wrong. So their measurement is stable, and so is ranging.

I don't know one way or the other, but are you sure about that?  Having looked at quite a few meter schematics, some pick off the signal for autorange quite a bit earlier and it is entirely possible that interference that causes auto-hunting will be rejected by the actual DC measurement circuit.  My Fluke 8846A is notorious for this, on the lower DC ranges it likes to jump around and flash 'OVERLOAD' but then it takes perfectly stable readings in 3 different ranges.  The test for rejection vs autorange would be to see how stable the actual readings are in manual range mode.  I would expect any decent meter to reject quite a bit of AC in any DC measuring mode, including ohms.

Just for comparison I tried this with my F289 and got very little change with the lead-wiggle/shirt rub and nearly none with the AC coupling.  The autorange has quite a bit of hysteresis, ranging up at 480-490K and ranging down around 450K.  I was not able to get it to waver with anything from 450K to 600K.  The F289 has very good AC rejection--there's one procedure on a calibrator repair that requires zeroing (balancing) out a signal +/- 5mVDC in the presence of 35Vrms, and it works perfectly.  It might be worth testing the Brymen for normal-mode AC rejection.

[2N3055]

Just to make sure, BM869 is hunting because it reads wrong value in presence of AC.
It's not problem in autoranging. Problem is that at that range, some other meters can reject 10-20X more interference without measurement being wrong. So their measurement is stable, and so is ranging.

I don't know one way or the other, but are you sure about that?  Having looked at quite a few meter schematics, some pick off the signal for autorange quite a bit earlier and it is entirely possible that interference that causes auto-hunting will be rejected by the actual DC measurement circuit.  My Fluke 8846A is notorious for this, on the lower DC ranges it likes to jump around and flash 'OVERLOAD' but then it takes perfectly stable readings in 3 different ranges.  The test for rejection vs autorange would be to see how stable the actual readings are in manual range mode.  I would expect any decent meter to reject quite a bit of AC in any DC measuring mode, including ohms.

Just for comparison I tried this with my F289 and got very little change with the lead-wiggle/shirt rub and nearly none with the AC coupling.  The autorange has quite a bit of hysteresis, ranging up at 480-490K and ranging down around 450K.  I was not able to get it to waver with anything from 450K to 600K.  The F289 has very good AC rejection--there's one procedure on a calibrator repair that requires zeroing (balancing) out a signal +/- 5mVDC in the presence of 35Vrms, and it works perfectly.  It might be worth testing the Brymen for normal-mode AC rejection.

I measured. If you fix it in 5MOhm range, by injecting signal from siggen, sweeping from 20Hz to 10KHz, it would drop to 460-480 kOhm (resistor value is 564 kOhm).. That is real problem here. Not autoranging.
Autoranging hunting was just a glitch that made us pay attention to the fact that in that resistor range it  doesn't suppress AC interference well...

Suppression of AC component in DC voltage mode is fine. It's just ohm mode..

[bdunham7]

I measured. If you fix it in 5MOhm range, by injecting signal from siggen, sweeping from 20Hz to 10KHz, it would drop to 460-480 kOhm (resistor value is 564 kOhm).. That is real problem here. Not autoranging.
Suppression of AC component in DC voltage mode is fine. It's just ohm mode..

Yes, that seems a problem.  Next I'd want to know how large an electric field is actually needed to cause a significant error, assuming no particular attention is given to test lead routing, as in normal usage.  And does a similar error appear in other ranges?

[dcac]

It’s clear to me BM869 is using dynamic filtering (or oversampling as it often is). So ADC values used to change range has less filtering than the number value that’s finally shown. I notice this as I get reasonable stable and correct value even for 500K-550K but only for a short time then the meter suddenly change range anyway and this of curse destabilizes everything and most of the time causes hunting for a while.     

To compare with first gen firmware for the 121GW which did not use dynamic filtering and also had an added delays after each range change - but this then also took forever to show the final result. This was changed in later FW so it has only one delay when it thinks it found the correct range and also the update rate were increased for some ranges. But overall the 121GW auto range is still slower than BM869 and it’s less stable in some ranges but at least it’s not at all that prone to hunting.

So the dynamic filtering in BM869 is used to speed up auto range - but it’s probably (slightly) too aggressively implemented - so instead of speeding up the auto range it sometime wastes time by hunting.

[EEVblog]

BTW, Dave do you have the BM869 schematics or perhaps you know where in the circuit C28 sits?

No one has schematics for any Brymen meter, not even a sizeable dealer like me.

[EEVblog]

I measured. If you fix it in 5MOhm range, by injecting signal from siggen, sweeping from 20Hz to 10KHz, it would drop to 460-480 kOhm (resistor value is 564 kOhm).. That is real problem here. Not autoranging.
Autoranging hunting was just a glitch that made us pay attention to the fact that in that resistor range it  doesn't suppress AC interference well...
Suppression of AC component in DC voltage mode is fine. It's just ohm mode..

Seems like a very specific issue around that small resistance value band for some reason. Strange.

[2N3055]

I measured. If you fix it in 5MOhm range, by injecting signal from siggen, sweeping from 20Hz to 10KHz, it would drop to 460-480 kOhm (resistor value is 564 kOhm).. That is real problem here. Not autoranging.
Autoranging hunting was just a glitch that made us pay attention to the fact that in that resistor range it  doesn't suppress AC interference well...
Suppression of AC component in DC voltage mode is fine. It's just ohm mode..

Seems like a very specific issue around that small resistance value band for some reason. Strange.

Current that BM869 uses in that range is 100 nA, voltage drop over Rx is 56 mV.. Not hard to upset that.
When I connected MTX3293 in series to measure current through Rx, it was enough to go crazy. But, if I connect 0.1uF capacitor in parallel with Rx, it stops hunting....

I really would like to know where is the famous C28 connected... Is it in the autoranging part of circuit or Ohms ..?

[Martin72]

Not hard to upset that.

Absolutely.
Just a few minutes ago I´ve measure a 560K SMD with the flexible testleads coming with the 869s.
Put the resistor on the table, connect the testpins to it - 560K and stable.
Moving with the knee the flexible leads while holding them to the resistor, value´s getting unstable.
So don´t move the leads and evrything is fine.

[2N3055]

Not hard to upset that.

Absolutely.
Just a few minutes ago I´ve measure a 560K SMD with the flexible testleads coming with the 869s.
Put the resistor on the table, connect the testpins to it - 560K and stable.
Moving with the knee the flexible leads while holding them to the resistor, value´s getting unstable.
So don´t move the leads and evrything is fine.

Hehe are you also using trick with shallow hole on the board for measuring SMD components?

[Martin72]

 Well......yes. 

[bdunham7]

Seems like a very specific issue around that small resistance value band for some reason. Strange.

Is it though?  Or does it just become most noticeable at that point?

2N3055:  Can you repeat the sig-gen experiment using 470K and 680K resistors in the 5M range?  And if/when the injected signal lowers the reading, is the (wrong) reading stable?

[2N3055]

Seems like a very specific issue around that small resistance value band for some reason. Strange.

Is it though?  Or does it just become most noticeable at that point?

2N3055:  Can you repeat the sig-gen experiment using 470K and 680K resistors in the 5M range?  And if/when the injected signal lowers the reading, is the (wrong) reading stable?

456k resistor in series with 25 Ohm (terminated siggen) and 500 mV 50 sinus signal. Signal generator amplitude is P-P.

5 MOhm range:
No generator reads 454,1k, with generator reads 515k when it settles. It takes few seconds to settle, it first shoots up to 550k and then settles down back to 515k and than stays stable there..

500 kOhm range:
No generator reads 452,86k, with generator set to up to 200 mV it fluctuates slowly from 452,0k to 454,00 and back. Movement seems sinusoidal, it is is some sort of aliasing between A/D and input signal..
Higher signal level from generator will start fluctuation of reading,at 300mV periodically going out of range.

Edit: some additions, please reread.

[joeqsmith]

Same here. If someone thinks that it is ok to have 30%  error in measurements as long as autoranging is slow, that is their prerogative.
...
If C28 mod makes meter better reject AC superimposed on circuit while measuring in 500kOhm range, and doesn't interfere with measurement settling time, bandwidth of meter in other ranges or what not, than it would be something that I would be tempted to do. But autoranging doesn't seem to me to be the source of problem..
...

Dave, have you asked Brymen to provide further details about changing this part? 

[cluca1969]

I notice that everyone is silent, no one has anything to say about the BM869s anomaly. Personally I am very affected by this hidden factory defect. The BM869s was the first multimeter which I bought after 27 years, the first was a HUNG CHANG HC-81. I had high expectations from this Brymen, as a result I also bought BM235 and BM857 based on my first experience with BM896s. I was glad if someone tried my temporary method to eliminate this defect / anomaly. I managed to get a provisional result by putting a 100nF capacitor directly at the input of the multimeter. I'm curious about the official position of the Brymen leadership because they can't ignore us indefinitely because this multimeter is their top of the range. I hope that the appearance of the new BM786 model will not divert attention from this problem with BM869s.

[AVGresponding]

I notice that everyone is silent, no one has anything to say about the BM869s anomaly. Personally I am very affected by this hidden factory defect. The BM869s was the first multimeter which I bought after 27 years, the first was a HUNG CHANG HC-81. I had high expectations from this Brymen, as a result I also bought BM235 and BM857 based on my first experience with BM896s. I was glad if someone tried my temporary method to eliminate this defect / anomaly. I managed to get a provisional result by putting a 100nF capacitor directly at the input of the multimeter. I'm curious about the official position of the Brymen leadership because they can't ignore us indefinitely because this multimeter is their top of the range. I hope that the appearance of the new BM786 model will not divert attention from this problem with BM869s.

I think rather than being silent, people are waiting to see what Brymen come back with.

If it's affecting you in your measurements, you can mitigate it greatly by twisting your test leads together. There is still an effect, but once you and your leads, and the DUT have stopped moving, the reading settles, and is accurate.

[Fungus]

I'm curious about the official position of the Brymen leadership because they can't ignore us indefinitely because this multimeter is their top of the range.

The official position was posted earlier.

https://www.eevblog.com/forum/testgear/brymen-bm869-resistance-quirk/msg3342566/#msg3342566

[cluca1969]

I was referring to the silence of the Brymen, not ours. We don't think our little tricks solve this major problem. I have enough multimeters and I don't think I can work without BM869s. I'm still waiting for an official response from Brymen.

[cluca1969]

A picture with a very low resolution is a more than offensive answer to us. I hope Brymen's position doesn't stay that way.
Their solution is rainwater, the trick with the capacitor I discovered before them, so it's not a big deal. I put it at the input and the anomaly disappeared.

[bdunham7]

I think rather than being silent, people are waiting to see what Brymen come back with.

If it's affecting you in your measurements, you can mitigate it greatly by twisting your test leads together. There is still an effect, but once you and your leads, and the DUT have stopped moving, the reading settles, and is accurate.

But is it accurate?  Part of 2N3055's test showed that the extreme amount of interference he injected would result in a stable but grossly erroneous reading, which is different than a noisy or hunting reading.  The unresolved question is does a much smaller amount of interference results in a small but out-of-spec error that would go unnoticed by most users?  Or is is able to reject lower levels of interference and only plays up when the interference reaches a certain threshold?  The former would be a real issue, IMO.  In real life we often work in pretty noisy environments and we don't always know it. 

[AVGresponding]

I think rather than being silent, people are waiting to see what Brymen come back with.

If it's affecting you in your measurements, you can mitigate it greatly by twisting your test leads together. There is still an effect, but once you and your leads, and the DUT have stopped moving, the reading settles, and is accurate.

But is it accurate?  Part of 2N3055's test showed that the extreme amount of interference he injected would result in a stable but grossly erroneous reading, which is different than a noisy or hunting reading.  The unresolved question is does a much smaller amount of interference results in a small but out-of-spec error that would go unnoticed by most users?  Or is is able to reject lower levels of interference and only plays up when the interference reaches a certain threshold?  The former would be a real issue, IMO.  In real life we often work in pretty noisy environments and we don't always know it.

Well, according to my 7075, using a 5-wire guarded lead set, yes.

I didn't calculate out the uncertainties, but a quick look showed it to be good enough for the girls I go out with.

[_Wim_]

I can't bring myself to see it as a "problem that needs fixing", it's just a quirk to be aware of.

Same here. I have 2 BM869s which I use almost daily, and never really noticed it (at least not in a disturbing way that I would remember). As this is a very popular meter and already quite a while on the market, I guess we are also not alone in this...

[2N3055]

I think rather than being silent, people are waiting to see what Brymen come back with.

If it's affecting you in your measurements, you can mitigate it greatly by twisting your test leads together. There is still an effect, but once you and your leads, and the DUT have stopped moving, the reading settles, and is accurate.

But is it accurate?  Part of 2N3055's test showed that the extreme amount of interference he injected would result in a stable but grossly erroneous reading, which is different than a noisy or hunting reading.  The unresolved question is does a much smaller amount of interference results in a small but out-of-spec error that would go unnoticed by most users?  Or is is able to reject lower levels of interference and only plays up when the interference reaches a certain threshold?  The former would be a real issue, IMO.  In real life we often work in pretty noisy environments and we don't always know it.

I measured with FIXED 500k and 5Mohm range to be able to get a reading. Otherwise will hunt.
That is why I said that hunting is actually PREFERED to just silently making erroneous reading.

If autoranging you will know something is wrong, because it will hunt.

In 3 years of everyday use i haven't had a problem. If you don't wrap cables around power cables and transformers on purpose, this is not an issue.

All I see here is that yellow meter fanboys are starting to gloat because "there is finally proof  that Brymen is not as good as they say"....

I'm going to repeat: if you are injecting enough interference to cause this problem into measurement circuit, while measuring Megaohm level resistance, you're doing something wrong.
If you have enough magnetic field in free air around your desk to trigger this, you have bigger problem...

[Martin72]

If you don't wrap cables around power cables and transformers on purpose, this is not an issue.

This is the point, normally nobody would do so while measuring.
This thing will happen under special circumstances and suddenly all people going 

Me, I´ll keep cool and use it like before with no problems so far. 

[2N3055]

I can't bring myself to see it as a "problem that needs fixing", it's just a quirk to be aware of.

Same here. I have 2 BM869s which I use almost daily, and never really noticed it (at least not in a disturbing way that I would remember). As this is a very popular meter and already quite a while on the market, I guess we are also not alone in this...

I agree. It is unwanted quirk. I'm not happy about it but in normal circumstances you will never have problem, if you use normal measurements practice.

[bdunham7]

All I see here is that yellow meter fanboys are starting to gloat because "there is finally proof  that Brymen is not as good as they say"....

Oh come on now.  There hasn't been a peep of that.

I understand the behavior with the gross interference you introduced.  What I'm wondering is what happens when the interference is less, not enough to make the autoranging hunt or for there to be an obvious error in the measurement.  After all, a DMM were off by 2 or 3% on ohms, many if not most people wouldn't notice that for a long time.  Only once we know how much interference it takes to introduce any significant error can it be concluded that there should be no issue under 'normal' conditions--whatever those are.  It may, in fact, not be a big deal at all.

[bdunham7]

So I tried a brief experiment to see if this behavior is found elsewhere.  I connected a Fluke 116, a 470K resistor, a Siglent AWG and a 50R terminator in the manner suggested by 2N3055, so the AWG+terminator are in series with the resistor and the DMM.  I used a 60Hz sine setting and voltages starting at 100mVp-p.  I wasn't looking for autoranging anomalies, but rather the stable but erroneous shift in readings.

With no interference, I had 471.9K.  With 100mV, it dropped one digit, to 471.8.  From then on, in 100mV steps up to 3.9V, the reading varied from 471.8 to 470.4, with no particular correlation that I could see--but what was interesting was that the value for each step was very repeatable, so for example at 900mV I would get 470.8 each time.  Then, above 3.9V, the reading dropped into the 440K range.  It wobbled around but stayed in that range up to 5V.

This meter uses a 1uA constant current for the 600K range, has a specified accuracy (resistance) of 0.9%, a specified normal mode AC rejection of >60dB (in DC volts mode) and is 6000 count.  So it remains well within spec with up to 3.9Vp-p interference, and doesn't work well above that.

The point is, that as I suspected, the meter would reject AC well enough up to the point that it didn't.  I think the reason is that somewhere in the signal chain, if the AC component gets high enough, it will get clipped or clamped, perhaps by an op-amp hitting the rails, or protective diodes clamping it somewhere, or something like that.  When that happens, it may be asymmetric--one polarity gets clamped first and then the difference in those clamp or clip levels creates a DC bias. 

If the Brymen uses a 0.1uA test current at this range, then it has to be 10X more sensitive and that additional gain would result in the sensitivity to the AC component being greater.  So the question is at what level does AC interference like this cause a significant error in the reading and what type of situations would result in interference at or above that level creeping into the measurements?

Just for fun I also tried this on an old 8842A that I'm fixing, which also uses 1uA test current on its 2M range. While 5Vp-p was enough to knock it out of its much tighter specs, even 10V was not enough to induce a gross error--it still read 470.65 or something, as opposed to 471.87 without interference.

[2N3055]

All I see here is that yellow meter fanboys are starting to gloat because "there is finally proof  that Brymen is not as good as they say"....

Oh come on now.  There hasn't been a peep of that.

I understand the behavior with the gross interference you introduced.  What I'm wondering is what happens when the interference is less, not enough to make the autoranging hunt or for there to be an obvious error in the measurement.  After all, a DMM were off by 2 or 3% on ohms, many if not most people wouldn't notice that for a long time.  Only once we know how much interference it takes to introduce any significant error can it be concluded that there should be no issue under 'normal' conditions--whatever those are.  It may, in fact, not be a big deal at all.

Not addressed to you at all..You are being calm and trying to figure out the facts...
All the best,

[2N3055]

So I tried a brief experiment to see if this behavior is found elsewhere.  I connected a Fluke 116, a 470K resistor, a Siglent AWG and a 50R terminator in the manner suggested by 2N3055, so the AWG+terminator are in series with the resistor and the DMM.  I used a 60Hz sine setting and voltages starting at 100mVp-p.  I wasn't looking for autoranging anomalies, but rather the stable but erroneous shift in readings.

With no interference, I had 471.9K.  With 100mV, it dropped one digit, to 471.8.  From then on, in 100mV steps up to 3.9V, the reading varied from 471.8 to 470.4, with no particular correlation that I could see--but what was interesting was that the value for each step was very repeatable, so for example at 900mV I would get 470.8 each time.  Then, above 3.9V, the reading dropped into the 440K range.  It wobbled around but stayed in that range up to 5V.

This meter uses a 1uA constant current for the 600K range, has a specified accuracy (resistance) of 0.9%, a specified normal mode AC rejection of >60dB (in DC volts mode) and is 6000 count.  So it remains well within spec with up to 3.9Vp-p interference, and doesn't work well above that.

The point is, that as I suspected, the meter would reject AC well enough up to the point that it didn't.  I think the reason is that somewhere in the signal chain, if the AC component gets high enough, it will get clipped or clamped, perhaps by an op-amp hitting the rails, or protective diodes clamping it somewhere, or something like that.  When that happens, it may be asymmetric--one polarity gets clamped first and then the difference in those clamp or clip levels creates a DC bias. 

If the Brymen uses a 0.1uA test current at this range, then it has to be 10X more sensitive and that additional gain would result in the sensitivity to the AC component being greater.  So the question is at what level does AC interference like this cause a significant error in the reading and what type of situations would result in interference at or above that level creeping into the measurements?

Just for fun I also tried this on an old 8842A that I'm fixing, which also uses 1uA test current on its 2M range. While 5Vp-p was enough to knock it out of its much tighter specs, even 10V was not enough to induce a gross error--it still read 470.65 or something, as opposed to 471.87 without interference.

That is nice data, thanks.
Just to make sure, on higher ohm ranges, with low current, meters should have same sensitivity to injected voltage amplitude as on lower ranges, when voltage is injected directly like that from low impedance circuit. It is all about proportion of interference voltage and working voltage drop on Rx at that range .
Difference is in normal work where low current and high resistances make circuit high impedance so those ranges are more likely to pick up interference, and induce higher amplitudes of interference.

This goes well with my theory that problem lies with AC signal suppression in what is supposed to be DC measurement.
From what I can see (maybe over weekend I'll have a bit more time to do more investigation), common mode disturbances are suppressed well enough.

I suspect rectification happens somewhere in a circuit (maybe protection components, or current source  or measurement path itself). I want to try to quantify how much error is there, and more importantly what is direction of error...  I will be sure when I compile a table, but my quick estimate is that error is low if it doesn't hunt in autoranging. That is why I keep repeating that.

Of course, that is for exact resistance I used. I would need to try it systematically with a range of resistors, try to maybe inject signal trough some signal transformer to reduce coupling to ground, and generally make a systematic, controlled experiment. I'm very busy at the moment, so I'll see if I can organize something. No promises. If I do I will publish details here.

[Dr. Frank]

Hi,
On my BM 869, I had a closer look on 3 separate circuit aspects of this effect:

- ranging method
- stability of the Ohm measurement function
- over range detection

I also see this annoying effect, even in my basement lab, which is relatively quiet concerning EMI, and even w/o any provoked AC mains signal injection.

Down-ranging always occurs at a fixed / digital count of about < 5000 counts.
Up-ranging occurs at random counts. In DCV mode, that's about > 53125 counts, in Ohm modes it's about > 53300 counts.
In the first place,  I tripped about other contributions here, which reported >56000 counts, or so.

Up-ranging is therefore initiated by an over range condition, not by a fixed upper digital A/D value, like in other DMMs (e.g. 1 200 000 ... counts for all HPAK bench DMMs).
Obviously that over-range detection inside the BM869 is realized analogously by a window comparator (OpAmp) which probably monitors directly the input signal over an additional input resistor.

In fixed 500kOhm range, as far as no over range occurs, the measured values between 500 and 530kOhm are always quite stable, due to the averaging characteristics of the multislope A/D. Of course, for any DMM with such a low sensing current of 1µA, you can easily provoke this reading to fluctuate by injecting large AC signals. Latter I would regard as being completely normal, and not being a 'Resistance Quirk' at all.

Approaching these 53300 counts, the BM869 quickly switches between OL and a mostly correct measurement value.

So the culprit lies only in this analogue overload comparator, which reacts much too sensitively on induced EMI signals.
I assume, that the additional capacitor simply gives a low pass RC filter to suppress that effect, making OL detection slower, but would probably not greatly affect the measurement.


Btw.: We have discussed such OL detection problems as well on the 121GW for AC+DC signals, but the other way round.
There, the OL detection is done digitally only, which leads to an undetected OL condition and a saturation of the input buffer, and in turn to (partly dangerous) false readings.
There, an additional analogue OL detection would have solved the problem.

So in the end, this is a weak design on a very limited use case, but it does not greatly affect the usability of this DMM.. simply switch to manual ranging under this condition.

Frank

[Fungus]

simply switch to manual ranging under this condition.

This.

[dcac]

Obviously that over-range detection inside the BM869 is realized analogously by a window comparator (OpAmp) which probably monitors directly the input signal over an additional input resistor.

Yes that’s an interesting solution giving optimum speed for the range detection. But it’s also much more sensitive to noise as the comparator have to rely solely on analogue filtering and cannot benefit from the digitally Sinc averaged values from the ADC.

This solution also explains why you can have an almost stable reading on the number display but then suddenly auto range kicks in anyway if its close to the tipping points of the comparator.

The effect others describe where a higher injected noise signal will cause the number value to shift up or down - is likely due to clipping if the signal exceeds positive or negative Vref for the ADC. The Sinc average filter is only effective as long a the noise (sine) wave can sampled correctly without clipping.

[Martin72]

Hi,

simply switch to manual ranging under this condition.

When I remember it right, as I took the brymen to work for measure on a resistor decade, fixing the range won´t fit the "problem".
But I can do it again next week.
Apart from this, normally I´m testing resistors out of the circuitry, on the bench and got no problems at all, as I don´t have any radiation fields here or "dancing" around the table while putting the probes on the DUT.
My 869s is calibrated according to ISO without any problems.
Next month, I´ll order my second 869s.

[bdunham7]

So in the end, this is a weak design on a very limited use case, but it does not greatly affect the usability of this DMM.. simply switch to manual ranging under this condition.

I agree that it is likely not a huge issue in normal usage, but I'd object to the conclusion that manual ranging is a complete solution, at least at this point.  joeqsmith has actually done some tests but they are in the meter robustness thread, so anyone that is interested can look there.  The issue I have is that at some level of interference, there appears to be a constant induced error in the resistance ranges that results in an apparently valid, steady, reading--but one that is wrong.  Before this thread I would have assumed that this would only happen at highly abnormal EM field levels, but now I'm not so sure and it might be worth quantifying both the levels of interference required to get a wrong reading and the EM field levels that would be needed to cause that interference with typical test setups.

[Wytnucls]

The meter passed the 3 A/m EM field strength compatibility already.

[cluca1969]

No matter what tests were or will be done by us, Brymen had to be aware of these fields and interferences.
I personally am still waiting for an official response from Brymen.

[bdunham7]

The meter passed the 3 A/m EM field strength compatibility already.

What does that mean?  As in--does the test involve just the meter or the test leads too, and what required for it to pass the test?

[Martin72]

Here it is...

Read the norms, then you´ll find it.

[bdunham7]

Here it is...

Read the norms, then you´ll find it.

Gee thanks.  For only $375 per standard I can read them.  Maybe there's a summary somewhere so I can whittle it down to 3 or 4 to keep it affordable?

EDIT: OK, it looks like of the ones it lists, EN61000-4-8 is the relevant one. https://www.iecee.org/dyn/www/f?p=106:49:0::::FSP_STD_ID:4229

Anyone have a copy?  Does the test require the instrument to remain accurate to specs even with external leads and devices, or just not fail or quit working?

[Wytnucls]

The meter passed the 3 A/m EM field strength compatibility already.

What does that mean?  As in--does the test involve just the meter or the test leads too, and what required for it to pass the test?

Severity Level 2   3 A/m
Performance criterion required: A
Criterion A: The multimeter shall continue to operate as intended without operator intervention

Multimeter modes: Voltage, Frequency, Resistance, Current, PC connect
Multimeter placed inside induction coil (1mx1m)
Coil orientation XYZ in turn
Test duration 5 minutes in each direction

[_Wim_]

Anyone have a copy?  Does the test require the instrument to remain accurate to specs even with external leads and devices, or just not fail or quit working?

Indian standards are a good read...

https://www.eevblog.com/forum/metrology/emi-measurements-of-a-volt-nut/msg3308348/#msg3308348

[dcac]

Down-ranging always occurs at a fixed / digital count of about < 5000 counts.
Up-ranging occurs at random counts. In DCV mode, that's about > 53125 counts, in Ohm modes it's about > 53300 counts.
In the first place,  I tripped about other contributions here, which reported >56000 counts, or so.

Up-ranging is therefore initiated by an over range condition, not by a fixed upper digital A/D value, like in other DMMs (e.g. 1 200 000 ... counts for all HPAK bench DMMs).
Obviously that over-range detection inside the BM869 is realized analogously by a window comparator (OpAmp) which probably monitors directly the input signal over an additional input resistor.

I think what we’re seeing actually is digital switching points for the Up-ranging after all.

What gives the impression of somewhat diffuse switching points that doesn’t seem to occur at the same value each time - is caused by noise and/or low (lower) resolution from the ADC.

What I meant earlier by “dynamic filtering” should rather be called “two stage filtering”. 121GW uses this too - the first stage is the HY3131 Sinc2 filter which is set to give 40 SPS output rate and is used for the Bargraph. The same value is then filtered by the STM32 MCU but with a simple Sinc1 stage and this gives the display output rate of 5 SPS. And 121GW uses this 5 SPS rate + some delays for the auto range which makes it rather slow.

I believe BM869 is configured in a similar manner - but here the first stage filter gives 80 SPS for the Bargraph but this value is then also used for the auto range and will fluctuate just like the Bargraph does when subjected to interference. Then for the second stage filter BM869 likely uses a Sinc4 (or higher) as it has quite noticeably better noise rejection than i.e. 121GW.

It’s also possible the auto range has its own filter settings but one that’s still not nearly as effective as for the final value shown on the display - all to make the auto range faster. But doing it this way the display will never show the exact value that actually caused the range to change.

If BM869 is in fact configured like this it should also be fixable in firmware. It quite strange though that current FW allow auto range to act this fast and hunt back and forth between two ranges without some kind of damping routine i.e. stopping the hunt if it changed back and forth several times within a couple of seconds.

[bdunham7]

Severity Level 2   3 A/m
Performance criterion required: A
Criterion A: The multimeter shall continue to operate as intended without operator intervention

Multimeter modes: Voltage, Frequency, Resistance, Current, PC connect
Multimeter placed inside induction coil (1mx1m)
Coil orientation XYZ in turn
Test duration 5 minutes in each direction

OK, I guess that is just based on the verbiage of the standard.  I found this:  http://www.compliance-club.com/pdf/EMCTestingPart6.pdf
which indicates that 'industrial' meters may be tested to 30A/m, if not more.  Out of curiousity, I got one of those hand-held trifield EF/EMC/RF Power meters and checked around my bench and office.  There are areas that exceed 3A/m (4.75uT) and when I checked on my bench, I just tested near 2 DMMs where they happened to be sitting at the moment, hooked up and testing something.  The highest one read 24A/m (30uT) and 99% of that was caused by the transformer in my Hakko FX888, which was about a foot away.

So, compliance with a nominal standard doesn't imply that it just barely met that standard--it could be a lot better.  But apparently being immune to 3A/m wouldn't keep you out of trouble, at least on my bench.

[Wytnucls]

The point is that manufacturers can't be held responsible if their meters are used in an environment that exceeds the IEC standards.
The onus is on the operator to comply with safety requirements.

[bdunham7]

The point is that manufacturers can't be held responsible if their meters are used in an environment that exceeds the IEC standards.
The onus is on the operator to comply with safety requirements.

I'm more interested in the technical issues and determining what the actual performance issues are here, not assigning blame or denigrating or promoting products..  As far as standards, they are just minimums.  You don't have to live down to them.  And the onus is not on the operator to use the meter within its standard, but rather to select a meter that meets their particular requirements even if those requirements exceed some common standard.

[Wytnucls]

The standards are minimum published safety requirements that manufacturers have to comply with.
While meters may well exceed those requirements, measurement accuracy and operator safety is not documented.

Electromagnetic Compatibility: In an RF field of 3 V/m total accuracy = specified accuracy + 20 counts (Fluke)
What is the accuracy at 30V/m?

[bdunham7]

The standards are minimum published safety requirements that manufacturers have to comply with.
While meters may well exceed those requirements, measurement accuracy and operator safety is not documented.

Electromagnetic Compatibility: In an RF field of 3 V/m total accuracy = specified accuracy + 20 counts (Fluke)
What is the accuracy at 30V/m?

Well, I don't know and I would like to.  'Not documented' can happen for a variety of reasons, not the least of which is the legal department discouraging the publication of any more information than is necessary.  Every claim you make becomes a potential liability.

The link I had was for power line EM fields at 50/60Hz and A/m.  However, my trifield gizmo measures V/m as well and if that is an actual spec from Fluke, then the question is certainly germane because I also see EF readings well over 3 V/m all over the place.  One point in my office had 981V/m, but that is an extreme example and I'm not likely to be making a measurement right in front of my air purifier.

[2N3055]

The standards are minimum published safety requirements that manufacturers have to comply with.
While meters may well exceed those requirements, measurement accuracy and operator safety is not documented.

Electromagnetic Compatibility: In an RF field of 3 V/m total accuracy = specified accuracy + 20 counts (Fluke)
What is the accuracy at 30V/m?

Well, I don't know and I would like to.  'Not documented' can happen for a variety of reasons, not the least of which is the legal department discouraging the publication of any more information than is necessary.  Every claim you make becomes a potential liability.

The link I had was for power line EM fields at 50/60Hz and A/m.  However, my trifield gizmo measures V/m as well and if that is an actual spec from Fluke, then the question is certainly germane because I also see EF readings well over 3 V/m all over the place.  One point in my office had 981V/m, but that is an extreme example and I'm not likely to be making a measurement right in front of my air purifier.

 981V/m would for sure be dangerous to humans..

[joeqsmith]

... because I also see EF readings well over 3 V/m all over the place.  One point in my office had 981V/m, but that is an extreme example and I'm not likely to be making a measurement right in front of my air purifier.

There is a major story here that you need to tell!

Many years ago, I was working on a design for 100V/m.  Something was going on and I had to use an outside lab.  Ended up at Cincinnati Electronics's in Mason.  In the main entryway there was a full size model of a missile.  Their lab at that time could support up to 1000V/m.   You don't want the radar knocking out your missiles electronics.   I went on about my humble 100V/m.     

[bdunham7]

[2N3055]

(Attachment Link)

I have no doubt you can read...

BUT....

Maybe that instrument is reading static electricity field from that IONIZER in the purifier....... ??

[bdunham7]

(Attachment Link)

I have no doubt you can read...

BUT....

Maybe that instrument is reading static electricity field from that IONIZER in the purifier....... ??

Yes, hopefully and obviously, it is a static field.  If it were RF you wouldn't need a radio, or at least you wouldn't need to turn it on! V/m doesn't actually tell you the whole story--frequency counts.  You can see pretty high V/m just from power line effects.  Also, just the great outdoors on a sunny day will give you a vertical field of over 100V/m.  None of those things would matter much to most DMMs, but the air purifier and my Keithley 414a are best friends--the Keithley can tell if the ionizer is on from quite a distance.  The air purifier just sits in the corner quietly until it gets dirty and usually doesn't bother anything, but I have to shut it off a good while before I do anything with the picoammeter or the 1G range on my DMM.

I've no idea how accurate the instrument is, but it seems to identify things correctly and at least sort of matches up with my preconceived notions of what familiar things should emit.  It does tell me that where I sit at my desk I'm exposed to 57V/m of 'mixed' EF and 13mW/m² of RF.  I'm not sure what to think about that.

[2N3055]

(Attachment Link)

I have no doubt you can read...

BUT....

Maybe that instrument is reading static electricity field from that IONIZER in the purifier....... ??

Yes, hopefully and obviously, it is a static field.  If it were RF you wouldn't need a radio, or at least you wouldn't need to turn it on! V/m doesn't actually tell you the whole story--frequency counts.  You can see pretty high V/m just from power line effects.  Also, just the great outdoors on a sunny day will give you a vertical field of over 100V/m.  None of those things would matter much to most DMMs, but the air purifier and my Keithley 414a are best friends--the Keithley can tell if the ionizer is on from quite a distance.  The air purifier just sits in the corner quietly until it gets dirty and usually doesn't bother anything, but I have to shut it off a good while before I do anything with the picoammeter or the 1G range on my DMM.

I've no idea how accurate the instrument is, but it seems to identify things correctly and at least sort of matches up with my preconceived notions of what familiar things should emit.  It does tell me that where I sit at my desk I'm exposed to 57V/m of 'mixed' EF and 13mW/m² of RF.  I'm not sure what to think about that.

I'm just saying ...

That static field would not impact BM869S much, unless in nS mode....

The only thing that would have major impact would be induction in wires from magnetic field..
And that only if it is not common mode..

[Martin72]

Today I did a normal resistance measure in practice, without any disturbances - No wonder, that this "quirk" wasn´t detect in the first time.
Also today, I´ve finally wrote to welectron what´s about this quirk from brymen´s official side, because I want to order a second one.
Apart from this, I got another 100Khz multimeter for free
A fluke 187 was sorted out last week because of fading LCD segments (I know what to do with this) and I raised my hand and shout, here! to me ! 

[2N3055]

F 187 ? Nice!!!
It is same as 189 except logging, which wasn't much anyway, only 100 points..
Congratulations, St. Nicholas was good to you this year.. You must have been good  .

[joeqsmith]

(Attachment Link)

Ok, not too much of a story.   meter = distance.   More it a meter away from the source and you read??

[bdunham7]

Ok, not too much of a story.   meter = distance.   More it a meter away from the source and you read??

Well, V/m doesn't imply an actual whole meter.  981mV/mm would be the same field.  But yes, the field drops rapidly as you move away.  The real field is inside the unit between the plates and what I'm reading is just a bit of leakage loops.  And I don't actually know that it is a static field--when the unit gets dirty it snaps and hums, so maybe rectified but not filtered DC? 

[joeqsmith]

Ok, not too much of a story.   meter = distance.   More it a meter away from the source and you read??

Well, V/m doesn't imply an actual whole meter.  981mV/mm would be the same field.  But yes, the field drops rapidly as you move away.  The real field is inside the unit between the plates and what I'm reading is just a bit of leakage loops.  And I don't actually know that it is a static field--when the unit gets dirty it snaps and hums, so maybe rectified but not filtered DC?

I thought the standards call for a meter minimum to setup the UFA, 3 meters preferred?

[bdunham7]

I thought the standards call for a meter minimum to setup the UFA, 3 meters preferred?

I wouldn't know, but sure, maybe.  If you want a uniform field for some standard test, you'd need a big area.  Same for the EM test, big coils to get a uniform area in the center.  But that's just a standard, and standards are designed to be, well, standard--accurate, repeatable tests, not to necessarily a reflection of any real-world situation.  I'm sure most real-world EMC issue don't involve huge, uniform fields. 

[kwinz]

So no firmware/software update for this?
But I can solder an SMD capacitor in parallel to fix it without messing up the multimeter calibration?