Cycle analyst v2.3 current sensing with a hall sensor
- Thread starter SpmP
- Start date
cal3thousand
10 MW
SpmP said:
The CA does current sensing. Are you doing something else that I don't quite understand?
And as for 5V, yes there is a 5V pad on the CA's PCB that you can tap into.
CA does current sensing through an external shunt for me (High power version)
I need current range up to 500 or 600A, the shunts are horribly expensive and not available locally. BUT I have loads of linear hall sensors and thought I could make a circuit to monitor the current using this, the 5v supply a MAX232 to give me the inverted power that i need to onvert 0-Vcc to -Vcc - Vcc with 0 for 0 urrent (Hall sensor gives Vcc/2 as zer), but again, a big PITA to calibrate accuratly enought for coloumb counting etc.
So I wanted to know if anyone has done this, if there is any reason to DEFINATLY _not_ do this. Then I guess I need to make the output 'look' like a 50mV shunt.
THey say shunt input must be +-200mV, and it would be good to use the full range.
Also if anyone has done this it may save me some time trying to remember all my op-amp theory and making the level shifting (with trimpot) and amplifying circuit.
There must be a MAX232 on a pcb here somewhere 8)
One of my bigger problems is I dont have a 50A or 100A load to put on the wires to calibrate. Have a good fluke current clamp... but a load... hmmm...
I need current range up to 500 or 600A, the shunts are horribly expensive and not available locally. BUT I have loads of linear hall sensors and thought I could make a circuit to monitor the current using this, the 5v supply a MAX232 to give me the inverted power that i need to onvert 0-Vcc to -Vcc - Vcc with 0 for 0 urrent (Hall sensor gives Vcc/2 as zer), but again, a big PITA to calibrate accuratly enought for coloumb counting etc.
So I wanted to know if anyone has done this, if there is any reason to DEFINATLY _not_ do this. Then I guess I need to make the output 'look' like a 50mV shunt.
THey say shunt input must be +-200mV, and it would be good to use the full range.
Also if anyone has done this it may save me some time trying to remember all my op-amp theory and making the level shifting (with trimpot) and amplifying circuit.
There must be a MAX232 on a pcb here somewhere 8)
One of my bigger problems is I dont have a 50A or 100A load to put on the wires to calibrate. Have a good fluke current clamp... but a load... hmmm...
Dunno - seems you are turning a simple problem into a science project.
Unlike ammeter shunts which demand extreme accuracy, you can make calibration adjustments with the CA - this relaxes the solution requirements considerably.
It's fairly simple to fabricate a workable shunt - the trick, of course, is to use a material that gives temperature stability - particularly important for shunt heating at high currents. Copper is not a good choice - commercial shunts use Manganin. Alternatively, you can parallel up some lesser shunts to achieve the current capacity that you need - this will also give you the desired temperature stability. There will be some very small thermal error introduced by the copper wire/bus bars bridging the shunts, but this should be of little consequence due to the short length.
Unlike ammeter shunts which demand extreme accuracy, you can make calibration adjustments with the CA - this relaxes the solution requirements considerably.
It's fairly simple to fabricate a workable shunt - the trick, of course, is to use a material that gives temperature stability - particularly important for shunt heating at high currents. Copper is not a good choice - commercial shunts use Manganin. Alternatively, you can parallel up some lesser shunts to achieve the current capacity that you need - this will also give you the desired temperature stability. There will be some very small thermal error introduced by the copper wire/bus bars bridging the shunts, but this should be of little consequence due to the short length.
- Grin Tech offers a 300A (800A max) 0.25mOhm shunt for $18.00 (pn Shunt0.25). Using two of these in parallel will give you the required 600A continuous rating (1600A max) and a nominal 0.125mOhm resistance. This is above the CA V2/V3 shunt minimum on Hi Range of 0.077 mOhm.
Thank you for the info.
I ended up using the cable from the battery, about 2m long. But now I see why this is not normally done.
How much would you expect the resistance to change, and does this make a big difference?
Certainly buying one from grin is the way to go, but delivery would still be more than a week away.
I ended up using the cable from the battery, about 2m long. But now I see why this is not normally done.
How much would you expect the resistance to change, and does this make a big difference?
Certainly buying one from grin is the way to go, but delivery would still be more than a week away.
The 'use a piece of wire' gimmick is an old hack that many use. It gives fair results, but if you want to see the actual error you will need to work the math for you particular case using thermal coef - see something like 'resistivity' in The Engineering Toolbox or similar. Again, this is for vehicle not laboratory use so you can reasonably get by with a few percent of slop and never notice.
I did this for a winter night to summer day temperature range and get about 1.5% error. This cannot be right as otherwise everyone would do this!
What I cannot account for is heating due to current load. Then we get back into science experiment land, using an ntc or something to measure temperature blah... 8)
What I cannot account for is heating due to current load. Then we get back into science experiment land, using an ntc or something to measure temperature blah... 8)
Similar threads
- Question
