Battery voltage level to ECU

[Andrew]

I am looking to send the voltage of a 24V battery pack back to the ADC
of my microprocessor. This is to monitory the current voltage of the
battery, which will ultimately be used to display this voltage when
necessary. I recognize that when there is a significant current draw,
the voltage of the battery will be lower than under no-load conditions.
But, to measure the voltage level of the battery simply when it is in
a no-load state, can I just use a simple voltage divider, with a cap on
the lower leg to capacitate the voltage output to the ADC? What are
the drawbacks of this method?

Is there a better way to do this so that I can check the "power left in
the battery" at any given time, and not have to wait for a no-load
condition?

Thanks,
Andrew

 

I am looking to send the voltage of a 24V battery pack back to the ADC
of my microprocessor. This is to monitory the current voltage of the
battery, which will ultimately be used to display this voltage when
necessary. I recognize that when there is a significant current draw,
the voltage of the battery will be lower than under no-load conditions.
But, to measure the voltage level of the battery simply when it is in
a no-load state, can I just use a simple voltage divider, with a cap on
the lower leg to capacitate the voltage output to the ADC? What are
the drawbacks of this method?

What type of battery is this? If it's a lead-acid you can get a
reasonable guess at the capacity remaining by measuring the
open-circuit voltage. Most designs tend to switch in a dummy load
resistor if they're not actually running off battery, though.

 

[Mochuelo]

On 19 Jul 2005 13:32:16 -0700, "Andrew" <xxragexx@gmail.com> wrote:

I am looking to send the voltage of a 24V battery pack back to the ADC
of my microprocessor. This is to monitory the current voltage of the
battery, which will ultimately be used to display this voltage when
necessary. I recognize that when there is a significant current draw,
the voltage of the battery will be lower than under no-load conditions.

Right.

But, to measure the voltage level of the battery simply when it is in
a no-load state, can I just use a simple voltage divider, with a cap on
the lower leg to capacitate the voltage output to the ADC?

Right. The cap should be in parallel with the resistor that is in
parallel with the ADC input.

What are
the drawbacks of this method?

None, if the two resistors are high enough so that the divider does
not draw a significant amount of current from the battery. The higher
the resistance, the lower the load effect, but also the higher the
noise being picked up.

Is there a better way to do this so that I can check the "power left in
the battery" at any given time, and not have to wait for a no-load
condition?

It is hard to accurately know the energy left in a battery. The output
resistance of the battery is a better indicator (it grows as the
battery drains), but it is also more difficult to be estimated (for
this, you need to measure voltages and currents with two different
loads).

Thanks,
Andrew

 

[Roger Lascelles]

"Andrew" <xxragexx@gmail.com> wrote in message
news:1121805136.930899.244290@g14g2000cwa.googlegroups.com...

I am looking to send the voltage of a 24V battery pack back to the ADC
of my microprocessor. This is to monitory the current voltage of the
battery, which will ultimately be used to display this voltage when
necessary. I recognize that when there is a significant current draw,
the voltage of the battery will be lower than under no-load conditions.
But, to measure the voltage level of the battery simply when it is in
a no-load state, can I just use a simple voltage divider, with a cap on
the lower leg to capacitate the voltage output to the ADC? What are
the drawbacks of this method?

Is there a better way to do this so that I can check the "power left in
the battery" at any given time, and not have to wait for a no-load
condition?

If "ECU" means you are controlling an engine, then I think you are on the
right track. Trust me, there is an awful amount of R&D involved if you want
to get into true battery condition and battery charge monitoring - you can
build a whole $100 million a year company on it, with dozens of engineers
nailing down all the parameters.

For decent instrumentation of the electrics of an engine, I recommend you
display battery voltage and battery current, in real time. That will cover
all the following

- charge system faults (low voltage when on charge)

- low voltage during cranking, caused by battery fault, aging, corrosion of
terminals or starter fault.

- refusal of battery to accept charge, even though charge voltage is
correct.


I would not bother trying to relate battery OCV to state of charge - just
display the voltage. OCV is an excellent indicator of state of charge if
the battery has not recently been charged or discharged. Unfortunately
state of charge is not all that much use - a 5 year old worn out battery
still shows correct OCV when charged. I suggest you "don't go there" - just
have good metering.

The basics are : 1) if your charge system delivers the correct (temperature
compensated) voltage during charge, you are doing the right thing by the
battery. 2) if battery voltage holds up during cranking, battery and
starter and connections are OK. If either is out of whack you can display a
warning, but you don't have to diagnose exactly what is wrong.

Roger

 

[Andrew]

This system is not for an engine. I'm not sure that I mentioned ECU
but if I did, I meant MCU. The specific application and design of the
system will not support realtime display of the voltage and/or current
(or any other values for that matter). These numbers will be polled
every so often, as needed.

I realize that this subject can get into some depth, but this
application certainly does not warrant a full out battery monitoring
system. As long as the voltage of the battery can be monitored, I will
be all set. The exact amount of life remaining, and the such, is not
necessary. If one minute the battery says 25V and two minutes later it
says 20V, that is fine, I just need to know, and be able to report!

Thanks for your help though guys, I just needed some confirmation,
because it seemed too simple.