Ah batery capacity logger

[Mosaic]

hi all:
I am making a kit to do this.
http://www.clubjameco.com/index.php/projects/project_brief/24/21263/auto_battery_capacity_logger


please have a read and comment. or login and upvote it! if you like it.

thanks!
Ancel/ReVolt

 

[ronv]

I registered, but couldn't see much except the brief description. I voted it up anyway.

 

[Mosaic]

 

[ronv]

Okay, I got it now. You are just starting. A couple of things to think about:
Most lead acid batteries are spec'd at a certain current otherwise the amp hour rating will vary depending on how high the current is. This is probably only important if you want to return a battery under warrantee. Otherwise you just need some constant current that you can find on your battery curve.

 

[Mosaic]

I could implement constant current control since I am using PWM, but, since the approach reports the average current which is within the ballpark of the peak I didn't see it being critical to determining the Ah. If you dial up a 6A initial discharge, the avg at the end of the rating is around 5.3A. This is actually similar to a user applied 'fixed' load. By running two different discharges you can generate a Peukert index which will then predict any discharge current life.
The chart axes are Y= Voltage, X =# samples (225 sec period)

 

[NorthGuy]

Deep 100% discharge may be harmful to most lead-acid batteries, even for those marked "deep discharge". So, by the time you reveal the whole Peukert curve, the battery might be very seriously harmed. I prefer not to know the exact capacity of my batteries.

 

[alec_t]

I agree with NorthGuy, assuming you aren't just doing a test-to-destruction for academic purposes. As with all things electronic, it's always good to have a safety margin.

 

[Mosaic]

Yes, every time you deep cycle a battery it counts as a used cycle. The 10.5V reference is equiv. to 1.75V per cell which is the 'official' method of testing capacity:
**broken link removed**
Some times a deep cycle is required to help 'exercise' a sulfated battery or one that has been equalized or to simply accelerate a new battery to max cycle performance. Such an exercise helps reform the sponge lead and reduce agglomerations of hard sulfate thereby increasing overall battery performance after the batt. has been in use for some time.
I recommend a C/10 discharge rate as opposed to C/20 to reduce the actual % deep discharge of the battery but still give a useful indicator of viable capacity.

@ Northguy: I have been doing extensive lead acid battery testing and recovery over the last year and having 'processed' dozens of batteries I can say that they are fairly robust and a controlled cycling will not 'damage' anything. In fact, for sulfated batteries a post equalize deep cycle boosts the available capacity by around 5%.

 

[Mosaic]

Wow, only one upvote from the whole forum....

 

[Mosaic]

Hi All:
The Kit has gone public as of today:
http://www.clubjameco.com/index.php/projects/project_details/24/21263/auto_battery_capacity_logger

Full DIY details are in the article first published here at ETO.

 

[MrAl]

Hi,

That's interesting because i had made something like this several years back and sold it. It was a four channel unit that could measure four batteries at the same time because i wanted to do NiMH batteries to prove that the manufacturers where shipping out crappy batteries. That one interfaced to the PC computer though for display and control.
It could also be used as a charge controller for charging the batteries too, as well as measure their temperatures.
Basically it's just a four channel voltage measuring device with specialized software to serve for the different functions.

Good luck with your project.

 

[Mosaic]

Thanks, did you use bidirectional rs232 to control?

I built this cause I needed one for the lead acid R&D I am involved with. I didn't care for the method of using transistors & power resistors as loads with large aircooled heatsinks. The dichroic lamps serve as decent and cheap & visible loads @ .50 ea from elec. goldmine. I haven't burnt out one yet in many hours of operation. It seems that since I don't run them at FULL load most of the time they last quite long. Not being behind a glass lens and facing upward also keeps them cooler, from convection.

EDIT: I added an image of the item in the first post.

 

[MrAl]

Hi,

That's a nice looking board, what is the two digit LED for?

I didnt need bidirectional although i intended to add that at a later date.

For mine, the user would use whatever load they wanted, and whatever max voltage (set with external resistors of their choice) for versatility. Here's a screen shot of the software with the "Setup" tab showing. The "Capacities" tab is where the ampere hour capacities show up after the test is complete. It's a very simple layout but effective.

 

[Mosaic]

That interface looks based on Visual Basic or C++?

The two digit displays a few things.
1) Alternate battery voltage Integer & decimal for precise voltage calibration to compensate for cabling & connection losses . eg 12.74
2) Single integer & decimal for current load setting up to 9.9A.
3) Cumulative Ah while running.
4) Undervolt 'Lo', to avoid testing an undercharged battery.
5) Hi Ah >99 - 'Hi'
6) 'UP' when ready to uplink the discharge log to PC

It's a standard size, single sided board spaced for DIY , all thru hole 'standard parts'.
This is the PCB layout

 

[MrAl]

Hi,

I dont use Visual Basic or any other Basic anymore. So i used mostly C++ and another less known language called "Euphoria". This other language is an interpreted language which is fast enough for most human graphical interface programs.

The multi functional LED is good as that shows more than one thing, it's just not apparent from the photo because it just shows "12".

 

[dr pepper]

I've done something remotely similar, I built a display and control for an ebike motor, one of the functions was to show real time motor wattage consumption and also how many amp hours had been used since the last resset.

 

[flatfootskier]

I have an ACT gold battery tester. In the space of a few seconds it displays capacity in Ah for 6 or 12 volt batteries. Can anyone explain how it does it?

 

[dr pepper]

An estimation of ah can be done by disharging a battery at a rate and measuring the v drop over a short time, knowing the characteristics of the battery (sounds like lead acid) you can make an assumption as to the actual ah, but its only an approximation, the discharge curves will change as the battery ages.

However are you sure your tool measures ah, the automotive battery checkers that I have come accross do not measure ah, they measure cca, cold cranking amps, as when your starting a car engine its more important to know that your battery will put out the amperage it specifies when its cold than its actual ah, I'm not a pro on this however sulphation which is a common killer of vehicle batteries probably affects cca more than a/h as it insualtes the battery plates.

Cca is something like the max current the battery will put out for 10 secs at its min operating temp (-10c) without the voltage dropping below 10v, ie enough to fire up your old truck when its flipping cold.

 

[flatfootskier]

Definitely measures Ah. I find it surprising that the slope can be measured accurately within a few seconds. It seems accurate to with about 10%

 

[MrAl]

Hi,

They are probably using a capacitive model for the battery, so that discharging it a little measures the capacitance which correlates to the Ampere Hour capacity.

For example, if you have a capacitor of unknown value you can measure the capacitance using:
dv=I*dt/C
or:
C=I*dt/dv

where I is the current (constant), dt is the change in time, and dv is the measured change in voltage.

There is a slight catch here in that there is a smaller secondary capacitance that must be discharged first so this measurement can not be made in a short time like 100us like you could with a real capacitor, but instead have to allow time for that smaller cap to discharge first.