NEW Brymen Bluetooth BM787BT Multimeter

[joeqsmith]

Showing the modified BM786 to BM787BT with t8712 firmware, 1.20 radio firmware and 1k resistor.   Total time was a bit over 29 hours.  The batteries in the meter then too low and the connection dropped.  I suspect this is also why I was seeing an additional count of change over the last 12 hours or so, getting worse until the test stopped. 

This data was captured using the PC without the CRC correction.  However, no CRCs were detected during this test.   

The typical transfer time is 0.2 second or 5Hz (for the resistance function).  Notice that towards the end there were samples that required much longer but the software was able to recover the connection.   None of these caused any problems with the data.   

If anything, any problems continue to seem to point to the phone, not the meter.

[EEVblog]

Showing the modified BM786 to BM787BT with t8712 firmware, 1.20 radio firmware and 1k resistor.   Total time was a bit over 29 hours.  The batteries in the meter then too low and the connection dropped.  I suspect this is also why I was seeing an additional count of change over the last 12 hours or so, getting worse until the test stopped. 

I just tested for 3 hours or so and I heard the meter beep, turned out it had switched off due to the low battery alert, but the app lost all my data. I could save the data to CSV but it was empty. I'll report that as a bug.

[EEVblog]

Saw your latest video with the Keysight Bluetooth battery leakage.  Was your plan to use it to compare with the Brymen meter?

No.
Don't know if they are still shipping. I'd like to see the Windows and Apple apps first.

[joeqsmith]

Showing the modified BM786 to BM787BT with t8712 firmware, 1.20 radio firmware and 1k resistor.   Total time was a bit over 29 hours.  The batteries in the meter then too low and the connection dropped.  I suspect this is also why I was seeing an additional count of change over the last 12 hours or so, getting worse until the test stopped. 

I just tested for 3 hours or so and I heard the meter beep, turned out it had switched off due to the low battery alert, but the app lost all my data. I could save the data to CSV but it was empty. I'll report that as a bug.

I had reported it as well.  They responded today and it sounds like a fix will be soon.   

The problem with the BLU phone sounds like it will be more complex to solve, but they are working on it.   

I'd like to see the Windows and Apple apps first.

If you have an iPhone, I would think they would allow you to test that application as well.  I can't steal my wife's phone long enough to run any long term tests with it. 

I have not heard about a Windows, LINUX or MAC application. 

[EEVblog]

I had reported it as well.  They responded today and it sounds like a fix will be soon.   

Yes, it was already on the list of fixes for the next version.

If you have an iPhone, I would think they would allow you to test that application as well.  I can't steal my wife's phone long enough to run any long term tests with it. 
I have not heard about a Windows, LINUX or MAC application.

I don't have a modern Apple phone. They mentioned Windows to me, but I checked email and it's "iOS APP and/or Window base software"

[joeqsmith]

I see they updated their applications for both the iPhone and Android to correct the problems with data loss during a disconnect.   

They continue to investigate the compatibility problems with my BLU G64 and I understand have expanded to testing with other Android phones as well.  Being completely clueless about these phones, I was surprised to hear just how different the hardware is.

I have stopped all testing for now. 

[EEVblog]

I see they updated their applications for both the iPhone and Android to correct the problems with data loss during a disconnect.   

I just tried it with a flat battery meter and got 13,000 reading before it dropped out and the meter switched off. The app handled it just fine and allowed me to save the data no problem.
Also, no x10 problem with 0.123k

[joeqsmith]

Because of this new meter, I went from never owning a cell phone to now having three of them.  None with service.   

I did try the new app and also found it would allow me to save the data but my tests were minimal. 

Now that you know you will not loose your data, I guess you could try a long term test with new batteries to get a feel for life.  Maybe just use the same brand/model that were supplied.   What I saw early on is the link can start to drop before the low battery indicator comes on.   It acts like there is a surge of current.   I have not looked into yet.  May make for a decent short video on your second channel.

[EEVblog]

Now that you know you will not loose your data, I guess you could try a long term test with new batteries to get a feel for life.  Maybe just use the same brand/model that were supplied.   What I saw early on is the link can start to drop before the low battery indicator comes on.   It acts like there is a surge of current.   I have not looked into yet.  May make for a decent short video on your second channel.

I'll need to dig out an old photo for that if I'm actually logging. Not that that would make any actual difference to a battery life test, just leave the meter with data mode turned on.

[joeqsmith]

When the BM788BT is advertising, it broadcasts 20 bytes (+ overhead) every 2.5 seconds.  Once notifications are setup, the meter will send 152 bytes at up five times per second depending on the function selected.  I have yet to look at battery life and don't know how much it costs to send the notifications compared to the broadcast.  Also, it looks like when I measured the latest version of the BM789, the peak current for all the modes was 6.71mA (no backlight or beeper). 

In other words, consider running the test with the meter connected and logging.  Then maybe just document which function was selected.  If you choose to log the data, you may want to use the same 1k resistor you used when the 10X glitches were showing.  With it being a long term test and new software, it may help build confidence that the problem is solved.  The other thing that may be of interest is how the lower battery effects the readings.   I noted there was an error towards the end of that 24 hour test.  It may be helpful to quantify that error.

I also suggest running the test with logging using the latest application just to build confidence that the system is robust enough to actually collect data for an extended period of time.

[joeqsmith]

If you would like to compare some measurements, I set the battery voltage to 4.5V  Then monitored the current with the meter in the AC VFD mode while I changed the BLE from off to advertise to notify.  It appears to draw about 400uA more to notify vs advertise.  We knew it wouldn't be free.  I don't know what the cause of the current spike is when it first starts sending the data. 

With notifications, I then adjusted the supply down.  The low batter begins to flash as 3.52V.  At 2.4V, the LCD is washing out.  Maybe half contrast. At 2.3, I lost connection.

All of that seems fairly normal.  Now the strange part.  I adjust the voltage to 2.5, the meter will not turn on.  Eventually at 2.9 it does, but it turns back off.  I think that the meter will turn off the power, even if the auto power off is disabled depending how lot the battery is.     

If I let it continue to log data, I see these current spikes at some random interval.  What ever is going on, it seems to stall the communications.   Maybe an ADC autocal or something.

[joeqsmith]

Running the normal battery check sequence and comparing with/without the BLE active (notification when active).  Comparing with the released BM789, you can see that maybe Brymen turned down the backlight.  Other than that, when BLE is turned off, the two meters basically track.   

The battery life between the two is basically a wash when BLE is disabled.  When logging, it looks like you loose about 30 hours of run time.  Not too bad really.   

[EEVblog]

I think that the meter will turn off the power, even if the auto power off is disabled depending how lot the battery is.     

Yes, I've noticed that.

[EEVblog]

Brymen want to know if I want the meters in July, or wait until the Windows version is also available, but they can't give me a date on that. 
The Android/iOS app uses the exact same code base, so both of those are available at the same time with the same fixes/bugs.

[joeqsmith]

I understand Brymen is going to supply you with a list of cell phones that have been qualified to work with their application along with those with known compatibility issues.   

I have never talked with them about their Windows application.

[joeqsmith]

Just an FYI, with the new cell phone and latest Android application, using my modified BM786 (BM787BT) with all the latest firmware, I was FINALLY able to replicate your 10X failure using the resistance mode with 1k resistor.   I strongly suspect now that the problem lies within the phone application as I have yet to see it when using my own custom application with the PC.   I have passed that data onto Brymen along with some other test data I collected.   At least we now know it is not unique to your setup and the addition of the CRC did not correct it. 

2025-05-26,21:28:40.239,BM78xBT,#channel1,Resistance,9.9996,k,Ω

[EEVblog]

Just an FYI, with the new cell phone and latest Android application, using my modified BM786 (BM787BT) with all the latest firmware, I was FINALLY able to replicate your 10X failure using the resistance mode with 1k resistor.   I strongly suspect now that the problem lies within the phone application as I have yet to see it when using my own custom application with the PC.   I have passed that data onto Brymen along with some other test data I collected.   At least we now know it is not unique to your setup and the addition of the CRC did not correct it. 
2025-05-26,21:28:40.239,BM78xBT,#channel1,Resistance,9.9996,k,Ω

Wow, nice work, thanks, I wasn't going crazy.
I'm going to get an Apple shoe phone and they are putting me on the Apple beta app program.

[joeqsmith]

Brymen thinks the 10X bug may have been addressed with these latest changes.  They have also release a new application for the Android today (store only).  They have started to run long term tests and have asked if I could repeat it as well using the latest updates.   Good to see them getting these problems resolved prior to the release.  Seeing all three of those Android phones working smooth with both meters was a big milestone.   

[joeqsmith]

Before running this next test to see if the 10X bug will replicate with all the latest changes,  I hacked up a AA battery pack to give me an even longer run time.   Who has concerns for safety when measuring a 1k resistor??   

IMO, the highest priority problem I would go after is how the sampling rate slows the longer you collect data.  This is not a problem when using my custom software + PC.  Brymen is investigating this problem.     

[BILLPOD]

Good Morning Joeqsmith, I too never had a cellphone until I started accumulating meters that have Bluetooth and can display/record data.   My Mooshimeter was my first candidate.   So, I went to Amazon, looking for an inexpensive cellphone.  I got the BLU model G61S.  I have several Bluetooth capable meters and they all connect to the phone.   I haven't checked if there are any 'drop-outs' when recording data, but my problem is range.   I can't get a signal once I get approx. 6 ft. away.   What kind of range are you getting

[joeqsmith]

Good Morning Joeqsmith, I too never had a cellphone until I started accumulating meters that have Bluetooth and can display/record data.   My Mooshimeter was my first candidate.   So, I went to Amazon, looking for an inexpensive cellphone.  I got the BLU model G61S.  I have several Bluetooth capable meters and they all connect to the phone.   I haven't checked if there are any 'drop-outs' when recording data, but my problem is range.   I can't get a signal once I get approx. 6 ft. away.   What kind of range are you getting

Hello Bill.  I have not yet tried any range tests with the Motorola or the Unnecto.  The BLU G64 can communicate from my upstairs to my downstairs living room which is maybe 30 or 40 feet.   I think Dave was testing at 9 meters or 30 feet.    I took the BLU  and meter to work to test it in a high traffic environment.  I was at least 40 feet away but that was line of sight. 

I did a test several years ago using a cheap tablet to communicate with Dave's 121GW, Gossen Ultra, UNI-T UT181A.  All of those worked far beyond 6 feet.  Time stamped with Dave's meter, maybe 20' or so.

https://youtu.be/e_YzwO62feQ?t=622   

***
While on the subject, when I finally replicated the 10X fault Dave found,  I had the phone and meter right next to one another, at my home which is a very low traffic area.  What ever the cause, it doesn't seem related to a weak signal or heavy traffic.   I have been running with their latest firmware and application with the same phone and modified BM786 (to simulate Dave's meter) for about 2.5 days now.   The plan is to let it run for at least five days to see if the updates had any effect.   

[EEVblog]

While on the subject, when I finally replicated the 10X fault Dave found,  I had the phone and meter right next to one another, at my home which is a very low traffic area.  What ever the cause, it doesn't seem related to a weak signal or heavy traffic.

That's interesting.

[BILLPOD]

Thank you Joe,  the short (<6 ft), range is apparently the fault of my BLU61S phone, as none of my bluetooth equipped meters work beyond 6 ft.   I would be very happy if I could get even 20 ft.  I got the cheapest new phone I could find and I guess I got what I paid for.
However, all is not lost, as I get some use out of the phone as a 'tablet'.   Thank you for your response.

[joeqsmith]

I would like to know if your iPhone produces the 10x problem and if you experience a reduced sample rate for longer data collections. 

It's a bit odd that the 10X fault produces the exact same value every time with both your an my setups.  Their application appears to save everything into the phone's RAM while it is collecting.  It doesn't offload it to NVRAM until you complete the collection.  I don't know why they don't create the file at the start of the collection and stream the data to NVRAM, the keep a limited amount in RAM. 

Anyway, I wonder as the phone starts to clog up with all that data and slow down, is this somehow causing the false 10x reading.   When I replicated the fault, the phone had ran for a fair amount of time and was slowing down.  In my case, the phones have no SIM card, and very little software is loaded onto them.   If the phone was actually being used as a typical phone, maybe the resources required can cause the fault more often and earlier in the collection.   

All pure speculation on my part. 

[joeqsmith]

Thank you Joe,  the short (<6 ft), range is apparently the fault of my BLU61S phone, as none of my bluetooth equipped meters work beyond 6 ft.   I would be very happy if I could get even 20 ft.  I got the cheapest new phone I could find and I guess I got what I paid for.
However, all is not lost, as I get some use out of the phone as a 'tablet'.   Thank you for your response.

$50 Motorola MotoG Tracphon