Amper hour meter

[AmperMan]

Hello Folks,

First post so be gentle. Hope this is not a repete post too as I tried posting some time ago and got a blank page.

I want to make an Amper hour meter (not a Watt hour meter) to measure coulombs going into an electrolythic process. The voltages involved will be low (below 12 Volts).
The total Amper hours needing to be counted will be large at around 50,000 (100 Amps going for few weeks).
The whole outfit is DC and there is no requirement for to measure reverse current (not a battery charging/discharging application). I can program a PIC16F628 and have an LCD for display. I want nothing fancy but reasonable accuracy is required (a good stable resistor for running current through and measureing Voltage accross I guess). Is there any other method that someon would recomment other than the resistor to measure the currnt. Hall effect sensor?
I have googled and seen some designs using an op amp and a low value stable resistor. Could I not just use the Analogue to digital converter on the PIC and not bother with the Op-Amp.
Any one got any circuits that they would like to share.
Is there any simpler way it can be done so that the PIC is not needed.
I cannot use a KILL-A-Watt (off the shelf) as the power consumed by the system is not a constant in relation to the current going to where I need to measure it.
Thanks,
John

 

[crutschow]

Measuring the current across a low value resistor (shunt) is an easy and accurate way to measure the current. Look for a 50 or 100mV, 100A shunt.

You need a low DC offset type op amp to amplify the voltage across the shunt and get better precision and accuracy from the normal input voltage range (typically 0-5V) of a PIC A/D. Alternately there are some Sigma-Delta type high precision A/D converters which might work directly from the 50 or 100mV voltage from the shunt. But then you have to interface the A/D to the PIC.

A PIC is probably the easiest way to keep track of and display the total ampere hours.

 

[dougy83]

You would want to use an opamp to get the voltage across the shunt/sense resistor from a small value up the the 5V (or VCC) range of the PIC. W/o the amp, the 5V full scale range @ 100 amps needs a resistor capable of handling 500W, not to mention eating up 5V of your [assumed] 12V total.

If you get a 100A 100mV FSR shunt, it drops a max of 0.1V (still wastes 10W). To get to 5V FSR, opamp should have gain of 50. You can either an ultra low offset opamp (if you can find one) or you'll want to compensate for the input offset at the opamp. https://en.wikipedia.org/wiki/Operational_amplifier

 

[SABorn]

Look up INA195 it is a current shunt monitor chip from Texas Instruments, just made for this type of application, it will interface direct to a pic.

They are a wonderful chip and very easy to use, except for the fact it a smd chip but not to small to work with.

The other option would be to use a DC current transformer ( ct ) and run one of the supply cables through the ct, then there is no cutting into cables or power loss with resistors.

Have the pic send the data out serial and log the amp hours on a computer.
We already do this sort of thing for wind generators and solar systems so the software is easy to download free.

You could also log the 12 volt supply and get a watt/hour reading relevant to the operation.

Pete.

 

[AmperMan]

Fancy Hall effect current sensors

Hello,

Thank alot for the replies. I understand about needing the OP-Amp now.
Looking around I seen some fancy Hall effect current sensors here:
**broken link removed**
Very small Voltage drop accross them but they don't go from zero to (say) 100 but rather from 50 to 100 Amps etc. Not what I want.
Some fancy animations too.

John

 

[nickelflippr]

Allegro makes a series of uni and bi-directional hall effect sensors ACS758xCB that might do the job (along with a microcontroller).

 

[Reloadron]

You have a few options. You can run with a simple DC current transducer along these lines in a maybe 0 to 100 amp or 0 to 150 amp flavor that will give you a nice 0 to 5 volts output proportional to the current. Next there is the simple current shunt method where you can use maybe a 100 amp or 150 amp shunt where 0 to full scale current gives you either 0 to 50 mV or 0 to 100 mV full scale output. Obviously using the latter shunt method requires amplification of the mV output to be scaled to something a little more usable. Amplification can be an off the shelf solution or you can roll your own depending on how much work you want to do and if the project is commercial industrial or a home project.

Next you mention 100 amps or so running for weeks. How often do you plan to sample the current and store the result? While a PIC is an option you do mention weeks? A sample rate of one sample per second is 3,600 samples per hour and 86,400 samples per day and then 604,800 samples per week. Where do you plan to store the data? If you want stand alone I don't see a PIC as a very viable option depending on what the sample rate should be. Then too, I am not sure how you plan to average the data collected. Do you want to see dates and times of current peaks and valleys?

Given a choice I would get the current using an item like the first device mentioned. A current transducer with a scaled output. I would run that voltage into a simple data logger and set it and forget it.

Just My Take
Ron

 

[RCinFLA]

A battery monitor will do the AH measurement.

**broken link removed**

Shunts also available.

 

[MrDEB]

One issue with using a shunt resistor is heat. if I read it right, 100 amps would generate alot of heat. The resistor when heated will change the resistance value thus your data accuracy.
Ti as well as Maxim have chips that are designed for this.

 

[crutschow]

One issue with using a shunt resistor is heat. if I read it right, 100 amps would generate alot of heat. The resistor when heated will change the resistance value thus your data accuracy.

Shunts designed for 100A use low temperature coefficient resistance material so the temperature change from the 5 or 10W they dissipate has only a slight effect on accuracy.

 

[AmperMan]

Thanks for all the replys.
First I forgot to mention that I am a cheap skate!!!!!!!!!!!

Batter monitors ($150) are way too expensive.
I would make my own shunt from Nichrome wire (about 2mm dia., I forget exact gauge No.) available from ebay for bugger all.

About all the data storage. I will be storing very little data. Just obtaining a running total of Amper Hours. Wil sample current every 10 seconds (very approx.) and add Amper hours to total amper hours. Only one piece of data! (total amper hours) to be stored. I do not want a graph of current or voltage or power.

The DC current transducers look good. They go from Zero up to whatever.
Are they expensive. Who can supply.

John

 

[nickelflippr]

Looks like back earlier we cross posted on the Allegro hall effect sensors. If you happened to look at the data sheet, you would see that the ACS758-100U is a unidirectional sensor that goes from 0-100A with a Vout of 40mv/A.

 

[Reloadron]

The DC current transducers I linked to run about $100 USD.

Making a shunt from nichrome wire is not a good idea. For a current shunt you want a very low temperature coefficient for the sensing element:

The type of material selected for the elements has a very low
temperature coefficient and the size and number of
elements chosen give the optimum surface area to dissipate
the maximum specified power in air for a particular value.

Nichrome wire has a high temperature coefficient making it great if you want to build a toaster but lousy for a DC Current Shunt.

Ron

 

[jpanhalt]

Thanks for all the replys.
First I forgot to mention that I am a cheap skate!!!!!!!!!!!

Batter monitors ($150) are way too expensive.

What do you figure your time is worth, seriously? Have you ever programmed a microcontroller with an ADC? How long do you think it will take you to learn to do that?

If you are anywhere near Cleveland, OH, I would offer you a day-labor job at 150% of what that hourly rate would turn out to be.

John

 

[Reloadron]

What do you figure your time is worth, seriously? Have you ever programmed a microcontroller with an ADC? How long do you think it will take you to learn to do that?

If you are anywhere near Cleveland, OH, I would offer you a day-labor job at 150% of what that hourly rate would turn out to be.

John

Not if I get him first....... Ha, Ha, Ha

Ron

 

[AmperMan]

Thanks nickleflipper, we did indeed crosspost!

Will have to read up on Temp. Co. Eff.
I thought Nichrome had a low Temp. Co. Eff.

"If you are anywhere near Cleveland, OH, I would offer you a day-labor job at 150% of what that hourly rate would turn out to be."

But then I would not have thaught myself how to program a PIC.............

 

[crutschow]

Nichrome has a low temperature coefficient as compared to copper but it is still about .0004 per degree C. So, for example, a change in temperature of the wire of 25C would give 1% change in resistance, thus it wouldn't take a lot of temperature change from wire self-heating to seriously degrade the accuracy of a current measurement using a nichrome shunt.

 

[Reloadron]

Carl. at 100 amps he would need a seriously large nichrome wire wouldn't he? I just don't see nichrome wire as a viable solution for a 100 amp shunt. Maybe I am wrong?

Ron

Ron

 

[SABorn]

For around $10.00 off ebay you can have a 100A (or even 1000A) purpose built shunt, so i question why anyone would want to stuff around with making their own of a questionable quality.

Pete.

 

[crutschow]

Carl. at 100 amps he would need a seriously large nichrome wire wouldn't he? I just don't see nichrome wire as a viable solution for a 100 amp shunt. Maybe I am wrong?

Ron, I agree. That's rather what I was driving at. You would have do have rather large gauge nichrome wire and it would need to be fairly long to minimize the temperature rise with 100A of current.