He sees that the owner soldered onto the boarder and instead of saying. Well you might have damaged it doing that, warranty void. He say’s no worries. And acts grateful that the owner took a crack at diagnosing it. A true gentleman. Even in business. Well done.
GQ Electronics was like that with my Geiger counter with a faulty LED. Said the warranty would be preserved if I wanted to replace it myself. Sent me a new LED (and extra). Saved me the trouble of shipping it back, etc.
Going to say a ceramic capacitor in reset circujt or doing another power rail is cracked or gone low. So pulse to hold on is now too short to hold power rail till processor can latch on.
The second one is my multimeter. I'm glad/relieved you could replicate the issue. I was watching the video thinking "stop turning on, damn it! You're supposed to be faulty" haha.
I was watching the show and I was thinking if the owners of the meters where watching and when the 2 meter didn't work first and then it kept switching on that the owner thinking that's not good. That was a bit like taking the car to get fixed and the fault disappears when its there and show up when you get it home again
@@peewee3ie Yeah, I'm just glad I wasn't watching live. The suspense would have killed me haha. Everyone would have been thinking "some clown has returned a perfectly good multimeter".
I have owned several different brands of meters going back 30 years and currently use the EEVBlog BM786 Brymen. The meter is good and I have liked it, but if it ever breaks down "like new", I will never use that brand again because all the other meters still work.
So buy Brymen BM869s and never have any problems! At least for 2 years in the laboratory and at the workplaces, every day work I have not had any problems. There is even a case of working for 1 hour at - 5°C. My Brymen BM869s has 2.5 minutes of screen backlight support. Greetings from Bulgaria
17:13 To me this looks like a 'high side switch' with inrush current management, (RC on the gate of the high side FET) rather than a latch C58 is not populated, but I think there is supposed to be a cap there. Without it, the FET switches too abruptly, causing a dip in the power-supply, causing the controller to screw up.
Dave, regarding the first four digits of the serial number being the date code, are you certain the last two of the first four digits are the year? I have a BM786, and the first four digits of mine are 2132. Possibly a 2-digit year followed by a 2-digit week. Incidentally, I've had no issues with my BM786. I just love that meter!
I have seen fault like this on other equipment where the problem was dirty push buttons. The CPU were trying to boot but by having a push button pressed when it was not expecting it (during boot) it would not work. After cleaning the push button contacts, it would work.
C67? vs C109. They don't exist but the poles C67 and C109 are interesting. As soon as the Q18 is unlocked, I'm curious about the status of the C65. Oscilloscope please. Why is resistor installed in C61. next to C3 and C2. It's as if the automatic parts selector, after looking around for a bit, replaced the C61 with a resistor without showing it to anyone. R78 in leg 1 of U20? If there is an R78, the regulator installed here is the adjustable one. Then where is the R76. 41:07 Where is the crystal (4.9152Mhz)? Do you think I installed the same software on thousands of devices and used the external oscillator, and then decided to use the internal oscillator on a few? What engineer does this? The polarity of the battery is marked on the back of the board, whereas the C54 was not marked on the side.
7:39 So you edited this and still left the part in where you say they're a year apart? They're a week apart, the first two digits indicate the year. It makes no sense to restart the serial numbering at 01 at the beginning of a new year.
Have experienced uC startup problems back in 1980s, as the crystal oscillator was flakey startup. Ultimately traced to uC manufacturer specified crystal was poorly matched to the uC oscillator on the chip. As a result occasionally the oscillator did not reliably start at power up. Correct crystal and matching capacitances are important to assure oscilator to begin. A marginal change to crystal Rs characteristic and the matching capacitors in oscillator circuit allowed more reliable oscillator startup. After changes only at very sub zero temperatures was the crystal oscillator able to rarely fail to start. Most likely source of oscillator startup is incorrectly selected capacitors on the crystal path, a mere change, only for example 15pF vs 22pF may change startup significantly. Very dependant on the chip oscillator design, and that can even be affected by chip batches originating from different chip manufacturing plant /site or chip redesign shrink over manufacturing life of a chip
This. Feels like oscillator. Or possibly it isn’t starting up until there’s some physical vibration to get it going. That’s why turning it fast gets it going. Also, F Twitter. Cesspool. :/
1.202 mA. Reminded me of Apollo 11 - The 1202 program alarm was a warning from the Apollo Guidance Computer that its core processing system had been overloaded.
Do you have the ability to re-flash the meter firmware, and see if that fixes it? I've been noticing many things with flash seem to die after a couple of years, like the flash gets corrupted. It would be interesting if a reflash would solve the issue.
Maybe MOSFET Vgs(th) variation combined with the circuit being just on the edge means some aging changes the Vgs(th) enough to not conduct? Or it might be a processor startup issue, or the crystal not starting up.
Could be something in FW boot sequence not following the ok/normal code path. E.g something not initialize correct and gets the FW stuck. I.e FW do not handle the situation correct, missing error handling.
I've got an ancient Extech meter that has an eerily similar problem. Powers up in only certain switch positions, not always the same ones. Works fine once powered up. And now that I think about it, it seems to happen more often if it's been turned off for a few hours. Almost certainly not a switch or battery contact problem. Actually been living with this annoyance for a few years. Hm.
Through some deep thought and a few conversations with ChatGPT for its amazing insights, I think I've figured out the root cause of the fault... It's buggered!
I'm wondering if something in the programming is having trouble coping with rapid switching of functions - but only on meters with certain switches. (Break before make, or the exact order contacts switch, etc.) Just something about the mechanical tolerances bringing it out in certain units. I just wonder if swapping the switch's rotor from another unit will fix this.
Very weird. 1.2 mA... maybe the processor is starting, but the multimeter chip isn't? I wouldn't be surprised to learn the display initialization only happens after the multimeter chip is known to be running. Could be a good troubleshooting video, especially since you have a good one to compare against... 🙄😉
@reedreamer9518 TES 2620 vs Tektronix DMM91x which one? Tektronix DMM 914 Pressing the blue button will reduce the resistor open circuit voltage below 0.3V normally 0.85 I liked it a lot. The TES 2620 cannot measure capacitors above 100uF, but after Tektronix I love it.
Only the users paying for the blue tick of shame get visibility now on Twitter. I don’t make money with my social networks, so I rather use a social network that doesn’t buries me if don’t pay. And we better not talk about the whole sea of scammers and extremists making a huge mess in there. I’ll keep paying for TH-cam Premium and watch your awesome videos on here.
He sees that the owner soldered onto the boarder and instead of saying. Well you might have damaged it doing that, warranty void. He say’s no worries. And acts grateful that the owner took a crack at diagnosing it.
A true gentleman. Even in business. Well done.
GQ Electronics was like that with my Geiger counter with a faulty LED. Said the warranty would be preserved if I wanted to replace it myself. Sent me a new LED (and extra). Saved me the trouble of shipping it back, etc.
And he's so modest about it. Other people would pin your post to brag about how good they are.
Going to say a ceramic capacitor in reset circujt or doing another power rail is cracked or gone low. So pulse to hold on is now too short to hold power rail till processor can latch on.
It is Amazon which decides it though, so Dave doesn't even get a choice.
The second one is my multimeter. I'm glad/relieved you could replicate the issue. I was watching the video thinking "stop turning on, damn it! You're supposed to be faulty" haha.
I was watching the show and I was thinking if the owners of the meters where watching and when the 2 meter didn't work first and then it kept switching on that the owner thinking that's not good. That was a bit like taking the car to get fixed and the fault disappears when its there and show up when you get it home again
@@peewee3ie Yeah, I'm just glad I wasn't watching live. The suspense would have killed me haha. Everyone would have been thinking "some clown has returned a perfectly good multimeter".
I have owned several different brands of meters going back 30 years and currently use the EEVBlog BM786 Brymen. The meter is good and I have liked it, but if it ever breaks down "like new", I will never use that brand again because all the other meters still work.
i was thinking about buying one, changed my mind now. thanks for the video :)
So buy Brymen BM869s and never have any problems! At least for 2 years in the laboratory and at the workplaces, every day work I have not had any problems. There is even a case of working for 1 hour at - 5°C. My Brymen BM869s has 2.5 minutes of screen backlight support. Greetings from Bulgaria
17:13 To me this looks like a 'high side switch' with inrush current management, (RC on the gate of the high side FET) rather than a latch C58 is not populated, but I think there is supposed to be a cap there. Without it, the FET switches too abruptly, causing a dip in the power-supply, causing the controller to screw up.
Dave, regarding the first four digits of the serial number being the date code, are you certain the last two of the first four digits are the year? I have a BM786, and the first four digits of mine are 2132. Possibly a 2-digit year followed by a 2-digit week. Incidentally, I've had no issues with my BM786. I just love that meter!
Well spotted. Correct, I got it backwards. First two must be the year. I have one here with 2044
Get Bryman to send chips back to silicon mfg for failure analysis
I have seen fault like this on other equipment where the problem was dirty push buttons. The CPU were trying to boot but by having a push button pressed when it was not expecting it (during boot) it would not work. After cleaning the push button contacts, it would work.
Deja vu. Like a lost dream where only parts are remembered lol
Just sold 3off old Fluke 85/87 and awaiting a 2nd Brymen 786. Must admit, I do like the Brymen DMMs.
They are very nice. Occasional failures happen of course.
i prefer the brymens over flukes, but i do miss the nice yellow color
however if someone offered me a free 289 or 101 😂
Unfortunately I have the exact same problem with mine. Was there ever a fix?
Don't turn it on, take... oh wait, never mind.
Don’t turn it on, shake it apaaaaart! 😂
Can't turn it on? Take it apart
C67? vs C109. They don't exist but the poles C67 and C109 are interesting. As soon as the Q18 is unlocked, I'm curious about the status of the C65. Oscilloscope please. Why is resistor installed in C61. next to C3 and C2. It's as if the automatic parts selector, after looking around for a bit, replaced the C61 with a resistor without showing it to anyone. R78 in leg 1 of U20? If there is an R78, the regulator installed here is the adjustable one. Then where is the R76. 41:07 Where is the crystal (4.9152Mhz)? Do you think I installed the same software on thousands of devices and used the external oscillator, and then decided to use the internal oscillator on a few? What engineer does this? The polarity of the battery is marked on the back of the board, whereas the C54 was not marked on the side.
7:39 So you edited this and still left the part in where you say they're a year apart?
They're a week apart, the first two digits indicate the year.
It makes no sense to restart the serial numbering at 01 at the beginning of a new year.
Have experienced uC startup problems back in 1980s, as the crystal oscillator was flakey startup. Ultimately traced to uC manufacturer specified crystal was poorly matched to the uC oscillator on the chip. As a result occasionally the oscillator did not reliably start at power up. Correct crystal and matching capacitances are important to assure oscilator to begin.
A marginal change to crystal Rs characteristic and the matching capacitors in oscillator circuit allowed more reliable oscillator startup. After changes only at very sub zero temperatures was the crystal oscillator able to rarely fail to start.
Most likely source of oscillator startup is incorrectly selected capacitors on the crystal path, a mere change, only for example 15pF vs 22pF may change startup significantly. Very dependant on the chip oscillator design, and that can even be affected by chip batches originating from different chip manufacturing plant /site or chip redesign shrink over manufacturing life of a chip
This. Feels like oscillator. Or possibly it isn’t starting up until there’s some physical vibration to get it going. That’s why turning it fast gets it going.
Also, F Twitter. Cesspool. :/
1.202 mA. Reminded me of Apollo 11 - The 1202 program alarm was a warning from the Apollo Guidance Computer that its core processing system had been overloaded.
Do you have the ability to re-flash the meter firmware, and see if that fixes it? I've been noticing many things with flash seem to die after a couple of years, like the flash gets corrupted. It would be interesting if a reflash would solve the issue.
Tried it, doesn't solve it. But powering through the programming power eliminates the power-on problem.
Maybe MOSFET Vgs(th) variation combined with the circuit being just on the edge means some aging changes the Vgs(th) enough to not conduct? Or it might be a processor startup issue, or the crystal not starting up.
Could be something in FW boot sequence not following the ok/normal code path. E.g something not initialize correct and gets the FW stuck. I.e FW do not handle the situation correct, missing error handling.
I've got an ancient Extech meter that has an eerily similar problem. Powers up in only certain switch positions, not always the same ones. Works fine once powered up. And now that I think about it, it seems to happen more often if it's been turned off for a few hours. Almost certainly not a switch or battery contact problem. Actually been living with this annoyance for a few years. Hm.
I'm inclined to think it's an MCU problem if not a dodgy capacitor (I haven't seen the whole video yet)
So is it worth buying the BM786? I am currently choosing between the BM786 and the UNI-T UT71C. Which one should I choose?
Through some deep thought and a few conversations with ChatGPT for its amazing insights, I think I've figured out the root cause of the fault... It's buggered!
I appreciate your deeply technical analysis! :)
Either that, or it's rooted.
I'm wondering if something in the programming is having trouble coping with rapid switching of functions - but only on meters with certain switches. (Break before make, or the exact order contacts switch, etc.) Just something about the mechanical tolerances bringing it out in certain units. I just wonder if swapping the switch's rotor from another unit will fix this.
Maybe some bad batches? The BM786 is my daily driver, and I haven't had any issues.
Very weird. 1.2 mA... maybe the processor is starting, but the multimeter chip isn't? I wouldn't be surprised to learn the display initialization only happens after the multimeter chip is known to be running. Could be a good troubleshooting video, especially since you have a good one to compare against... 🙄😉
no twitters dying, everybody's bailing on it and hates it now
LOL, nope. All my metrics are increasing.
@@EEVblog2 - well, of course... from a low base almost anything is an increase.
It's just the left wing nut jobs that are fleeing. They can't stand getting their feelings hurt when someone speaks the truth.
Dump it and buy a Kaiweet.
Go, take it apart!
Intermittent short between mcu pins maybe?
Keep posting to twitter as watch on TH-cam as TH-cam notifications are very hit miss.
So you killed it by powering it from 12V?
WTF didn't you try removing and replacing the batteries once you got it into the switched on but not working state???
0:50......Yeah, sure Dave....That'll be it....🤔🤔🤔 Don't patronise buyers like that....
I have many meters, but my Fluke 107 is my daily go-to meter. Just say'en.
I would love to have a fluke 107. But I can’t afford it after buying 5 $30 meters.
🤣😂🤣😂😊
@@ecospider5 Haha - yeah, at some point I too realized that buying cheep can be more expensive.
As my mate Harry with his Russian wife says: I do it again so you can see it is not a Flu*e !
@reedreamer9518 TES 2620 vs Tektronix DMM91x which one? Tektronix DMM 914 Pressing the blue button will reduce the resistor open circuit voltage below 0.3V normally 0.85 I liked it a lot. The TES 2620 cannot measure capacitors above 100uF, but after Tektronix I love it.
F Twitter. Elon destroyed it
LOL, nope. All my metrics have increased since he took it over.
It'll be better now a WEF puppet is the CEO.
Only the users paying for the blue tick of shame get visibility now on Twitter. I don’t make money with my social networks, so I rather use a social network that doesn’t buries me if don’t pay. And we better not talk about the whole sea of scammers and extremists making a huge mess in there.
I’ll keep paying for TH-cam Premium and watch your awesome videos on here.
And the root cause is....?????
Still, TL;DW. Dave being clueless.
Dave is investing in stock, like Elon is convincing to buy Twitter stock, Dave also thinks he will be Richie rich.
Déjà vu
F twitter.
Crap
Wow - that now makes TWO boring and pointless videos from Dave on the same subject. Gee, are you paying us viewers for our time you wasted? 👎👎