What do you think of your EEVblog Brymen BM235 Multimeter ?

[Hobby73]

I would like to use Lithium batteries to avoid the leakage risk with Alkaline batteries.  The manual says:

Batteries:
1.5V AAA Size battery x 2
240 hours continuous on Alkaline
NOTE: Lithium AAA's may need to be drained a small amount before use to drop the initial terminal voltage.

As you know, Lithium cells tend to have excessive initial voltage (for some reason I never understood, maybe to extend shelf life?).  I just made these OCV measurements on brand new AAA cells:

Alkaline AAA 1.60V
Lithium AAA 1.84V

Should I discharge the Lithium to match the Alkaline at 1.6VDC?  I feel like I'm throwing away $$$ by draining it, but I want the DMM to receive power within spec.  If the DMM works with 10% tolerance, then I could drain it to 1.65V.

What shall I do?

[bitwelder]

What shall I do?

Try the Batteri... uh, no nevermind. 

[IanB]

I would like to use Lithium batteries to avoid the leakage risk with Alkaline batteries.  The manual says:

Batteries:
1.5V AAA Size battery x 2
240 hours continuous on Alkaline
NOTE: Lithium AAA's may need to be drained a small amount before use to drop the initial terminal voltage.

As you know, Lithium cells tend to have excessive initial voltage (for some reason I never understood, maybe to extend shelf life?).

This is not something within the ability of the cell manufacturer to control. The cell voltage is a fixed property of the cell chemistry, a function of the anode, cathode and electrolyte choices. They did pretty well to get the voltage as low as 1.8 V. As you know, many types of lithium cell are over 3 V.

I just made these OCV measurements on brand new AAA cells:

Alkaline AAA 1.60V
Lithium AAA 1.84V

Should I discharge the Lithium to match the Alkaline at 1.6VDC?  I feel like I'm throwing away $$$ by draining it, but I want the DMM to receive power within spec.  If the DMM works with 10% tolerance, then I could drain it to 1.65V.

What shall I do?

I tried lithium AAA's in the BM235 and found that, sadly, it does not operate properly. The circuitry detects an over voltage condition and goes into an error state.

Rather than trying to drain the lithiums down to a lower voltage I would suggest using rechargeable Eneloops instead. These seem to work fine.

[tautech]

It'd be interesting to know the exact voltage threshold the 235 throws up a battery overvoltage warning, then maybe a series diode could be inserted to drop the voltage some but still use and have the superior capacity of Lithium cells.

[EEVblog]

It'd be interesting to know the exact voltage threshold the 235 throws up a battery overvoltage warning, then maybe a series diode could be inserted to drop the voltage some but still use and have the superior capacity of Lithium cells.

I'll have to test this.

[EEVblog]

Just remembered I asked Brymen about this way back:

I was talking to engineer and was advised that BM235 battery voltage can not be more than 3.5V out of its current design. The reasons are:

1. Input warning detection uses a comparator inside main chip. While A or mA/uA input terminal is without being plugged-in by test lead, the compared signals are from battery voltage and from a fixed voltage around 3.6V. If battery voltage is more than 3.5V, that comparator may misjudge and display “InEr”.
 
2. The maximum working voltage of BM235 main chip can be only 3.6V.   

[apelly]

Just remembered I asked Brymen about this way back:

Brymen are bloody awesome aren't they? Some office dude actually went and asked someone what AND why. And they took the time to pass that on as well. I have only the best things to say about them.

[joeqsmith]

Just remembered I asked Brymen about this way back:

Brymen are bloody awesome aren't they? Some office dude actually went and asked someone what AND why. And they took the time to pass that on as well. I have only the best things to say about them.

Same

[irakandjii]

Funny mine is black and blue all over!... it must have taken a beating during the canoe ride across the pacific..

I use it all the time.  It is my go to meter. Light weight, rugged and just the right size.  All the little numbers say exactly what they are supposed to...

I am very pleased with the purchase.   

[EEVblog]

Just remembered I asked Brymen about this way back:

Brymen are bloody awesome aren't they? Some office dude actually went and asked someone what AND why. And they took the time to pass that on as well. I have only the best things to say about them.

Yep, they always come back with an answer to any technical question in record time. It's a small company I believe, so yeah, probably just leaning over the cubicle to the design engineer.
They don't seem to have any red tape at all which is great.

[Hobby73]

Just remembered I asked Brymen about this way back:

I was talking to engineer and was advised that BM235 battery voltage can not be more than 3.5V out of its current design. The reasons are:

1. Input warning detection uses a comparator inside main chip. While A or mA/uA input terminal is without being plugged-in by test lead, the compared signals are from battery voltage and from a fixed voltage around 3.6V. If battery voltage is more than 3.5V, that comparator may misjudge and display “InEr”.
 
2. The maximum working voltage of BM235 main chip can be only 3.6V.   

Very good.  In that case, my math skills tell me that if each AAA cell is lower than 1.75V it should operate fine.

[tautech]

It'd be interesting to know the exact voltage threshold the 235 throws up a battery overvoltage warning, then maybe a series diode could be inserted to drop the voltage some but still use and have the superior capacity of Lithium cells.

I'll have to test this.

If you can get away with a Schottky instead of a silicon diode then even a bit more Lithium cell capacity may be available.

[Crumble]

I recently received my unit from Dave, and I like it! Rugged and accurate bit of kit with documentation. I won't review it here because I hardly have the knowledge/equipment to do a review that adds anything to the remarks already made. Three things stood out to me though:

• When a small change occurs, the meter takes a long time to get a final reading. I found this while testing a power supply kit I had built for which I used a Spectrol 10-turn pot. When turning it quickly (jump of about 1V), the value takes 1 or 2 display updates to show. When tweaking the voltage by nudging the knob (change of 10s of mVs), it takes several seconds to get the proper reading. I initially thought it was a delay caused by capacitance in the power supply, but other meters show the value changing instantly. Something similar happened when joeqsmith upon checking high value resistances and low value capacitors. He shows it here (from the 8:48 mark). In resistance or capacitance mode it doesn't really matter to me, but in voltage mode it is actually quite annoying, especially when you want to set an accurate voltage. It seems there is some kind of digital noise filtering takes place inside of the unit. I have no idea if that can be switched off/enforced.
• The low impedance mode does not do anything under 8V. I had assumed it would have a resistor in parallel with the measument probes and switch it out of circuit according to the voltage causing a roughly similar drain at every voltage range. It is mentioned in the manual it does not work under 8V, but it is quite annoying because it beeps every time some small voltage is present (for example by touching the probes), implying something is happening while in fact no low-impedance source is present. It seems to be only usable to detect whether if no AC mains is present rather than filter out high impedance voltage sources, but discharging a small capacitor (to under 8V) is not an option. It seems to be usable for rough electrical work only.
• One minor thing which was not obvious to me before: it is brown-ish rather than grey. Very minor thing, but somehow I only noticed it when I received it. I had never seen it on the pictures or video clips. I may want to have my monitor calibrated, because it seems like my cheap-ass monitor is not completely accurate.

I only consider the slow updating of small changes actually annoying because none of my other meters seem to do it, but mentioned the other points for those interested. Maybe there is a filter I can turn off to make it update faster (at the cost of jitter maybe), but I could not find it in the manual.

Regards,
Crumble

[evava]

It is very interesting, no one mentioned it before (IIRC)...
What are your other meters you spoke about, if I may ask?

I hope that behaviour can be switched off, or maybe small firmware update would help?

Can someone please try that on higher Brymen meters, especially on BM867(869) - thanks!

I only consider the slow updating of small changes actually annoying because none of my other meters seem to do it, but mentioned the other points for those interested. Maybe there is a filter I can turn off to make it update faster (at the cost of jitter maybe), but I could not find it in the manual.

Regards,
Crumble

[Crumble]

No, that's the reason I brought it up. And yes, you may know my other meters . I compared it to a Gossen Metrawatt Metra Hit One, a Duro AL205 and an old Uni-T M840D I had for years. None seem to rely on big jumps to show the final value quickly. Also, when I look at the update rate on the sceen when you are monitoring a slowly changing signal it looks like the screen updates are irregular. It looks like there are about 5 updates per second as specified, but they all seem to occur in the first half of the second, reducing the usefulness of this update rate. Don't know if this behaviour is as intended. 

edit: joeqsmith tested a 150pF capacitor which took a long time to settle, but in this clip (2:12:50) a 220pF capacitor read instantly. It looks like the behaviour changes between 150pF and 220pF.

edit2: I dug up a 100pF and a 220pF cap, and found this behaviour to be reproducable with my unit.

[IanB]

I just did a few tests on my unit. It seems there may be some kind of adaptive filtering happening. If I probe a solid value the display updates instantly to the expected measurement. However, if I do a bit of scratching with the probes so that the meter may think the measurement is "noisy", then it looks like some extra filtering cuts in to filter out the noise. In that case it takes the meter a few seconds to settle down on the final value.

[MacMeter]

It is very interesting, no one mentioned it before (IIRC)...
What are your other meters you spoke about, if I may ask?

I hope that behaviour can be switched off, or maybe small firmware update would help?

Can someone please try that on higher Brymen meters, especially on BM867(869) - thanks!

I only consider the slow updating of small changes actually annoying because none of my other meters seem to do it, but mentioned the other points for those interested. Maybe there is a filter I can turn off to make it update faster (at the cost of jitter maybe), but I could not find it in the manual.

Regards,
Crumble

The BM235 is NOT firmware updatable. One of the few negatives. Otherwise, I've been very pleased with its performance.

[joeqsmith]

I don't want to waste too much time on this without having more data about exactly what you were doing.   Here is mine along with a few other meters (including the BM869s) showing various level changes and settling times.  You may have stumbled onto a problem but we would need more details.   

Removed video, see post below

I recently received my unit from Dave, and I like it! Rugged and accurate bit of kit with documentation. I won't review it here because I hardly have the knowledge/equipment to do a review that adds anything to the remarks already made. Three things stood out to me though:

• When a small change occurs, the meter takes a long time to get a final reading. I found this while testing a power supply kit I had built for which I used a Spectrol 10-turn pot. When turning it quickly (jump of about 1V), the value takes 1 or 2 display updates to show. When tweaking the voltage by nudging the knob (change of 10s of mVs), it takes several seconds to get the proper reading. I initially thought it was a delay caused by capacitance in the power supply, but other meters show the value changing instantly. Something similar happened when joeqsmith upon checking high value resistances and low value capacitors. He shows it here (from the 8:48 mark). In resistance or capacitance mode it doesn't really matter to me, but in voltage mode it is actually quite annoying, especially when you want to set an accurate voltage. It seems there is some kind of digital noise filtering takes place inside of the unit. I have no idea if that can be switched off/enforced.

...

Regards,
Crumble

[/list]

[IanB]

I don't want to waste too much time on this without having more data about exactly what you were doing.   Here is mine along with a few other meters (including the BM869s) showing various level changes and settling times.  You may have stumbled onto a problem but we would need more details.

In your test you were switching cleanly between two solid and stable voltages. However, try "confusing" the meter by making the voltage wobble in a more or less random way in between measurements. In my test this seemed to activate some kind of low pass filter on the BM235 causing a much longer settling time. The same thing did not happen on the BM869s.

[joeqsmith]

Now you know very well there is no such thing as a clean switch between levels.   
Does it matter what levels you ran at and how much you deviated by?  Any idea on the ramp rate?  I should be able to change the rise and fall times to replicate it. 

[IanB]

OK, the way I tested it was by touching meter probes to battery terminals by hand. To make a "noisy" signal I simply scratched and tapped the probe on the terminal to make the meter reading jump all over the place. After doing this for a second or two I then touched the probe firmly to get a clean measurement. This is where I saw a longer settling time.

If I simply touched the probe to a terminal in one clean motion the display settled almost instantly.

[joeqsmith]

Interesting.  I will try something a little more controlled and see if we can determine what conditions put it into this long settling time. 

If when you notice it is in the settling mode and you move it quickly off the battery post and back on, does it then settle quickly?   

[joeqsmith]

If I slow down the edge rates enough I can not only get the BM235 to exhibit the long settling times, so does the BM869s.   Interesting enough, I have not been able to cause it with my UT181A or the TIP194II.   Something unique to Brymen's code base?  Maybe Dave has enough pull to get an answer to this one.   

As I shorten the times, the meters are less prone to enter the long settling time mode.

I pulled the first video and added a section to the end of it and reposted. 

https://www.youtube.com/watch?v=_O4auxR1hsA&feature=youtu.be

[Crumble]

Wow, thank you Mr Smith for your prompt and detailed reply! I must excuse myself for not having providing a more detailed report, but sadly I lack the proper equipment here, so I turned out just tweaking an analog pot, making it difficult to quantify the issue.  Turns out that for tweaking circuits the BM235 (or BM869s) might not be ideal, because they settle slowly after tweaking with a slow ramp-up. Not a major issue, but useful to know about it when using it in analog circuits because it consequently seems to lag behind in the value on the screen when faced with slowly ramping signals.

I guess their AD converter switches in some kind of averaging when it detects little change in value, maybe to increase accuracy when doing so. I just happened to stumble across it building this power supply kit. These 10-turn potentiometers allow setting it within a few mV with a 19,5V range, but this issue will likely not be noticable with a single turn unit or a more "scratchy" unit. I guess using a digitally controlled power supply will not lead to similar problems because of the steps. I guess the capacitor measurement also takes place with a circuit that ramps slowly when applying a small capacitor.

Nonetheless I still love my little Brymen, and might purchase more in the future. This is just something one needs to know, maybe this can be included in the manual in the future? I am curious what kind of system they use that causes this though.

[joeqsmith]

No problem. Glad to provide what little information I can.  It appears another person had posted about it some time ago.  Fan boys down voting? Strange. Seems easy enough to reproduce. I'm surprised the battery noise test gets it into this mode as well. It's a strange problem.   

I can see your point about using the meter to made fine adjusting and having it get into this mode and then having to wait for it to settle.  So far it has not bit me.