Ampere-Hour meter

[IRQ57]

Hello,

I need a help to build an Ampere-Hour meter, with electronic display only, without memory, the simplest one, just start and reset.

Thanks in advance.

 

[Russlk]

Is this AC or DC? I guess it doesn't matter, you would rectify the AC with a precision rectifier and feed into a A/D. You will need a microprocessor to accumulate the readings and adjust them for ampere-hours.

 

[IRQ57]

It is for a DC current, max 15A 300V. May I use a PIC ? or something less complex?.
I need to accumulate AH for about 20 hours, then reset it and restart.

Up to now I could not find a diagram in the Internet, even a block one to start. Any help in this sense will be appreciated. Thanks.

 

[IRQ57]

It is for a DC current, max 15A 300V. May I use a PIC ? or something less complex?.
I need to accumulate AH for about 20 hours, then reset it and restart.

Up to now I could not find a diagram in the Internet, even a block one to start. Any help in this sense will be appreciated. Thanks.

 

[Nigel Goodwin]

If it's for DC it makes it a lot easier - for a start you need to generate suitable signals from the voltage and current you wish to monitor.

Opamps will probably be required, one monitoring the voltage drop across a small resistor in series with the load, and the other monitoring the actual voltage across the load. For the load voltage, a simple resistive divider is probably all that's required.

Measuring these with two analogue inputs on a PIC will give you the voltage and current, you may need to scale the readings to get the actual values - then simply multiply them together to get the power consumed.

If you only want the current, you don't need to measure the voltage, but if you are monitoring power you will need both.

 

[chemelec]

There is a Watt-Meter Circuit on my site. Its for 50/60 Hz and for a nominal 110 volts, but can be modified for 220/240 if needed.

Go to my Projects area at: **broken link removed**

Take care.........Gary

 

[IRQ57]

Chemelec,

Thank you very much, I visited your site and found plenty of very interesant projects, really very good. I saw the circuit diagram for the Watt-Meter, and it is for instant readings; I need to accumulate data during periods of about 20 hours.

Nigel,

Thanks too for your approach, I need to read a current sample, may be each second, then store it in a register and add to another. the next step would be to divide the latter by 3600 (seconds/hour) and display it. Am I right?

 

[Nigel Goodwin]

Thanks too for your approach, I need to read a current sample, may be each second, then store it in a register and add to another. the next step would be to divide the latter by 3600 (seconds/hour) and display it. Am I right?

It depends entirely what you are trying to do, could you be more specific?.

 

[IRQ57]

Hi Nigel,

I need to register the total charge that a string of batteries are receiving, and when the number of AH required is reached, stop the process. That is all. Thanks.

 

[Nigel Goodwin]

Hi Nigel,

I need to register the total charge that a string of batteries are receiving, and when the number of AH required is reached, stop the process. That is all. Thanks.

So this is just for a battery charger then?.

 

[bogdanfirst]

if it is for a battery charger than i don't think this is a good idea.
first, the charging process is not done with 100% efficency. so you need more than the capacity of the battery. energy is lost as heat(the battery heats) and in the conversion from electric to chemical energy.
i think that the best way to monitor when a battery is charged is to monitor the voltage.
i think that this idea is good for discharging.
but anyway, why do you need 300V?
now, i think that your way is the best.
you can monitor the current and read the value at different intervals. that way you consider that between 2 intervals the current has stayed constant.
so you add the values of the current mesured and then divide the sum to the time. i think this is oky, if the currend doesn't vary very fast. for improoving the accuracy you could decrease the interva vetween 2 readings.
anyway, what do you need this for?

 

[spuffock]

suggest a differential amplifier to produce a voltage proportional to current, feeding an integrator. When the output of the integrator reaches a preset level, a constant current source is switched to the integrator for a short known time, and at the same time a counter is advanced. It will cope with fast variations in current, even spikes, without losing accuracy.

 

[IRQ57]

This is not exactly a battery charger, It is to formation of battery plates. This involves an electrochemical process which needs a very tight control in current, temperature, and time .

For each pound of material processed, there is a stablished number of AH needed, and due to the temperature rise in the containers during the formating process (exothermal reactions + electrical power losses), the current fed to the string of plates can´t be constant, unless using a very low value (and a long process time). That´s why it is necessary to compute the total AH given to the plates to be sure that the chemical conversion is complete, avoiding damage caused for incomplete process or wasting energy in unnecessary and harmful overheating (max 55°C).

 

[IRQ57]

To Spuffock,

Your suggestion sounds very interesting and not complicated. But, please, could you explain a little more the function of the constant current sent to the integrator?

Thanks a lot.

 

[spuffock]

The constant current for a fixed time is a constant amount of charge. The charge to be measured is integrated and drives the integrator output positive. When the threshold is reached, the constant current for fixed time puts a standard "bucket of charge" into the integrator to drive the output negative. Unless the integrator gets to hit the rails, charge in=charge out, and the number of"buckets" * constant current * fixed time = measured charge. The device I built is used to measure the battery consumption of micropower devices that use a microprocessor that spends much of its time asleep. It gets pulses of several milliamps for milliseconds on top of a constant few microamps. It can measure to 1/10 microcoulomb per count, accurate to a couple of percent.

 

[bogdanfirst]

what about using a computer?
you could mesure the value of the current much faster. than you could do the rest in softwre...
though, this solution is not portable

 

[spuffock]

I bet I could make something to draw spikes of current and make a sampled reading useless :twisted:
I built my device ten years ago. I didn't include an on/off switch, so it's on.
It runs on batteries so it can float at a couple of kilovolts if you like.
And I'm just thinking about changing the batteries. The original ones might leak!!

 

[IRQ57]

Spuffock,

Thank you very much for your support, I will start right now working with your idea. If it were possible, to shorten the way, could you send me a diagram?. Thanks so much anyway.

 

[Nigel Goodwin]

Right, from what I can gather, you want to feed a certain amount of AH into a device - your existing idea is to measure the current over time and calculate the AH, once AH equals your required value - you stop the current. Is that correct?.

This seems a very long winded way of doing it, how about this much simpler way:

Feed the device from a constant current source, and simply time how long it's on for. As an example - if you want 10 AH, use a 1A constant current for 10 hours - or a 500mA constant current for 20 hours - or any combination you want.

Far simpler than trying to measure things!.

 

[IRQ57]

Thanks Spuffock,

I got very clearly your description of the "buckets", now I am working in this idea. I will use an Op Amp to take a sample in the current sensor resistor. the output will feed a capacitor thru an adjustable resistor. The voltage will be sensed by pins 2 and 6 of a 555 (in a monostable conf), and pin 7 will discharge the capacitor, giving pin 3 a pulse to be counted as a unit of charge and displayed in a seven segment led display.

Regards and my thanks to all the participants in this forum.