BM036 Clamp Meter Magnetisation Test
Вставка
- Опубліковано 17 бер 2024
- Testing if a returned BM036 Clamp Meter has had its core magnetised.
If you find my videos useful you may consider supporting the EEVblog on Patreon: / eevblog
Web Site: www.eevblog.com
Main Channel: / eevblog
EEVdiscover: / eevdiscover
AliExpress Affiliate: s.click.aliexpress.com/e/c2LRpe8g
Buy anything through that link and Dave gets a commission at no cost to you.
T-Shirts: teespring.com/stores/eevblog
#ElectronicsCreators - Наука та технологія
The amount by which the offset has drifted just in the course of your two videos is notable and more indicative of a failing component than anything like magnetisation.
Yes.
The huge amount by which it drifted between the customer posting and in Dave receiving it already showed it wasn't core magnetisation.
@@ferrumignis That should be the first clue, absolutely right. But we don't know how long the meter was in a box before Dave got it out to look at it. The rate of drift within the video just puts a scale to that.
I guess you could try swapping parts with the good unit. Be nice to hear back from Brymen though.
If the clamps are magnetised then it'll just be that to swap to prove the point.
I bet on hall sensor failed internal offset compensation
@@viperwizard491I am going with that also, due to the way it drifted.
I was thinking, move the guts of the good one into the bad one, then you will know if its the clamp or the electronics! If the hall sensor is embeded into clamp....
Hi Dave,
You should try this with the meter switched off, as if it uses a Zero-Flux topology (which it most likely will in order to compensate for core parasitics) it will always keep the magnetic field inside the core at zero, and you won’t be able to measure anything.
Unless it's just using e.g. Hall sensor directly, it will be closed loop. That is, there is a coil (say, 2000turns for a 2000:1 current ratio) that produces the opposing flux, since the magnetic sensor (also probably Hall effect) and the core itself is non linear. And it could also be that those 3 adjustments are for frequency response, since the coil is of course reactive (or the gain of the feedback loop per current range)... Anyway gapsem is correct, with the meter running if that feedback loop is working properly, the flux will be cancelled and constant (near zero). But when the meter says OL (in the low range), you might find the loop saturated, in which case the flux will deviate from that closed loop zero.
They likely wanted to measure peak current while connecting some large capacitive load like an inverter to the battery and that will generate quite a few kA likely above what the meter can handle. Most likely the amplifier after the hall effect got damaged due to fast transient from large capacitor connected to the low internal impedance battery (basically a short circuit).
Peaks could easy be in the 12 to 24kA for those batteries order of magnitude above the meter max current and the rate of change is also huge.
I'd bet money on your and @AnnaVannieuwenhuyse 's comment about "the raw value [being] too far outside the nominal or expected value" to zero out.
There should be an input protection in place for the opamp from voltage spikes (protection diodes) but if there aren't then that could be possible
From your previous video, the fact that you can't use the delta function to null it to zero, and, the 86 amp value didn't change when you opened the jaws, definitely leads it to a hardware issue besides any residual magnetism in the jaws. I am thinking it's a wonky op-amp, or part of it's circuitry, that is causing the issue, seeing that a small change/defect in the op-amp itself, or an associated component, can amplify a defect on a large scale.
It would be worth looking at the magnetic fields with one of those magnetic field viewing cards. I think we are all interested in your final resolution to this mystery. We like puzzles!
Thanks,
Carl
Interesting that the display didn’t change to DC_0 on the faulty meter when pressing the zero button like the working unit does!
Yes, forgot to mention that. It refuses to do the zero for some reason, so the routine doesn't display it
I would think measuring a high AC current for a while would naturally degauss the loop.
The symptom it has where the offset keeps getting larger sounds more like a component failure. The loop would not be getting more magnetized as it sits. Not being able to zero probably means a software limit has been reached and it has applied the largest number it can to bring it back to zero, so internally that thing must be reading a very large current.
While the AC is still flowing, slowly open the clamp and pull away from the wire. The slowly reducing field will work to demagnetize. Switching off a big AC draw while still clamped can make it worse - you won’t know where you are in the cycle.
I doubt it. We're only talking the current you measure flowing in a one turn coil. You're probably not going to get anywhere near enough magnetism to degauss it. Plus, you need the alternating magnetic field not to suddenly stop at a random point and gradually reduce to degauss.
As I suggested in in a comment on the first video -- hit that AMD part (probably a precision op-amp) with some hot/cold. If it's a leaky semiconductor junction that'll likely cause a dramatic change in that offset reading.
Surely the core having somehow got magnetised wouldn't account for the offset steadily increasing?
Make a degausing coil by getting a 50-100 watt power transformer with "E" and "I" laminations. Pull it apart and toss out all "I" laminations. Then restack with all the "E" laminations in one direction. Power it with mains. It will buzz like hell but great for demagnetizing by waving it over magnetized device and slowly moving it away. Only use it for
Glad I'm not the only one to think of using a degaussing coil on it. I'd think a tech like Dave would already have one around.
in previous video it looks like the 3 pin board to board connector had a dodgy solder joint. (visible @ 08:51 in part 1, one of the pin is not soldered on the bottom side).
Just taking an analytic approach: DC-Clampmeters need an integrator circuit, usually realized by one of their OpAmps. For obvious reasons the "Nulling" should not only be done in Software, but also the integrator circuit should be reset.
Because the button beeps, the button and the communication to the CPU are okay. The fact that the value of the integrator is not changing at all, indicates the "reset" signal is not reaching the OpAmp-circuit at all. So principally, I would trace the reset line from the microcontroller to the opamp...
I use a vintage tape head demagnetizer to demag my clamp meters and various tools that get magnetized.
check the hall effect sensor output voltage and connect it to the other meters circuit.
since you have already desoldered the hall effect's front end, why not try and replace it with one from a good one to rule that bit out?
Considering the earth's magnetic field produces a 200mV from that probe when rotating the probe and you weren't even getting half that from the clamp, I'd say it's not magnetised as if that much magnetic field made a difference, the earth's magnetic field would cause the meter to read 180a just from rotating it
My money is on the hall sensor DC preamp; the hall sensor itself would give the same failure mode for ac and dc, and anything after the amp would also be affected equally. The only part in there that knows the difference between AC amps and DC amps is that little preamp (or, more likely, those two little preamps, one for each). Couple thousand amps thru there from some giant f'n forklift battery or whatever and if the preamp isn't suitably protected it's toast.
Do these meters use the hall sensors for AC? I'd have thought they would use the coils shown in the previous video.
Wouldn't measuring a big AC-current work as a degaussing?
Wrap the lead a couple turns to increase the field.
However it´s strange that the offset increases over time. Unlikely that a potential static magnetic field will increase over time...
For what it is worth, an internal oscillator may be used on DC Amps to create an artificial AC magnetic field in the clamp. The AC field is then modified by the DC current field which allows it to be measured. If I saw it correctly, your good meter was on AC, then not generating an AC field. If an oscillator is used, the field might be picked up with a coil and shown on an oscilloscope.
On the working unit you could try zero it while measuring 8 amps. If that works then it's a firmware fault on the broken one
The 'Gauss' is named after the German physicist Carl Friedrich Gauss and his name is pronounced 'gowse' or the 'au' has the same sound as the 'au' in Audi, not Audio or 'gauze'.
There is a smaller 'degausser' used on magnetic tape heads. I did have a look on Amazon and found watch and jewellery demagnetizers/degaussers for around 40AUD. They're certainly cheap enough and may have other uses on the workbench.
I would think that, particularly since support from the plastic jacket means that you don't have a need for mechanical strength, one would choose a "soft iron" class material that doesn't retain permanent magnitization.
If the Hall sensor is indeed buried, it might have a bad solder joint, initially just mechanically in contact, oxidation growing over time Or a fault in the Hall chip's case, allowing moisture incursion.
You need one of those vintage Tandy handheld degaussers for VHS/beta tapes. I’d swear that you got one once in an aging retail cardboard box.
If the core was magnetized, wouldn't the reading drop as soon as you open the clamp?
You can try to find an old CRT TV or monitor; there is a demagnetization coil around the kinescope connected in series with the PTC in the mains. Using the same coil, but connected to DC current, you can try to shift the readings to 0 (so as not to try to find 90 amperes). But I don't think the problem is the peremanent magnetization of the core, it is made of a soft magnetic material with a very low coercivity. The problem is probably either in the sensor itself or in its signal amplifier.
Oddly enough, after your previous video I decided to check my cheapo meter of a different brand and turns out it has the same freaking issue while measuring DC. It constantly shows ~130A and refuses to zero out. It measures AC with no problems, which is what I mainly use it for. The meter is "Forscher 209".
parts swap or get a bigger sample size for the magnetic reading. if good ones all read within a tolerance, and the bad one is outside that range, its a magnetic issue.
if parts swaps fix it, something in the little board or the clamp arms is bad.
if not, its drift in something on the main board.
I guess you can detect the magnetic field of 90A DC with screws and nuts or any small steel parts.
Surely logic would say if this was some issue with magnetisation then the reading wouldn't keep getting larger daily? Seems like some component degrading. How about swapping the hall effect sensor just for fun? This would be really neat to see you actually fix it by parts swapping from another meter so the cause could be found.
My bet is on the hall effect sensor failing. They can and do fail (either mechanically, through corrosion ingress, etc) or electronically (electrostatic failures, etc). Perhaps there was some sort of moisture ingress over time, explaining the slow bias creep.
And I wondered some 35 years ago why measurement cores were made from very brittle ferrite..
Yes... What a memory. With a lot of mates still alive.
Degauss it with an old soldering gun, AC drill, dedicated degaussing coil, etc.
Surely if the core was magnetised, just opening the clamp would massively reduce the offset?
If the clamp meter's ferrite were easily magnetized, it wouldn't be much use because it would constantly be getting magnetized when measuring larger currents.
Wind 4 0r 5 turns of wire around the clamp and connect in series to a heavy AC load. { ie an electric heater would work fine} , There should be more than enough magnetic field to degauss the clamp. Using 4 or 5 turns of wire to replace the soldering tip. on a soldering gun also works well as a degausser
I think that calibration data is stored in eeprom(U5 24L02), I would check if the record is going on or not.
Without knowing how the software is calculating the offset but i doubt there is any NV RAM holding the data. So the offset is being read directly from a part of the device. Not sure what the lam-core is made of for saturation but worth swapping clamp heads, but if the whole thing was subject to very high magnetisation it could be somewhere else, again did not see any lc network but who knows. Have you a Gaussmeter?
Well, I am physicist with a career working on particle accelerator magnets.
Maybe you can try to degauss it by wrapping as much turns around as possible, hook it up to a signal generator or DC power supply.
Go +I, -I, +0.75I, -0.75I, +0.5I, -0.5I, ... 0.
Where "I" is sufficient current to saturate the core.
I've had two DC clamp meters that ended up with offsets like this. One of them was so large the meter always showed OL. The other one only has a 23 amp offset, and I can zero it out before I use it. Would be nice to find a way to fix these!
i have this same issue with my ANENG AN8008 i think it appeared after i measured some frequency on ultrasonic cleaner board and it sparked while measuring since then i noticed that Amperemeter has an offset and it's still working although now it's measuring 0.7amp per amp on top of the offset 🤷♂
You could use a transfomer with a burnt secondary. Taked appart as degausung coil or a squirrel cage motor. Take the rotor out and power it? Then you can just refit the rotor and the motor should still be fine.
I've noticed that the offset seems to be getting worse as time goes on. The customer said it originally had a small offset (about 10 I believe), but by the time of your first video, the offset was around 80. On today's video it seems to be hovering around 90. If it is picking up and retaining a magnetic field, have you tried using a degaussing coil on it to see if that has any effect?
its increasing, it was shipped at 17 or whatever, the last video was at 89 or whatever and this video is at 91, would not be a magnetic field
only thing i can think of is some leakage in the opamp or in the hall sensor it self
i would be curious if it still has its full AC measurement range, granted im not sure how one would test that in the home lab ... maybe a function generator and quite a few wraps of wire
How about trying the old school trick of iron filings and the magnetic field? You would probably take out the iron core and just test that. But its way more fun and visual and lot less sensible to other sources.
My take is, the circuit starting with the "zero" switch and the debouncing cap back to the controller. Q: does the zero function work for ohms?
Give it some sharp whacks with a screwdriver handle to shake the internals about.
either there's magnet particles stuck inside or else hall ic are damaged or zeroing op amp may have faulty caps
Could this be a crack in the core?
You can try connecting a cable from the hall sensors of a faulty device to a working one and from a working one to a faulty one. I'm almost sure the problem will be in the core or less likely in the hall sensor
Would not help to fix it, but i would like to see can you zero it when you get the current measurement near zero.. ie. by loopping 30 loops wire and feeding negative (negative) 3amps to push the reading down and trying to zero it then.
I don't really think it's magnetized. I think it's either a faulty hall effect sensor, or even a faulty DAC. Considering the guy said it got worse over time, it sure does sound like a faulty DAC. And that offset sounds exactly like it could be this, too. Might be that Analog Devices thingamajig, or the other Microchip one, whichever is the DAC. I'd test it to see if it has some uncompensated bias, maybe its input is getting offset in some way (maybe it really is the hall effect sensor).
Maybe a deguassing coil may demagnetise it...???
The type that they used to use for CRTs.
If the cause was the magnetization it makes not much sense to me that the measured current keeps increasing. I don't know how they measure a non altering magnetic field, but my guess is that there is something wrong with the circuit that does that.
AC and DC mode are totally different (transformer vs hall sensor). The faulty one is in dc mode in the video. You could maybe clamp them together to measure with the good one
Will the broken one still measure? If so, try offsetting the reading with a magnet to a value that will zero. Keep doing that until you lower it enough to zero without offsetting.
Maybe the gain setting resistor in the front end opamp is wrong or open?
...Does the offset change when meters circuit board hit with spray freeze / hot air.....I often find this sort of fault can be narrowed down as faulty components are often Heat/Cold sensitive ...
De Gauss the core of the clamp with an audio tape demagnetizer. (not an audio tape head demagnetizer) We all need to be prepared for this possibility. Ron W4BIN
If you used degausser on that, would the hall effect sensor and other circuitry survive? :)
In school in the 80"s we were shown any old electric motor could be used for degaussing. Shown using drill and a vacuum cleaner degaussing a TV!
Maybe ask Brymen if they can suggest possible cause(s)....
I successfuly demagnetized the metal casing of a compass by using a strong magnet attached to a drill. While rotating the magnet I approached the case and then moved away.
But, if the clamp's core was magnetized, wouldn't the reading be influenced by opening and closing the clamp ?! I did not see anything like that.
Just out of curiosity if you try zeroing the clamp while your measuring current through it does it effect the displayed value at all?
I would also try swapping the daughter board. Any ideas as to the function of the AD "09" chip?
I was thinking the same thing. If Dave is willing to do some part swapping, that should narrow down the suspect list very quickly - even if the adjustment pots on the daughter board are not perfectly tuned for the other unit.
Have you checked the contacts of the "ZERO" switch for continuity when the switch is pressed?
it goes beep when button is pressed
Back in the 90s we used a degaussing magnet to wipe floppies. Probably in the 1000s of floppies.
If the clamp meter became magnetised, wouldn't that affect DC current measurement as well as AC??
And surely wouldn't cause the reading to keep increasing over time after the meter has been removed from the strong magnetic field?
I noticed a poor solder joint on the three-pin header of the board you removed. Could that be an issue?
What about the pivot pin or case screws?
My bet is still a defect or failure in the hall effect sensor. Is there a way to evaluate the sensing component for failure?
That would be the next culprit you'd suspect.
@@EEVblog2 I went straight there after the last video. I didn't even consider that the hoop could have taken a magnetic charge. I always learn something new in your videos! 🙂
Thanks again for all you've taught us! 🤓
@@EEVblog2 Are there separate sensors for AC and DC?
My thought was the sensor or opamp factory defective
Board/clamp swap?
I don’t think it was magnetized. For it to be offset by 90A it would have needed to see an equivalent field of thousands of amp turns through it. Maybe a failing Hall effect sensor or amplifier or something.
This video is super well timed for me. I’m designing a current sensor using a gapped ferrite core, and trying to account for the offset induced by remanence field (with open loop architecture) is really a challenge. Especially at cold temperatures ferrite holds a lot of remanence field. I’m curious if anyone knows what material the laminated core is in this sensor. SiFe?
You can change the jaws from ome to the other one to see.
Go the iron filings test and then break out the old color tv deguassing coil.
Board swap. It probably doesn't take too long. It is interesting that the faulty meter does not react to zero-ing as another comment points out.
Cool probe. Try substitution testing to pin that pesky little bugger down... hope you won't need 1.21 jiggawatts to degauss the clamp if that's the case :)
If it was at 86A earlier and is now at 92A, then it's not getting magnetized. Unless you also happen to have a strong magnetic field in your lab?
What happens if you clamp the good through the loop for the bad one?
My guess is that the Hall Effect Sensor has come loose.
Just drive up the road to tricky Dicks house. He must have some spare coils laying around. Just avoid bringing it in an Aldi bag.
Where can I purchase one of the shotty units? (:
You look too keen for that probe in the thumbnail...
My guess would be a failure of one of those surface mounted capacitors.
The good meter shows "dC_0" when you zero the offset. The faulty one didn't. A clear firmware fault.
Or rather than a fault, it could be refusing to zero the measurement out because the raw value is too far outside the nominal or expected value.
@@AnnaVannieuwenhuyseyes, I suspect its this
Plot twist: It is working correctly now. But pots were clogged at factory during calibration, so it was miscalibrated and pots slowly unclogged over time and vibration... Unit just need to be re-calibrated using these pots (these are likely offset+scale?).
Easy way to test is use the probe while passing current thoguh the clamps
But WHY tf hall effect was there at first?
Just why?
You should measure the magnetisation with the instrument off. When it is on it generates a inverse magnetic field to null the existing field in the jaws
Does it measure ACA properly?, If it is DC saturated it shouldn't that would be an interesting test, also pass a demagnetizer close by, I don't think it's magnetic after this video but I have been wrong before 😂
The zero button should work. Magnetized or not.
Use a degaussing coil like for the old CRT or I used a Weller soldering gun move it around in circles then pull away.
Maybe try to demagnetize it with one of those cheapo tool magnetiser/demagnetiser gadgets I think you have.
Clamp it around a high load AC line and let it sit there for a few hours.
Simply put it near a compass and see how magnetic it is.. lf it is de-gauss like we did with WW2 ships?😂
Should be careful he doesn't end up in Philadelphia 😂😂😂 Well, could be a Fran collab then 😀
Cant you just demagnetize it clamping for a moment ac wire with many amps and move it away
I think there is something wrong with one or both of the coils in the clamp.
It ruined anyway maybe you could try to put a magnetic screwdriver or something else close to the clamp, to see if something happens.
You can just wrap some wire around a big screwdriver, attach it to a mains plug. It's what ElectroBOOM would do lol
This looks like a software issue. Possibly caused by bad flash memory in the main multimeter chip.
Since it changes itself upwards, magnetic problem should be ruled out from the beginning.
Looks like a leaky capacitor in zero reference side of hall circuit.