Brymen BM789

[2N3055]

But does the 869 correctly shows "OL" for AC in this scenario?  Did you try to switch polarity?

With the 869  it makes sense to have everything on the same switch position, but with the single display if there is no way to show overrange they certainly could put in a separate AC mV position.

I think we have a language barrier here.

BM869S has 2 different positions for DC and AC mV, but both are nevertheless still DC coupled .

It does show OVL if you apply larger AC voltage. If you combine it with DC, that combined goes over range it also behaves funny.

For instance 1V P-P 400HZ square wave, will show correct 500mV RMS. If I add more than 0.6 V DC to it (offset) it will start showing wrong numbers. Funny enough, I can add -2,3V offset before it starts making problems. So it is not symmetric.

But that is all getting into non important detail. Instrument shouldn't be used outside it's specified range. Good practice is to measure in Volts range all the time and only jump to mV range when we know there won't be hidden problems.

[2N3055]

Other even cheaper ones (UT81E and Aneng 8009) don't, they correctly display OL (kudos to them). Fluke 189 (in AC mV) is not tricked by a high DC offset either. Fluke 87V doesn't have this issue at all because it doesn't have a separate AC mV position on the dial in the first place 

Are you sure about F87V ? Connect other meter in ohms mode and measure input resistance of F87V. Go through modes and ranges. Not only input resistance won't be constant 10 MOhm, but when it goes Hi impedance then it is AC coupled.

I only meant that Fluke 87V doesn't have an "AC mV" mode on the dial at all, only a "DC mV" one (which has a single 600 mV range). It does measures AC mV in "AC V" mode, which goes down to 600 mV range (and all AC ranges are AC coupled). The lowest DC V range in "DC V" mode is 6 V, hence the separate "DC mV" 600 mV DC mode.

Not sure why they did that - maybe it was their way of dodging the bullet. They must have figured the problem and this is how they avoided it. They don't call themselves Fluke for no reason   

Doing AC mV right is not easy. And also, we keep forgetting, as much as people love F87V, it is not a meter for electronics. It is industrial electrician workhorse meter. That is what Fluke says too. For that audience, AC mV is not a thing.

[bdunham7]

he lowest DC V range in "DC V" mode is 6 V, hence the separate "DC mV" 600 mV DC mode.

Not sure why they did that

It would be interesting to see what its capability for normal mode AC rejection is in that mVDC range.

[Caliaxy]

Other even cheaper ones (UT81E and Aneng 8009) don't, they correctly display OL (kudos to them). Fluke 189 (in AC mV) is not tricked by a high DC offset either. Fluke 87V doesn't have this issue at all because it doesn't have a separate AC mV position on the dial in the first place 

I'll have to take back the statement that UT81E doesn't have this issue. I just realized that 500 mVrms in my quick test was way over its AC mV range (200 mV or so), which might have contributed to its displaying OL. At 100 mVrms it also displays nonsense at DC offsets higher than 2V (but it seems a bit more tolerant to DC bias than the Brymens). Aneng 8009 might behave similarly (haven't rechecked, but its upper AC mV limit is only 100 mV or so...).

Fluke 189 works fine (and so does Keysight U1282A, if anyone wonders; U1252A doesn't).

[bdunham7]

Well, I guess part of the answer is "RTFM"!  At least they've documented this issue.  I still think the OP may have an issue though, as I think it should indicate OL, but I'll await someone else chiming in with some tests.  The F116 and F189 don't have this limitation--I tried a 0-2V 5% duty cycle signal (CF~5) and they both showed approximately the expected readings. 

[Neutrion]

I think we have a language barrier here.

BM869S has 2 different positions for DC and AC mV, but both are nevertheless still DC coupled .

It does show OVL if you apply larger AC voltage. If you combine it with DC, that combined goes over range it also behaves funny.

For instance 1V P-P 400HZ square wave, will show correct 500mV RMS. If I add more than 0.6 V DC to it (offset) it will start showing wrong numbers. Funny enough, I can add -2,3V offset before it starts making problems. So it is not symmetric.

But that is all getting into non important detail. Instrument shouldn't be used outside it's specified range. Good practice is to measure in Volts range all the time and only jump to mV range when we know there won't be hidden problems.

So if I understand you right, the 869s also does this or something similar? What does it show on the AC mV range if you feed in a similar 30V p-p squarewave?

Again I don't want to use the meter above its specified range (1 V p-p offset included) all what I would like, that the meter shows me that it overranged. Like it shows it correctly on DC mV or AC+DC mV.

Othervise when one measures around 5-600mV AC(correctly, because that is still within its range), and something goes wrong, and let's say 230V appears, but one won't have any clue about that. Because the meter shows almost exactly the same values.
But even a 30 V extra can be quiet problematic of course.

My question about the meter was also about the possibilities of showing "OL" above the 1V p-p level, and not whether it would be possible to measure mV AC on top of some dc offset above the specified 1V p-p treshold.

[2N3055]

I think we have a language barrier here.

BM869S has 2 different positions for DC and AC mV, but both are nevertheless still DC coupled .

It does show OVL if you apply larger AC voltage. If you combine it with DC, that combined goes over range it also behaves funny.

For instance 1V P-P 400HZ square wave, will show correct 500mV RMS. If I add more than 0.6 V DC to it (offset) it will start showing wrong numbers. Funny enough, I can add -2,3V offset before it starts making problems. So it is not symmetric.

But that is all getting into non important detail. Instrument shouldn't be used outside it's specified range. Good practice is to measure in Volts range all the time and only jump to mV range when we know there won't be hidden problems.

So if I understand you right, the 869s also does this or something similar? What does it show on the AC mV range if you feed in a similar 30V p-p squarewave?

Again I don't want to use the meter above its specified range (1 V p-p offset included) all what I would like, that the meter shows me that it overranged. Like it shows it correctly on DC mV or AC+DC mV.

Othervise when one measures around 5-600mV AC(correctly, because that is still within its range), and something goes wrong, and let's say 230V appears, but one won't have any clue about that. Because the meter shows almost exactly the same values.
But even a 30 V extra can be quiet problematic of course.

My question about the meter was also about the possibilities of showing "OL" above the 1V p-p level, and not whether it would be possible to measure mV AC on top of some dc offset above the specified 1V p-p treshold.

Problem is that by pushing 30V in mV range meter circuitry is overloaded in such a way that not even OVL detection works anymore.
OVL is simply coming from ADC overrange register. If preamps are overloaded in such a way that that they are producing something inside ADC input range, meter (ADC) "doesn't see" overrange. That is simple fact based on fact that you need amplifier in front of ADC for sensitive range.

To circumvent around that, a separate, parallel circuitry could be added that would supervise input and monitor for overload. That circuit could be even used for autoranging.. But that would just add to the price. It is responsibility of the user not to connect meter to 10000V in millivolt mode.
It is your responsibility to know the 230V cannot appear from nowhere. Not only because of meter (it will survive) but because of you. You have to have clue. Or big badaboom...

You should be using old moving coil meter for 6 months and you would understand me well.
On those, OVL indicator was smoke coming from it...

Don't get me wrong, it would be nice if they could make completely foolproof meter.. But with good measurement practice that is not necessary.
Also good measurement practice takes into account your safety. Which always comes first.

[bdunham7]

OVL is simply coming from ADC overrange register. If preamps are overloaded in such a way that that they are producing something inside ADC input range, meter (ADC) "doesn't see" overrange. That is simple fact based on fact that you need amplifier in front of ADC for sensitive range.

To circumvent around that, a separate, parallel circuitry could be added that would supervise input and monitor for overload. That circuit could be even used for autoranging.. But that would just add to the price. It is responsibility of the user not to connect meter to 10000V in millivolt mode.

I think some systems do use as separate path for the autorange to make it faster, but it is filtered.  You'd need to tap into it earlier for the overload indicator.  Some meters are better than others at dealing with this--try a 50Hz 1% duty cycle 0-5V square wave if you want a brutal test.  (CF = 10) Even my 8846A didn't range correctly on AC with that one.  However, the issue with the OPs example is that it seems to be a pretty basic example that I wouldn't expect it to not perform well at--but there it is in the specs.  A crest factor of only 1.6 might not be enough to measure some modified sine wave inverter outputs or triac circuits, among other things.

[joeqsmith]

1.If I use the mV Scale and feed in a squarewave with 50% duty cycle without(or very minimal like 100-200mV) zero crossing ...

I started to read your original post but the was not sure with all the missing details.  mV (rms, p-p, average, DC).  I wonder is scale DC, AC, AC+DC.   Want me to try and guess?    What does zero crossing mean to you?  Maybe use terms like offset.   

I have no problems running what ever test cases you want but you need to be VERY clear about what you want to see.   Even a scope shot would help.   


****
So, the BM789 is set to ACmV autorange.  You inject a 40Hz to 500Hz, 50% duty squarewave with no offset and 100mVp-p (+/-50mV) and the meter reads fine.  You then increase the amplitude to 1Vp-p and your meter displays 650mV.  You then continue to drive it up to 30Vp-p and the meter continues to display 650mV.   Then you flip the leads (even though there is no offset and they waveform is symmetrical) and the meter starts to read correctly.   You have verified the wavefrom you are injecting by other means (scope) and know it to be symmetrical.   

Then you go onto to talk about  " I assume it depends on which polarity the squarevawe has."  which now I am thinking it is not a symmetrical waveform but rather switching from 0 to some peak value.       Then later you mention  " (1 V p-p offset included) "   

I thought if I waited and reread your post I could make sense of it.   My guess is you are adding an offset in which case, I bet you could make the meter do all sorts of strange things and I can believe that flipping the leads would make a difference.   A bit of an odd use case but I won't judge. 

Maybe come at it from another direction and post your actual requirements.  It seems you can't use a blocking cap which is typical and you don't want to have to get into the practice of starting in the high ranges.   So, I would say it sounds like you bought the wrong tool for the job. 
So maybe start with the requirements and someone could maybe recommend something. 
 

[Neutrion]

You should be using old moving coil meter for 6 months and you would understand me well.
On those, OVL indicator was smoke coming from it...

Don't get me wrong, it would be nice if they could make completely foolproof meter.. But with good measurement practice that is not necessary.
Also good measurement practice takes into account your safety. Which always comes first.

Obviously I don't use the meter with starting on the millivolt scale, but things can happen in life (at least with me) which we don't foresee so I would be rather on the safe side when measuring something unknown, and be sure that the values around and above 500mV AC are also corect, and I don't have to worry that the reality is something entirely different. With a dual display, as I said it might won't be an issue as you could see on the DC or AC+DC value, that you are out of range but with the single display this might pose a problem.  So this is just safety, like the fuses in your meter, which are also not supposed to be blown- at least regulary.

Joe:

Without zero crossing I meant an offset, so it is an assymetrical squarewave. Meter scale is at AC mV.

An example:Assymetrical squarewave duty cycle 50% p-p 5,28V average 2,52V max 5,12V min -160mV rms 3,55 V
AC V scale correct display: 2,5-2,7 V(STdev on Siglent) which is correct, mV scale: 650mV
Will add a scopeshot if needed later on.
At 1V p-p the meter measures about correctly(not 650mV) so I am not talking about it not meeting the specs, things start to happen if I go above the limit.

I don't have a signal gen, so I can't just test it with different offset levels. But within the 1V p-p including offset there seems to be no problem with the measurement. So the specifications are seemingly met.
I can not tell you what it does if I feed in the same signal without offset and raise it above the 1V p-p level, but it seems, that at least if I go above the specified limit with an assymetrical signal (which again I in a normal case obviously don't want to do, but could happen during a measurement) the AC mV scale does not show that it overranged, but stays like that at around 650mV and changes only minimally.
If I flip the leads it shows "OL" what should be the correct behaviour.
So with changed polarity I meant that the sqarewave is assymetrical signal starts almost at zero, and goes up to above 1V, positive, so the meter gets the signal with a given polarity, but if I flip the leads the meter gets the signal with the opposite polarity.

[joeqsmith]

Now that it is clear you are adding a DC offset,  none of what you wrote surprises me.     At first I thought it was the wrong tool for the job but not posting your requirements, it seems you're just playing around with your new toy.   Most of the responses here seem correct. 

[Fungus]

AC V scale correct display: 2,5-2,7 V(STdev on Siglent) which is correct, mV scale: 650mV

When I teach my basic multimeter class I always tell them to use the 10A input jack for measuring current and only move to the mA input jack if they've seen a signal too small to measure properly on the 10A scale.

The same reasoning applies here too. mV/mA ranges are designed for sensitivity, not to handle overloads.

Now that it is clear you are adding a DC offset,  none of what you wrote surprises me.

You've done several similar tests on bigger scale, adding hundreds of volts of offset on the ACV scale. Many meters fail to handle it gracefully.

it seems you're just playing around with your new toy.

Which is fine! It's good to know the limitations of the tools we own, playing is a good way to find them.

[joeqsmith]

Now that it is clear you are adding a DC offset,  none of what you wrote surprises me.

You've done several similar tests on bigger scale, adding hundreds of volts of offset on the ACV scale. Many meters fail to handle it gracefully.

I certainly have and in cases where the meter doesn't want to switch to the next range,  it makes for a less than impressive review.   I was glad Brymen's engineers were able to sort it out before the meters release.   

The BM789 is not a bad little meter.  Add a few changes like split display and I think I would have a new favorite.    Considering the abuse that both BM869s have seen and their continued trouble free service,  its proving to be more difficult to find something I would take over them.

[Neutrion]

Now that it is clear you are adding a DC offset,  none of what you wrote surprises me.     At first I thought it was the wrong tool for the job but not posting your requirements, it seems you're just playing around with your new toy.   Most of the responses here seem correct.

So do you agree with the others, that there is no way of showing the overload under these circumstances?
Does this happen with every major brand with similar funcions?
Playing: Yes, it is not smart to start using a meter or tool if you are not avare of its limitation, or functions so I am just playing. But watched you playing with the 786 before I bought it
Thats why I was curious of your oppinion.

AC V scale correct display: 2,5-2,7 V(STdev on Siglent) which is correct, mV scale: 650mV

When I teach my basic multimeter class I always tell them to use the 10A input jack for measuring current and only move to the mA input jack if they've seen a signal too small to measure properly on the 10A scale.

The same reasoning applies here too. mV/mA ranges are designed for sensitivity, not to handle overloads.

I also do that with the 10A scale, or actually using a clamp meter if possible, and oviously generally not starting to use the meter on the millivolt scale but again, the problem is that measured values can change while one measures them, and even values in the mV range could suddenly change even dramatically, even if they are not supposed to. So I am only missing the overrange indication and asking whether it would really be impossible with the FW update to solve this, or it would require a major HW update.

But I start to have a feeling that what I wish for is something extraordinary .
That is why I was curious how other meters (with a mV scale of course!) would perform here.

If we are talking about best practice, with the 789 it seems, that if one reads anything on the mV AC range higher than 500mV one should not trust those numbers, and constantly doublecheck it on an other scale, or in the dc range as well. Just as a sidenote. If no FW update is possible for this.

[Neutrion]

I certainly have and in cases where the meter doesn't want to switch to the next range,  it makes for a less than impressive review.   I was glad Brymen's engineers were able to sort it out before the meters release.   

The BM789 is not a bad little meter.  Add a few changes like split display and I think I would have a new favorite.    Considering the abuse that both BM869s have seen and their continued trouble free service,  its proving to be more difficult to find something I would take over them.

This missing split display was for me the only major point why I could not decide for a while between the 789 and the 869. Is there any major technical reason they decided not to go that way?
Or is it that in some industrial application maybe the big numbers are the number one priority?
I also really like the warm white backlight, which is uniqe.

[bdunham7]

mV/mA ranges are designed for sensitivity, not to handle overloads.

Typical designs of AC/mVAC ranges include blocking capacitors at some point to prevent any overload by a DC bias.  This is how you would read ripple on a power supply, right?  This meter is atypical, although documented in the specs, and would be unable to read power supply ripple in the mVAC range without an external blocking capacitor.  This may not be a deal killer for most users, but it is a shortcoming that cannot be overcome with 'correct technique'.  You have to add an external component or move up to a higher range and lose resolution.  So now we know.

You've done several similar tests on bigger scale, adding hundreds of volts of offset on the ACV scale. Many meters fail to handle it gracefully.

That was simply a ranging issue.  The vast majority of meters will not be able to handle over-range normal mode AC rejection, it's just an inherent limitation that can't be easily solved with one component.  The Brymens can't handle over-range normal mode rejection either, but the autoranging appears to respond to the AC enough to move up to a higher range, at least under those test conditions.  That's an advantage of certain types of fast autoranging circuits. 

[AndrewBCN]

...
Which is fine! It's good to know the limitations of the tools we own, playing is a good way to find them.

Except in this case it's not exactly a "limitation" of the BM789. The OP stumbled upon an edge case which is very similar to an operator error, where the BM789's autorange algorithm doesn't compute an appropriate response. The obvious "fix" for this issue is not to reprogram the MCU inside the multimeter (because there is an almost unlimited number of similar edge cases), but to operate the Brymen BM789 following standard good practices, which even a beginner can learn is less than 15 minutes.

Also note that the edge case / wrong operation of the BM789 that the OP stumbled upon is completely different from the methodical stress testing that joeqsmith performs on his DMMs. Joe simulates real-life accidents and situations that the meters are supposed to be able to cope with, within and sometimes beyond their specifications.

Joe stresses the hardware and points to flaws in the design and safety of DMMs, the OP just found an operator error that the programmers at Brymen didn't account for in their autorange algorithm.

I very much enjoy Joe's videos where he goes through various steps to test his DMMs, because what I see is a talented engineer making sure his tools are fit for the job, and this information is very valuable. An operator error that leads to a DMM incorrect display is anecdotal information of next to zero value.

[Fungus]

I also really like the warm white backlight, which is uniqe.

OT, but... Brymen do some orange backlights, too, eg. my BM857. 

[bdunham7]

So do you agree with the others, that there is no way of showing the overload under these circumstances?
Does this happen with every major brand with similar funcions?

To be clear, since you are implying what 'others' agree on and I'm an other, I don't agree and think the meter should show an overload indication.  Whether that's a design issue or a one-off defect with your unit, I don't know.

And no, this doesn't happen with most meter designs AFAIK, at least not the ones I have with an explicit mVAC function.  I think it is a result of 'bangforbuckitis' or the manufacturer trying to add as many features as possible at a price point.  Something has to give.

If we are talking about best practice, with the 789 it seems, that if one reads anything on the mV AC range higher than 500mV one should not trust those numbers, and constantly doublecheck it on an other scale, or in the dc range as well. Just as a sidenote. If no FW update is possible for this.

This shouldn't be necessarily be dealbreaker, its just another characteristic you have to learn to deal with--if your individual meter is not, in fact, defective.  I think the easiest way to guard for it would be to always take your reading initially  in mVAC+DC, then if it reads in range you can switch to mVAC and get the correct reading.  This actually isn't a bad idea for any AC range on that meter, since it is easy to do with just two button pushes.

And, b/t/w, contrary to some earlier discussion, if this behavior is characteristic of all the units, I think issue of the reading hanging without an OL indication might well be resolvable in firmware. 

[joeqsmith]

Now that it is clear you are adding a DC offset,  none of what you wrote surprises me.     At first I thought it was the wrong tool for the job but not posting your requirements, it seems you're just playing around with your new toy.   Most of the responses here seem correct.

So do you agree with the others, that there is no way of showing the overload under these circumstances?

I believe that given a set of requirements, a good group of EEs could design products to achieve them. 

Does this happen with every major brand with similar funcions?

As I already mentioned, its not how I would normally use a meter.   Another good example of what I would consider a fringe case, someone had posted how they were trying to measure a high value resistor that was right on the edge of the range switch point.  The meter was picking of the AC mains and had problems.  The tried a similar test using other meters and they did not have the problem.  Something like that.      In this case, I had the same meter and I think I was able to replicate it by twisting the probes with a lamp cord while trying to make the measurement.  I think in that case, the company came up with a hardware solution but it's nothing I considered adding as I am not normally wrapping my leads around lamp cords.   

[Fungus]

This meter is atypical, although documented in the specs, and would be unable to read power supply ripple in the mVAC range without an external blocking capacitor.

Yes, but it's a 50,000 count meter so you should be able to read power supply ripple perfectly well on the normal ACV range.

The only problem is that you have to read the fine manual.

You've done several similar tests on bigger scale, adding hundreds of volts of offset on the ACV scale. Many meters fail to handle it gracefully.

That was simply a ranging issue.

I wasn't referring to that specific case of the 87V giving incorrect readings.

The vast majority of meters will not be able to handle over-range normal mode AC rejection, it's just an inherent limitation that can't be easily solved with one component.

Yep. Joe gave the 87V a pass on the incorrect reading because it's quite atypical.

[bdunham7]

Except in this case it's not exactly a "limitation" of the BM789.
An operator error that leads to a DMM incorrect display is anecdotal information of next to zero value.

The meter is unable to read AC ripple on a DC bias in the mVAC range, such as a very common PSU ripple measurement.  How is that not a limitation?

As for the value of the information, I don't know why you think it has zero value.  It certainly is instructive about how to avoid a particular issue in using the meter.  Of course, as I posted, it's in the manual too, but who reads those?

[Fungus]

To be clear, since you are implying what 'others' agree on and I'm an other, I don't agree and think the meter should show an overload indication.  Whether that's a design issue or a one-off defect with your unit, I don't know.

That assume the meter can detect the condition, which is an unknown.

The way to find out would be to have Joe do a video on it. If Brymen fixes it and sends him another meter for his collection then we know it's a firmware issue.   

And no, this doesn't happen with most meter designs AFAIK, at least not the ones I have with an explicit mVAC function.  I think it is a result of 'bangforbuckitis' or the manufacturer trying to add as many features as possible at a price point.  Something has to give.

Yep. This is a massive bangperbuck meter. Maybe that capacitor plus a few other components was just a bit too much.

[Neutrion]

...
Which is fine! It's good to know the limitations of the tools we own, playing is a good way to find them.

Except in this case it's not exactly a "limitation" of the BM789. The OP stumbled upon an edge case which is very similar to a operator error, where the BM789's autorange algorithm doesn't compute an appropriate response. The obvious "fix" for this issue is not to reprogram the MCU inside the multimeter (because there is an almost unlimited number of similar edge cases), but to operate the Brymen BM789 following standard good practices, which even a beginner can learn is less than 15 minutes.

Also note that the edge case / wrong operation of the BM789 that the OP stumbled upon is completely different from the methodical stress testing that joeqsmith performs on his DMMs. Joe simulates real-life accidents and situations that the meters are supposed to be able to cope with, within and sometimes beyond their specifications.

Joe stresses the hardware and points to flaws in the design and safety of DMMs, the OP just found an operator error that the programmers at Brymen didn't account for in their autorange algorithm.

I very much enjoy Joe's videos where he goes through various steps to test his DMMs, because what I see is a talented engineer making sure his tools are fit for the job, and this information is very valuable. An operator error that leads to a DMM incorrect display is anecdotal information of next to zero value.

Sudden unforeseen  levels of voltages are also real life issues, which you don't want, but it just happens.

Now with this meter if you are measuring 300 mV AC, and the value goes over 500mV, you have to doublecheck it on an other scale.

But we are arguing here about whether it is the intended use of the meter: No it is not (except if one wanted to buy the meter for the mentioned ripple measurement.)
But the real question in the whole topic is whether the missing "OL" indication is solveable in FW.
According to Dave possibly not, 2N3055 provided a possible technical description why not, Joe seems to agree?.

I also really like the warm white backlight, which is uniqe.

OT, but... Brymen do some orange backlights, too, eg. my BM857. 

 I also like the orange more than the cool colors. Especially during the winter season it is more confy. But still the warm white is my favourite.
Just waiting to see a user configurable rgb backlight  Deep purple would be also nice before going to bed and taking a final look at the meter before sleeping.

[joeqsmith]

I certainly have and in cases where the meter doesn't want to switch to the next range,  it makes for a less than impressive review.   I was glad Brymen's engineers were able to sort it out before the meters release.   

The BM789 is not a bad little meter.  Add a few changes like split display and I think I would have a new favorite.    Considering the abuse that both BM869s have seen and their continued trouble free service,  its proving to be more difficult to find something I would take over them.

This missing split display was for me the only major point why I could not decide for a while between the 789 and the 869. Is there any major technical reason they decided not to go that way?
Or is it that in some industrial application maybe the big numbers are the number one priority?
I also really like the warm white backlight, which is uniqe.

Outside of any influence my small YT channel and posts on DMMs may have had on peoples purchasing decisions, I am not involved with marketing or sales.    I don't work for Brymen and really have no understanding of the hows and whys they or any other DMM company make the choices they do. 

Again,  I find meters that can display more than one parameter at a time allow me to work more efficient.   That's my personal choice.   As a consumer, it's good to have such choices to make!