Brymen BM037 clamp meter bug: false readings in Crest mode?

[shapirus]

I was going to post this a long time ago, and shot a demonstration video, but never posted.

My Brymen BM037 has a strange bug: in Crest mode, which is made to measure momentary max and min values (within a few ms window IIRC), it shows readings ~4-5 times higher than the actual value is. Normal continuous measurements work fine. It's not a user error: there's no unaccounted for inrush current, it can be seen with plain DC constant current.

Now, the meter shown in the video has this bug in both the 60 A and 600 A ranges. I did communicate with the distributor and they did their best to resolve this. They even contacted Brymen I believe, but it resulted in pretty much nothing. They proposed me to return the meter and refund the money, but I didn't want the money, I wanted a nice clamp meter with good peak current registration.

I don't remember precisely what happened then, but I believe they exchanged the original unit (shown in the video) with another, and the one that I currently own has this bug only in the 60 A range. The 600 A range works fine, and Crest measurements aren't different from normal ones when measuring a stable DC current. I think that was where I considered the issue settled: resolution in the 600 A range is sufficient, as I don't usually need precise measurements of inrush current, and for anything precise and/or low-current I'd need to use a dedicated shunt or Hall sensor connected to an oscilloscope anyway.

Does anyone else have this bug? If not, what's your firmware version (hold the Hold button during power-on)? Mine shows "030d".

Any idea if it can be fixed?

Here's the video showing the issue with the old unit that had it in both ranges. Unfortunately I didn't record what its firmware version was.

[electr_peter]

How do you know which meter shows correct reading? What are the specs?
Why would you expect that a cheap DC load supports ideally constant DC current?

It could be a bug, but there is not enough evidence to support it.

[shapirus]

How do you know which meter shows correct reading? What are the specs?

BM869s:

Code: [Select]

Crest mode (Instantaneous Peak Hold)
Resolution: 5000 counts
Accuracy: Specified accuracy ± 100 digits
for changes > 0.8ms in duration

DC current base accuracy is 0.5%+20d in the 10A range.

BM037:

Code: [Select]

CREST (Peak-Hold)
Applicability: Voltage and Non-invasion Current functions
Accuracy: Add +/- 250 digits to specified accuracy for changes > 5ms in duration

DC current clamp base accuracy in 60A range is 1.0% + 5d.

Why would you expect that a cheap DC load supports ideally constant DC current?

Because it does, it can be verified with an oscilloscope. Of course I did verify this. Not ideal (nothing is), but quite sufficient.
Besides, this bug can be reproduced with any other load, including a plain resistive load (and I did that as well, of course). Take a wire with a dead constant DC current flowing in it, take a measurement with the BM037 in normal mode, and it's well within spec, then, without changing anything, enter Crest mode, and voila, it shows several times (it's not like 30% off) the actual value of the current. Besides, as I mentioned, the 600 A range in my current meter works fine, this issue is only present in the 60 A range. So I assume they maybe forgot to calibrate it at the factory or something like that. Or maybe there's a hardware or firmware bug -- that's why I'm interested to hear if others have seen this behavior too.

[electr_peter]

Thanks for clarification. I was thinking about BW limitations and momentary spikes with DC load, etc.
From you description this really does sound like a bug in BM037.