Brymen BM86x Backlight Duration Mod

[SMB784]

Hey everyone!

Like many of you, I love my Brymen BM867 multimeter.  Its a great meter for the money with loads of functionality, and even a nice, bright backlight.  However, for my purposes the backlight doesn't stay on long enough - a mere 15 seconds! - and there doesn't appear to be a way to adjust the duration from within the existing software.  As such, I would like to look in to how to adjust the firmware running on this device so that the back light duration can be modified to be of indefinite duration.  The current backlight can be toggled on and off by holding the select button, and times out after 15 seconds if it is not switched off manually.  Ideally the mod would increase this timeout to a very long time (a minute or more), possibly up to the auto-shutoff time for the meter itself.

As a caveat, I recognize that the 15 second duration is designed to limit battery usage and extend battery life, however I am not really concerned with that, and I am more than happy to replace batteries frequently.

The two chips that seem to run the thing are labelled BTC-85-4 and BTC-0197-0000

The first one is the ADC chip, which might also contain a microcontroller but more than likely seems to just sort out signal processing.  It is rumored to be a clone or close cousin of the FS970X

The second one is likely the display driver, and it may also serve as the master microcontroller that governs the operation of the device.  It has an EEPROM (a 4kB S24CS0A4) associated with it which may contain the firmware (located at U5)

Any advice on how to go about this?  Maybe reading the EEPROM is a good start.

[Corner]

I too I'm bothered by the short backlight duration. The backlight on the F87V is great to use in contrast. I'd like to mod it but I don't have the knowledge to do so.

[SMB784]

So it has been claimed that the EEPROM stores the calibration data for the meter; it is unclear whether or not it also stores the firmware. There is a tx/rx pair of pins going to the display, perhaps the main chip communicates with the display via SPI or UART, which can be tapped to see if there is useful information being transmitted.

[CDaniel]

Coud be much easier to add a small microcontroller to control directly the backlight from the backlight button, of course if the leds wires are accessible

[SMB784]

After doing some sleuthing, it would appear the large chip (BTC-0197) might just be an LCD display driver with no associated microcontroller capabilities.  That leaves the AD-85 chip to be the brains behind the operation; if it is truly an FS970X it should have an onboard microcontroller that could execute firmware.  One possibility I will explore is to trace the LED backlight signal to a chip and see where it ends up - at the AD-85 or the BTC-0197.

[chemik22]

After doing some sleuthing, it would appear the large chip (BTC-0197) might just be an LCD display driver with no associated microcontroller capabilities.  That leaves the AD-85 chip to be the brains behind the operation; if it is truly an FS970X it should have an onboard microcontroller that could execute firmware.  One possibility I will explore is to trace the LED backlight signal to a chip and see where it ends up - at the AD-85 or the BTC-0197.

Have you found anything useful to solve the backlight timer problem ?

[coromonadalix]

for some fluke you had a capacitor change who helped getting the power down delay longer,  for backlight ??   you have to find the datasheet  of the ic's ... if they exist

the BTC  chip seems custom,  send an email to Brymen and or Grenlee    since there is no schematics floating around ?

[IanB]

To the best of my understanding, the backlight timer is programmed in the firmware. It can only be modified with new firmware, and that is not an aftermarket option. The only option would be to buy a new meter from current production.

[chemik22]

Ech.. I have just had a hope that it can be finally solved by simple hard soldering and turn the light constantly without capability to off it. But probably if nobody did that think before I guess it is not possible.

[shakalnokturn]

First do a little reverse engineering see if it's an external RC monostable triggered by one of the "BTC's"? (Doubtful...)
Next some bus sniffing, if you're lucky and it is EEPROM set, it may fetch the constant each time backlight is enabled, in that case once you have the address it's a matter of playing with the data value. Easy enough.
If IanB is right chances of existing software modification are low. In this case I'd go for kludging-in a monostable of one sort or another. The PIC 10F MCU's in SOT-23-6 would be my choice for this kind of job. If you can tap in to the keypad easily it even allows you to set different modes depending on key usage. (ie: Long press->enable BL for 15 min, Longer press-> enable until DMM power off. It would also leave the option to disable BL manually without power cycle if needed.)

As for having the BL constantly on... If you really need it that much... I'd at least try to tie it to the auto power off timeout so you don't drain the battery if you leave for coffee and never return to work.

[joeqsmith]

To the best of my understanding, the backlight timer is programmed in the firmware. It can only be modified with new firmware, and that is not an aftermarket option. The only option would be to buy a new meter from current production.

https://youtu.be/bDIx4ik0vvA?t=698

[Fungus]

Add a flip-flip to the LED circuit so that pushing the backlight button toggles it on/off.

[mwb1100]

Has anyone dumped the EEPROM from any of the BM86x family (including the BM525/BM521)?  I suspect there's a half decent chance that there's a byte in the EEPROM for the number of seconds the backlight stays on.

[IanB]

I don't think anyone knows for sure that there is an EEPROM in the meter. Based on interactions between user joeqsmith and Brymen, it might be that the firmware is stored in non-erasable ROM and the only way to change it is by chip replacement.

[shakalnokturn]

There's a difference between firmware in mask ROM or flash in the MCU and configuration data that may be also stored in the MCU (though not in mask or OTP ROM) or in an external EEPROM.

If you ask me the BM867 has a 24Cxx on the top PCB...
https://www.eevblog.com/forum/testgear/brymen-bm-867-teardown-pictures/?action=dlattach;attach=31973;image

BM869:
https://www.flickr.com/photos/eevblog/8513690107/in/album-72157632878242654/

BM525 has a lot of EEPROM must be mainly for the logging.
https://lygte-info.dk/pic/Brymen/BM525s/DSC_8868a.jpg


I have a dump from a Greenlee BM251 if that's of interest.

[mwb1100]

I have a dump from a Greenlee BM251 if that's of interest.

If I'm not mistaken, that's a Greenlee DM-100A == BM251 with a backlight (approx 30 seconds)? 

If someone were willing to monkey around with a DM-100A EEPROM (surely there's calibration data in there, so some risk of really screwing up your meter): I'd look for something like 0x20 in the EEPROM (32 seconds) and see if changing that to a significantly smaller or larger value makes a corresponding change to the backlight duration.  Of course even if that works it only hints to the possibility of a BM867 doing something similar, and almost certainly not in the same EEPROM location.

I have a BM257, but don't have access to an EEPROM reader/writer.

[coromonadalix]

in the brymen eevblog model, i think DAVE stated at one time, it needed a programmer to access some .....  he would not go that far in explanations

For the higher models duno ??

[shapirus]

I don't think anyone knows for sure that there is an EEPROM in the meter. Based on interactions between user joeqsmith and Brymen, it might be that the firmware is stored in non-erasable ROM and the only way to change it is by chip replacement.

There should be. The meter has to store its calibration data somewhere, and there is a procedure to recalibrate them, so it's not read-only.

edit: ...but of course the delay time may be hardcoded in the firmware itself, which may be stored in a ROM, which apparently is the case.

[joeqsmith]

...I suspect there's a half decent chance that there's a byte in the EEPROM for the number of seconds the backlight stays on.

... Based on interactions between user joeqsmith and Brymen, it might be that the firmware is stored in non-erasable ROM and the only way to change it is by chip replacement.

The following was from my discussions with Brymen about increasing the BM869s's backlight timeout.   

Me:   

Is it possible to upgrade the two meters I have with the new firmware?   I would have no problems changing the IC if that is what it takes, assuming it would not cause a problem with the meter's alignment. Enough people have asked about it that I would take the time to demo this feature as I doubt very many people are aware of it.

Brymen:

... BM869s uses OTP chip. Its firmware can not be updated directly. The only way is to replace its U6. To change U6 may have the chance to affect its performance. Generally re-calibration and retest are recommended after replacing chip. ....

...new U6 were sent to you as a special arrangement. To prepare and to send new U6 would be a heavy service burden. We will not offer new U6 to other users to replace their BM869s. Please kindly understand.

Me:

I appreciate you sending out the new ICs for me to try it out.   I will make it very clear when I demonstrate the new feature that the ICs are not available to the general public.

[shakalnokturn]

I have a dump from a Greenlee BM251 if that's of interest.

If I'm not mistaken, that's a Greenlee DM-100A == BM251 with a backlight (approx 30 seconds)? 

It's a DM-200A, backlight is about 30s.
I have a second one around that I'll dump when I clean the switch. It will at least show the memory locations that have identical values on 2 different meters.

Brymen:

... BM869s uses OTP chip. Its firmware can not be updated directly. The only way is to replace its U6. To change U6 may have the chance to affect its performance. Generally re-calibration and retest are recommended after replacing chip. ....

That only makes half sense to me:
Calibration constants can't be in OTP MCU (at the best they could store coarse default values) and various photos show EEPROMs, replacing the MCU is not replacing the ADC so their comment is true only if the new firmware uses different EEPROM addresses for calibration.

[IanB]

Brymen:

... BM869s uses OTP chip. Its firmware can not be updated directly. The only way is to replace its U6. To change U6 may have the chance to affect its performance. Generally re-calibration and retest are recommended after replacing chip. ....

That only makes half sense to me:
Calibration constants can't be in OTP MCU (at the best they could store coarse default values) and various photos show EEPROMs, replacing the MCU is not replacing the ADC so their comment is true only if the new firmware uses different EEPROM addresses for calibration.

If you disassemble the meter, unsolder a chip, and solder in a new chip, that is major surgery on the meter. If you want to trust the meter after such work, it would clearly be sensible to do a re-calibration and retest. It's a recommendation, not a requirement.

[joeqsmith]

Brymen:

... BM869s uses OTP chip. Its firmware can not be updated directly. The only way is to replace its U6. To change U6 may have the chance to affect its performance. Generally re-calibration and retest are recommended after replacing chip. ....

If you disassemble the meter, unsolder a chip, and solder in a new chip, that is major surgery on the meter. If you want to trust the meter after such work, it would clearly be sensible to do a re-calibration and retest. It's a recommendation, not a requirement.

Agree.  During that video, after swapping the part, I run through the same basic checks and didn't notice it having any effect on the measurements.   

***

The only option would be to buy a new meter from current production.

That video is 3 years old.  I am a bit surprised that someone with a older meter would necropost about it today.
***

That only makes half sense to me:
Calibration constants can't be in OTP MCU (at the best they could store coarse default values) and various photos show EEPROMs, replacing the MCU is not replacing the ADC so their comment is true only if the new firmware uses different EEPROM addresses for calibration.

Calibration constants have nothing to do with the backlight timer, which is hard coded in the firmware.   Calibration does not suggest realignment.  As IanB mentions, calibration was only recommended to make sure the meter was still functioning. 

Add a flip-flip to the LED circuit so that pushing the backlight button toggles it on/off.

I like my auto dimming backlight on my UT61E. 

https://youtu.be/S36LfKmSjKw?t=332