Hi
I need some advice .
The voltage across my shunt (0.1 ohm) when amplified works fine between 0 Amps and 2 Amps.
When I replace the shunt with a much bigger one (15A 50mV) the amplification is not linear.
At 1A I get 400mV then 2A should be 800mA but I get 980mA.
Should I use a better op amp circuit...what needs to be improved?
Thanks.
what needs to be improved?
Your posting skills.
Post the circuit if you want help with that (hand drawn is fine, Fritzing is not).
A 15A/50mV resistor will only generate 3.3mV at 1A.
The input offset of the opamp could well be several mV.
And are you using 'Kelvin' 4-wire connections ? - essential in this sort of application.
Allan
I would also be very happy to see a picture of my circuit on this post.
I did make a hand drawing and take a picture,copy to my pc and attach it to my post but the size was above
the allowed size...so nothing was attached.
Then I did another post to explain why there was no picture and this was also denied ,I am not allowed to post
within 30 sec's.
I will work on something to somehow get a picture or even a link.
What is Fritzing ?????
New forum members have to wait five minutes between posts (for spam prevention), until you get to 100 posts or so.
You will know a Fritzing mess when you see it:
Hi
This is a link to the circuit I use ,sorry for the link but this is the best example I could find in a short time.
https://www.eetimes.com/document.asp?doc_id=1279415.
When working with quick reply there is no tools ,how do I make the link work (turn blue) ??
Thanks.
how do I make the link work (turn blue)
Use the insert link button (infinity, or chain link symbol).
this is the best example
We need to see your circuit, not someone else's.
Easiest thing to do is use the "Reply" instead of "Quick Reply". If you know the tags, you can use them in quick reply
mikedb:
I would also be very happy to see a picture of my circuit on this post.I did make a hand drawing and take a picture,copy to my pc and attach it to my post but the size was above
the allowed size.
What is Fritzing ?????
reduce the size.
Photoshop 4"x4" save as medium quality - 5
mikedb:
HiWhen I replace the shunt with a much bigger one
Thanks.
you put instead 0.1 ohm a 10 ohm ?
Picture of the resistors ?
Hi
This is my circuit.(attachment)
With bigger shunt I mean stronger in Amps ,as I said the shunt is rated 15A 50mV.
So the resistance is much lower than 0.1 ohm.
V = I x R ,the shunt is rated 15A 50mV.
R = V/I
50mV / 15A = 0.00333... ohm.
It seem like the op amp does not amplify linear at this low resistance.
Picture of big shunt ?
It is warm ?
There is nothing wrong with the circuit, it is working for 0.1 ohm, you changing only resistor so resistor is causing your problem. At 15 temperature of big shunt is increasing so resistance of it, put it in to water or use fan to cool down the shunt.
At 15 temperature of big shunt is increasing so resistance of it, put it in to water or use fan to cool down the shunt.
I doubt if this is the problem - the dissipation of the shunt with 15A at 50mV is only 0.75 watts. Not much. And I suspect it's physically quite large. And at 2A it's only 0.0133 watts. Tiny.
It's value will be 3.33 etc mohms .
So see my previous post for suggestions.
Note my recommendation of 4-wire connections. Invented by Lord Kelvin a century and a half ago for exactly this sort of problem. Want a circuit? Or look it up.
And what op amp are you using?
You'll get 3.3 mV per ampere, so opamp performance is significant. A cheapo LM358 won't do.
Allan
To find out i suggest to connect 0.1ohm shunts in parallel - 5 or 10
allanhurst:
15A at 50mV is only 0.75 watts.And at 2A it's only 0.0133 watts. Tiny.
Allan
This is a huge difference in power so can affect the temperature of the shunt.
If it's rated at 15A it shouldn't be worried by such a current, and shouldn't overheat sufficiently to affect accuracy.
Else the OP has been sold a pup, which I doubt.
A cheapo 1watt wirewound resistor sold for 10p wouldn't have any problems running at 0.75 watts and would probably give 100ppm/C accuracy or better. 0.05% or so with 50C rise...... Good enough? (You can buy 20ppm/C, but they're rather dearer.)
I suspect his shunt has at least as good a spec.
link, OP?
Allan
My point is described in first line of post # 12
To OP
Just blow on shunt and observe - the reading is stable or not.
If it's dissipating 0.0133 watts (the heating effect of 2A flowing) , any effect will be due to the temperature of your breath..... the resistor won't be significantly heated. And I doubt if anything will be observed unless there are other circuit problems.
OP - any comments? I'd worry about other stuff.
Ted - what about the temperature coefficient of resistors ? How stable are they? can you observe these effects, and if so how? What measuring equipment would you use? And how certain would you be of your measurements?
Resistor manufacturers over many decades have measured and documented their devices using labs equipped with the best kit costing a lot , accurate to < 1ppm, and traceable back to national laboratories such as the NPL here in the UK..
Reckon you can improve on their stuff with a breath?
NiChrome wire has a temperature coefficient of about 55ppm/C. Is your 1% ish multimeter good enough to measure this? Only if you heat it with a blowtorch.....
Allan
