Shunt measuring errors help needed...

[rg12]

Hey Guys,

I'm having problems with my amp hour meter (came with a 100A shunt) and was wondering if the following can cause such errors like I'm experiencing of 20-40% off in the amp numbers.

The shunt has 4 connections, two big ones and two small ones for the meter.
I have soldered another wire to one of the meter connections in order to pull a negative wire to power the meter screen.
Can that solder change the measuring somehow? if so, can it be by that much?

Also, I trippled checked and everything is installed correctly, but what would happen if the big wires that go through the shunt are reversed (flipping between the battery and load sides), what can happen then?

 

[nutnspecial]

You can pull meter ground from the input side of the shunt if you don't want that measured, or from output side if you do, but you shouldn't pull ground from the measurement wires. Or you could pull 5v and ground from the throttle in some cases, for the meter. It shouldn't matter what orientation the two shunt wires are afaik- it's just testing the difference in resistance between them.

This could verywell be the issue you mentioned before! If for nothing else imo it would be great for posterity to link the two subjects/topics.

 

[rg12]

Found the issue but not the solution...
Just found my old 100V 100A meter which always worked great and it was insync with my 60v one which was always reliable...seems that the new one that isn't accurate is showing lower amps when pulling high amps...
If I pull 200w for a minute or two, the new (non accurate one) is accurate just like the old reliable one, but when I pull about 50A+ it collects the Ah number slower than the rest.

 

[rg12]

Also, I noticed that the amp meters that were accurate had the positive wire going through the shunt and the non accurate meter had the negative wire go through the shunt.
The accurate numbers are the higher amp ones that went through the shunt using the positive wire as opposed to the lower amps measured through the negative wire.

 

[izeman]

can't really help with your special problem, but the current going through the positive and negative wire is exactly the same.
i don't know if you know how a shunt works: what it does is measuring the voltage drop through the shunt. a shunt is made of a special material that does not change its resistance over temperature. that is because the shunt gets warmer as current runs through it. a "regular" piece of copper changes its resistance by about 0.4% per kelvin.
you can't use the thick wires to measure the voltage drop directly as those wires have a voltage drop as well as current is going through them. so you need those small sensing wires which measure voltage only, and have no current going through them. the resistance of the shunt is fixed. and the meter is calibrated for exactly this shunt used. if you use a different shunt you need to recalibrate the meter if you need a precise measurement. if you put some solder on the shunt or add wires the resistance will change and the meter will read wrong values.
hope that helps

 

[amberwolf]

Unfortunatley on really cheap stuff actual shunt wire may not even be used, and they might just use regular wire...so at higher currents the readings will be different from lower currents just becuase of the heating. Ambient temperature will also make a difference to overall accuracy.

Most of the time, even when they use actual shunts or wire, the device has no way to calibrate it to the actual shunt resistance (which will be slightly different on every unit for various reasons), so every meter will read at least a little different, and sometimes a lot different.

Makes no difference whether you measure thru positive or negative wire; that's not the reason your meters are different accuracies.


Devices like the Cycle Analyst, which has a menu option to input the actual shunt resistance, can be much more accurate, as long as you test the shunt you're using for it's actual resistance and input that number in there. Ebikes.ca has info on how to do that, as well as the CA itself for purchase. But it's not cheap, becuse it's not just a wattmeter.

 

[rg12]

Actually the meter that is lagging on higher amps (counting less amp hours) is the more complex one with the external 100A shunt.
The accurate one has a shunt that looks like a fat FET with two thick legs and 3 smaller legs.
I guess I will solder the thing to 12AWG wire and hook it to my 90A rig and see if it survives as I pull high amps only in bursts.

 

[amberwolf]

The FET looking thing is actually a hall current sensor, and is usually more accurate than the shunt types, presuming the electronics reading it are calibrated for it. Just more expensive than a shunt.

The external shunt one...if it's inaccurate, make sure it's calibrated for the actual shunt resistance, meaning you'd need to follow one of the many guides on testing your shunt's true resistance (see ebikes.ca or poke around the forums in Teklektik's posts for examples), and then change the meter's shunt setting to match that. If it can't be changed, then that's why it's inaccurate, and is a bad design.