BM235 dead on arrival (but easy repair)

[jasonb]

I ordered a BM235 off of Amazon back in September. After it arrived, I noticed it was reading low on the AC and DC voltage ranges compared to other meters. At the time, I didn't have much spare time to investigate, so I set it aside. I finally got back to it this weekend and took a closer look at what was going on. I'd often get zero readings at low voltages (i.e. below about 10 volts), and then for higher voltages, it would read low (e.g. 40V for a 240V source). I also discovered that touching the buttons (not pressing them, just touching them) was often enough to change readings, sometimes resetting readings to zero and sometimes making them accurate.

I took apart the meter and discovered that R12 (edit: I initially said R18, which was incorrect) was loose. It's a large, through-hole resistor, and jiggling it a little, it came completely out of the via. It looks like there was no solder on that leg of the resistor at all. The attached images show how clean these are (the via above the SW5 lettering is where the leg should be).

I soldered the resistor into place and that seemed to fix the problem. I had to recalibrate the temperature range, but otherwise the meter seems fully functional after the repair.

Anyway, I just thought that was interesting and maybe this would be useful if anyone else is running into a similar problem.

[BeBuLamar]

Good info. However you must have gotten it for a very low price so that you didn't return it.

[Simon]

Just to be clear, was it EEVblog branded (blue) or brymen (red)?

[jasonb]

Just to be clear, was it EEVblog branded (blue) or brymen (red)?

It's a blue, EEVblog-branded meter. I'm in the US, and I'm not aware of any place to buy Brymen-branded meters here other than eBay. The only ones I've seen are rebranded as Greenlee meters with green holsters. But the blue, EEVblog meters are available on Amazon.

Good info. However you must have gotten it for a very low price so that you didn't return it.

I'm a terrible procrastinator sometimes, which is all there is to not returning it. I did pay full price for it on Amazon ($151 USD, with sales tax).

[NoisyBoy]

Unfortunately I can’t see what you’re trying to show in the picture. But glad that you realized there is a problem, as it could have seriously mess things up when it was so far off.

Even better, I am glad you found the problem, thanks for sharing.

[jasonb]

Apologies if the photos aren't clear. Here are slightly better photos with the relevant bits circled. I've also added a photo showing what the reference board looks like in the EEVblog manual. The first photo shows the resistor leg not in the through hole. The second photo shows the reverse side of the board. The third photo shows the reference board with solder on that through hole.

[EEVblog]

Apologies if the photos aren't clear. Here are slightly better photos with the relevant bits circled. I've also added a photo showing what the reference board looks like in the EEVblog manual. The first photo shows the resistor leg not in the through hole. The second photo shows the reverse side of the board. The third photo shows the reference board with solder on that through hole.

Yeah that's a problem. Seems like the leg has broken, possibly through shock or vibration. But given that there have been no other reports of this, it was probably a production issue. Possibly a kinked lead that weakened it and then shock took it out.
Resoldering it should fix it as you discovered.

[AVGresponding]

Apologies if the photos aren't clear. Here are slightly better photos with the relevant bits circled. I've also added a photo showing what the reference board looks like in the EEVblog manual. The first photo shows the resistor leg not in the through hole. The second photo shows the reverse side of the board. The third photo shows the reference board with solder on that through hole.

Yeah that's a problem. Seems like the leg has broken, possibly through shock or vibration. But given that there have been no other reports of this, it was probably a production issue. Possibly a kinked lead that weakened it and then shock took it out.
Resoldering it should fix it as you discovered.

It was clearly never soldered; if it had been, the back side of the board would have looked normal, but the through-hole is empty.

[EEVblog]

It was clearly never soldered; if it had been, the back side of the board would have looked normal, but the through-hole is empty.

Oh, right. Once of those missed production steps that can make it through calibration and testing.

[AVGresponding]

It was clearly never soldered; if it had been, the back side of the board would have looked normal, but the through-hole is empty.

Oh, right. Once of those missed production steps that can make it through calibration and testing.

The pic clearly shows there's no solder nor component lead in it   

So, either the OP is lying, and desoldered it himself, or it wasn't soldered in the factory and did pass because it was a good enough contact until a good bump during shipping dislodged it.

[jasonb]

Apologies for not responding sooner. I missed that there were more responses.

There was indeed no solder there. However, the component leg was in the hole when I received it. I mentioned in my original post that when I jiggled it, it came out, and I took the pictures after that. I certainly was able to get it to work properly sometimes before I took it apart, and it was correctly calibrated other than on the temperature range.

My belief is that it was never soldered (the pics seem to make this clear), but that it had good enough contact during calibration that it didn't affect it. Likely that was just a missed production step. It likely came loose during shipping.

[J-R]

My understanding would be that the SMD component side of the BM235 PCB would be done via solder mask, pick and place and reflow oven.  The through-hole components I'm on the fence a little but looking at various photos, I think the opposite side of the main PCB as well as the input PCB are populated and soldered by hand rather than wave soldered.  I'm basing this primarily on the inconsistency of the soldering.

From what I see on other tear-down photos, R12 appears to be soldered in vertically, then bent over afterwards.

If you zoom in on the photo of R12 provided by jasonb, you can just see a ring of solder on the pin showing that the component was only inserted a small amount and it also didn't receive a very large dose of solder.  Maybe there is a close-up available of that to make it more clear?