BM235 vs BM867s

[robdejonge]

For some more on-point arguments, it was recommended to me that I post this question here.

A decade ago I bought a $5 multimeter to keep around for occasional use like checking a battery voltage or how fast a phone was charging.

A few years ago I bought a 40,000 count multimeter for ~$60. I had started getting into electronics a bit. Low voltage stuff, mostly surrounding home automation and other simple circuits. The occasional more challenging gig. The meter still measures within spec, but I’ve noticed odd behavior. Mechanical issues perhaps.

Now, I'm in the market for a 'proper' multimeter. My initial selection thoughts:

• No need for it to be yellow for reasons Dave made a video about
• Reliable manufacturer
• Long-term reliable seller
• No more than €150
• Comfortable to stick the leads into mains
• 6000-60000 count
• Voltage, current, frequency, resistance, capacitance, diode, continuity
• Temperature is *not* important to me

I decided on Brymen. Their meters get great reviews, they seem to build quality products and I'm able to buy them through some reliable sellers in Europe when I visit there this summer.

I'm now trying to decide between the BM235 and the BM867s; for the extra €60 I'd get:

• An extra digit that I know I don't really need, but might be a 'nice to have' feature to, for example, see gradual changes in battery voltage or something like that
• Dual display that I've never had and so don't really know how useful it is. Comments have been made that I could just as easily add a second meter for the second parameter.
• 'Analog' type bar graph is a feature I could see being useful. I'm just not sure how to quantify how valuable exactly.
• Data logging is a feature I would really like. I've read however that Brymen data is really weird, so I'm wondering if I'll be able to get it working on a Linux machine without having to write software myself.
• I *think* it can detect one-of pulses better? That could be useful.
• Duty cycle seems useful if it can be measured without having to set up my scope.
• I really like all the whitespace on the meter display, for some reason
• Neither nS or dBm, or LoZ are things I've ever needed.

The meter will land on a hobby bench that has an oscilloscope, bench power supplies, other cheaper multimeters, logic analyser, soldering gear, etc. that is all used by a curious but decidedly ignorant individual. I am by no means qualified to be using any of this stuff, just like learning. Most of what I do is home automation stuff, plus the odd repair around the house.

I'm curious what arguments you all would offer up in deciding between these meters.

[Fungus]

I'm curious what arguments you all would offer up in deciding between these meters.

Me? I'd get an "industrial" BM857s. They're square and chunky. I like that. They're also quite a lot smaller than those massive curvy ones (literally half as thick... probably 1/3 of the volume of the 867s).

Also: The dial is simpler - less function overload on the ranges. The only thing I really need to switch is resistance/continuity. Beware that the pic on the Brymen web site is wrong, it shows the 859s which has temperature where the 857s puts the diode function.

It has the 500000 count mode and bar graph, etc. It's not dual display but I can switch between values with a button press (eg. press the Hz button to see frequency of an AC signal).

• Dual display that I've never had and so don't really know how useful it is. Comments have been made that I could just as easily add a second meter for the second parameter.
• 'Analog' type bar graph is a feature I could see being useful. I'm just not sure how to quantify how valuable exactly.
• Data logging is a feature I would really like. I've read however that Brymen data is really weird, so I'm wondering if I'll be able to get it working on a Linux machine without having to write software myself.
• I *think* it can detect one-of pulses better? That could be useful.
• Duty cycle seems useful if it can be measured without having to set up my scope.
• I really like all the whitespace on the meter display, for some reason
• Neither nS or dBm, or LoZ are things I've ever needed.

I'm curious what arguments you all would offer up in deciding between these meters.

Bar graph is useful.

Dual display? I'm never sure. It's OK when you're actually using it but otherwise it's just clutter. Me? I wouldn't use it much so that's why I go for the 857.

Brymen protocol is discussed here: https://www.eevblog.com/forum/testgear/brymen-ir-connection-protocol-anyone-sniffed-it-yet/msg208747/#msg208747

The 235 is a good meter but it has a lot less counts and no bar graph. To me it seems more like an electrician/workshop meter than en engineering meter. It also does a lot of things I'd never use like LoZ and VFD. All those functions clutter up the dial.

Bottom line: I guess I'm a low-clutter person. I don't see the point in having lots of extras that I'll never use but are constantly distracting me.

[sonpul]

My opinion is this. If the wonderful BM859s do not suit you, do not torture yourself with doubts. Buy BM869s. Let him know everything. Don't trade for the BM867.
I saw that you are unsure about many of the functions listed. Let me remind you once again about the BM789. It is very compact, thin, comfortable in the hand and not bulky on the table. Knows a lot.

[2N3055]

Temperature measurement is useful. Also with two probes for a measuring of temperature difference between 2 points.
And so is Lo-Z. For those occasional moments where you measure mains it is very useful..

Dual screen? It depends.. Measuring 2 things at the same time can be useful. Not necessary but useful.

Also I would take a look at BM257...

[mwb1100]

• Data logging is a feature I would really like. I've read however that Brymen data is really weird, so I'm wondering if I'll be able to get it working on a Linux machine without having to write software myself.

EEVblog user HJK has written a logging program that supports a wide range of meters, including the BM867s.  The software supports Linux, Windows, and Mac (it's written in Java).

  - https://www.eevblog.com/forum/testgear/program-that-can-log-from-many-multimeters/
  - https://lygte-info.dk/project/TestControllerIntro%20UK.html

If you look at the eevblog thread you will see that HJK actively supports the program.

Disclaimer: I have never actually used the program, but have no reason to believe it doesn't do what HJK says it does.

[robdejonge]

Brymen protocol is discussed here: https://www.eevblog.com/forum/testgear/brymen-ir-connection-protocol-anyone-sniffed-it-yet/msg208747/#msg208747

What a delightfully odd protocol indeed. Haha!

EEVblog user HJK has written a logging program that supports a wide range of meters, including the BM867s.  The software supports Linux, Windows, and Mac (it's written in Java).

Fantastic, that will prove to be super useful! Thanks for sharing. Will for sure check it out. A quick search also found this Python script, which looks useful.

[robdejonge]

Me? I'd get an "industrial" BM857s. ...snip... It has the 500000 count mode and bar graph, etc.

What do you say to the "6000 count is more than enough for a hobby user" comments I read here all the time?

[robdejonge]

Also I would take a look at BM257...

Interesting suggestion. That looks like a BM235 with data logging. Thanks!

[Fungus]

Me? I'd get an "industrial" BM857s. ...snip... It has the 500000 count mode and bar graph, etc.

What do you say to the "6000 count is more than enough for a hobby user" comments I read here all the time?

For real work? I agree with it.

6000 counts is ~0.02% accuracy. That's plenty. Most components are 5% tolerance and I never saw a 5V power supply that was exactly 5.00V under all loads.

But ... hobbies are supposed to be entertaining. Volt-nuttery is fun and a 50,000 count meter doesn't hurt.

It's also good to have a reference meter to compare all your other meters against.

[robdejonge]

Me? I'd get an "industrial" BM857s. ...snip... It has the 500000 count mode and bar graph, etc.

What do you say to the "6000 count is more than enough for a hobby user" comments I read here all the time?

For real work? I agree with it.

6000 counts is ~0.02% accuracy. That's plenty. Most components are 5% tolerance and I never saw a 5V power supply that was exactly 5.00V under all loads.

But ... hobbies are supposed to be entertaining. Volt-nuttery is fun and a 50,000 count meter doesn't hurt.

It's also good to have a reference meter to compare all your other meters against.

Excellent arguments to be added to my "this vs. that" list! Thanks ;-)

[mwb1100]

Regarding the BM257s there are two nitpick things which have bugged me about the meter:

1. it can only measure plain silicon diodes - it cannot light up even a red LED.  I knew this when I bought it and figured I didn't need that capability.  But then came a day when I actually tried to test an LED and I had forgotten about the limitation...

2. the backlight remains lit for only 30 seconds.  This is something many users don't care about, but I find myself using the backlight a lot.  And the short duration drives me a bit crazy.

[2N3055]

Me? I'd get an "industrial" BM857s. ...snip... It has the 500000 count mode and bar graph, etc.

What do you say to the "6000 count is more than enough for a hobby user" comments I read here all the time?

For real work? I agree with it.

6000 counts is ~0.02% accuracy. That's plenty. Most components are 5% tolerance and I never saw a 5V power supply that was exactly 5.00V under all loads.

But ... hobbies are supposed to be entertaining. Volt-nuttery is fun and a 50,000 count meter doesn't hurt.

It's also good to have a reference meter to compare all your other meters against.

6000 count is 0,016% resolution... Accuracy is specified separately.

And 6000 count is good enough for very professional work..  If that work does not require more...

Professional - hobby distinction comes in play with imperative to perform at different contractual level.
You can do same thing at home for fun and be hobby user, or if you do exactly same type of work for other people and get paid for it you are pro.

There are many professionals that can perform their contractual obligations with simple meters with 3.5 digits and 0.3% accuracy with no TrueRMS. And they do it well and reliably. Professionally. Because that meter is good enough for the job they do..

[robdejonge]

Regarding the BM257s there are two nitpick things which have bugged me about the meter:

1. it can only measure plain silicon diodes - it cannot light up even a red LED.  I knew this when I bought it and figured I didn't need that capability.  But then came a day when I actually tried to test an LED and I had forgotten about the limitation...

2. the backlight remains lit for only 30 seconds.  This is something many users don't care about, but I find myself using the backlight a lot.  And the short duration drives me a bit crazy.

You can’t just turn the backlight on permanently? Hm.

I believe these also apply to the BM867 / BM869. Can someone confirm?

[Fungus]

6000 count is 0,016% resolution... Accuracy is specified separately.

We weren't talking about any specific model so I assumed an ideal multimeter. 

[2N3055]

6000 count is 0,016% resolution... Accuracy is specified separately.

We weren't talking about any specific model so I assumed an ideal multimeter. 

Your numeric estimate was correct enough.. I just wanted to point out difference between accuracy and resolution. As a clarification, nothing more..

[sonpul]

Each brand has an engineer on the team who brings some kind of dirty trick to the ideal development.
For example, they make the backlight only 15 seconds, or wedged nS between measurement and continuity modes.
Probably in Briman this developer was fired. In recent versions, 869 has about 3 minutes. And the BM789 finally did the right thing, 30 minutes of backlighting. But they still left the catch, did not remember the previous state of the backlight when turned on. Тhat would be ideal.

[mwb1100]

You can’t just turn the backlight on permanently? Hm.

Not that I'm aware of.  If that can be done, I hope someone can fill me in.

I believe these also apply to the BM867 / BM869. Can someone confirm?

According to the manual the BM867s/BM869s models have a 256 second backlight.  4+ minutes seems reasonable to me.

[robdejonge]

I believe these also apply to the BM867 / BM869. Can someone confirm?

According to the manual the BM867s/BM869s models have a 256 second backlight.  4+ minutes seems reasonable to me.

I'd agree. I thought I had seen comments about shot backlight times somewhere, but looks like that was in older firmware versions.

[HKJ]

• Data logging is a feature I would really like. I've read however that Brymen data is really weird, so I'm wondering if I'll be able to get it working on a Linux machine without having to write software myself.

EEVblog user HJK has written a logging program that supports a wide range of meters, including the BM867s.  The software supports Linux, Windows, and Mac (it's written in Java).

  - https://www.eevblog.com/forum/testgear/program-that-can-log-from-many-multimeters/
  - https://lygte-info.dk/project/TestControllerIntro%20UK.html

If you look at the eevblog thread you will see that HJK actively supports the program.

Disclaimer: I have never actually used the program, but have no reason to believe it doesn't do what HJK says it does.

TestController only support HID interfaces on Windows and BM867 uses HID.
But BM257 has serial interface and works on any OS (Windows, Mac, Linux) supported by TestController.

[Fungus]

Yeah, the backlight thing is completely stupid, I don't know what goes through their heads. Let us choose!

Why can't we hold the "light" button for three seconds to make it stay on forever? (or something like that)

[J-R]

Running the backlight on the Brymen DMMs pulls a lot of extra power (2x-4x depending on the model), so the already short battery life would be greatly reduced.  It was perhaps a wise marketing decision to have a short backlight timeout to avoid having the Brymens be labelled as excessively battery hungry.

[Fungus]

... avoid having the Brymens be labelled as excessively battery hungry.

Instead they get labelled as having useless backlights?

I don't see how that's better.

[J-R]

Brymen's backlight and auto-power-off behavior is similar to many DMMs where it is a common design choice in order to help the user avoid constantly picking up a tool expecting to be able to use it but instead finding a dead battery...

[Fungus]

Brymen's backlight and auto-power-off behavior is similar to many DMMs where it is a common design choice in order to help the user avoid constantly picking up a tool expecting to be able to use it but instead finding a dead battery...

By that logic they should make flashlights with auto-power-off... 

[mwb1100]

Just for the record, the BM257s is a fine multimeter. 

But as often happens, I found imperfections that I cannot "unsee" - even if they are actually of little consequence.  Kind of like when a stupid song gets into your head and just plays itself over and over. It took me more than a year before I tried to light up an LED - and I probably wouldn't do it again for another year.  On the plus side, it actually won't happen again, because every time I pickup the meter now the earworm of "oh yeah - this doesn't light LEDs" pops into my head.  Even though lighting up LEDs was (until that moment) the furthest thing from my mind.  (Cripes - I just wrote an entire paragraph about lighting up LEDs... )

And in principle I have no problem with auto shutoff for backlights.  But the 15s timer that many meters implement is just plain stupid.  And 30s is irritating (bordering on stupid).  Give me at least a minute, or implement Fungus' excellent idea of an extra long button press to turn on the backlight with no timeout (or a 10 minute timeout  if the meter's designer is really concerned that I'll use up too many of my batteries).