Why does my shunt have blobs of solder on it? :p

[AmmoJammo]

Says EX530 on the board 





Do these meters have a current calibration? or is it assumed in the factory that the shunts are calibrated, so it could be reading wrong now?

[staxquad]

Do these meters have a current calibration? or is it assumed in the factory that the shunts are calibrated, so it could be reading wrong now?

we've seen shunts filed down, pinched with cutters to increase resistance to adjust the shunt, so I guess adding solder would decrease resistance

doubt the solder has the temperature stability of manganin though (adding non-linearity), so would be a quick fix only (the manufacturer doesn't care about the quality of their product)

proper companies usually start with more metal than needed and file it down or pinch it till the resistance is correct

[amyk]

As odd as it looks, it's for calibration.

[kizzap]

wouldn't it be more likely that the solder on the shunt is an incidental part of the manufacturing process, and was simply placed there by a clumsy hand?

It looks like one of the solder joints wants to break off too.

-kizzap

[AmmoJammo]

one is a blob... the other looks more like an accidental smear

This is a Digitech meter, sold by Jaycar, its the same as the Extech EX530... however, I dont think its the same build quality as a lower spec meter from the same range I bought ~5 years ago...

[pickle9000]

That one pad looks very poor.

[AmmoJammo]

which one?

edit: the solder must just be for calibration... the other meter I just pulled apart again has had the shunt pinched/cut to calibrate it.

[BravoV]

which one?

edit: the solder must just be for calibration... the other meter I just pulled apart again has had the shunt pinched/cut to calibrate it.

Nope, that is a shitty solder joint, what meter is that ?

[AmmoJammo]

Nah, that solder joint is fine

[Stonent]

During one of Dave's multimeter reviews there was a shunt that looked like someone had hacked away at it with an axe.

I believe it was discussed in the "shaved shunt" discussion on the forum. 

[AmmoJammo]

During one of Dave's multimeter reviews there was a shunt that looked like someone had hacked away at it with an axe.

I believe it was discussed in the "shaved shunt" discussion on the forum. 

I found this:
https://www.eevblog.com/forum/testgear/teardown-extech-ex570-my-first-but-also-last-extech/

Must say, I'm quite happy with the construction quality of my meter now 

[kizzap]

I can comprehend and completely agree with the fact that a saw or snips can increase resistance on a shunt, being that it gives a smaller path for the current to flow. I can't though comprehend how solder on a seemingly random spot, with no care taken to ensure the solder joint is uniform will decrease the resistance, especially when ideally the cross-sectional area of the original shunt would ideally be uniform. If the original shunt has a uniform cross-section, wouldn't the resistance be the same along the length of the shunt?

I would believe that it would help if some "subtractive adjustment" had been done on the shunt and it was a kind of surgery to fix it...but still...

Dave, do an experiment for us?

[codeboy2k]

I can't though comprehend how solder on a seemingly random spot, with no care taken to ensure the solder joint is uniform will decrease the resistance, especially when ideally the cross-sectional area of the original shunt would ideally be uniform. If the original shunt has a uniform cross-section, wouldn't the resistance be the same along the length of the shunt?

The blob of solder makes the cross sectional area at the point of the solder no longer uniformly the same as the cross-sectional area everywhere else
The formula for resistance of a uniformly circular wire is

where is the resistivity of the medium, is the length of the wire in meters and is the cross sectional area in square meters. Thus is the radius of the wire.

This works because it's uniform, and you can multiply it by the length. But the solder blob adds a cross sectional area that is changing along the length of the wire, now you have to integrate it. And the resistivity is different at that point too, so you would have to do the two regions separately to get the total resistance, and these two regions are in parallel. The math for an unknown blob of solder would be hard, at best.  But the resistance will still decrease as you increase the total cross sectional area at that point where the solder blob is.  Basically, it's adding a small parallel resistance at that point, it has its own resistivity different from that of the base shunt wire, and it has its own tempco, and its own total resistance along its length, which is hard to calculate because it's oddly shaped.

Adding a blob of solder may not seem precise, but they are tuning it at the factory, so they are watching the measurements as they tune it.

As an example that is easier to visualize, consider the image below, I've soldered 4 chip resistors together, the size doesn't matter, assume they are all value 100 ohms. You have no trouble believing that the structure in the image is 250 ohms: 100+50+100. without the top resistor, it is 100+100+100 = 300 ohms. So adding the top resistor lowers the resistance because there is a parallel resistance there. You can visualize increasing the length of the lower section with more and more 100 ohm resistors in series, but the section with two in parallel will always be less resistance, i.e. 50 ohms, at that spot.

Dave, do an experiment for us?

Yes, Dave can prove it with an experiment

[AmmoJammo]

I'd say that this meter needed the resistance of the shunt reduced, due to the long negative WIRE running from the ground terminal on the multimeter, to the shunt

Whereas, most other meters need the resistance increased to calibrate it (bits ground/cut out).

[kizzap]

Awesome explanation, thanks for taking the time there. I completely forgot that resistance is also dependant on the length of the conductor 

[AmmoJammo]

This meter iiiiissss ssoooooooooooo ssssssllllllllloooooooooooooooooooooooooooooooooow... 

[Harvs]

Can anyone tell me why they would bother trimming the resistance of the shunt? 

First thing that would happen is the shunts voltage would be digitized (probably with some initial analog gain and CMR.) Then offset and scale would be applied in software.

How far off would the shunt have to be for this not to be easily compensated in software?  Like 20%?

[retiredcaps]

This meter iiiiissss ssoooooooooooo ssssssllllllllloooooooooooooooooooooooooooooooooow... 

Yes, that is what Martin commented on as well.

[AmmoJammo]

This meter iiiiissss ssoooooooooooo ssssssllllllllloooooooooooooooooooooooooooooooooow... 

Yes, that is what Martin commented on as well.

I know... but its not till you actually use it that you realise just how incredibly slow it is!

I might end up getting a lower spec meter for daily use, and just use this when I need high resolution...