battery capacity tester

I am trying to design a relatively simple capacity tester for a 12v battery. I want to put the battery under a load and measure the time it takes to drop the voltage to 10-10.5 v. I thought of the circuit in the first attachment, but can't work out the zener voltage or resistor values to get it to shut off at the target voltage.
A more knowledgeable person suggested the circuit in attachment 2. I could not make it work without reversing the PNP transistor (collector to emitter). It then seemed to work, but will not shut off anywhere close to the desired voltage.
Any suggestions appreciated.

your original circuit is suitable for testing.

What is the NPN transistor type.number you are going to use ?
based on that the resistor values can be tried.

N.B:if possible attach the schematic in png/gif/bmp formats than the doc.

You might want to look at one of these Turnigy power analysers.
It monitors volts, amps, watts, and records watt hours and amp hours with very fine resolution.
Perfect for monitoring and recording the results of a battery discharge test.

**broken link removed**

You will still need your circuit to switch off the load at the minimum discharge voltage, but once it does switch off the exact amp hour and watt hour figures and length of discharge time are right there on the screen. Very accurate too. And you don't need to hold the discharge current exactly constant during the test, so a simple resistor or lamp load is all you need.

I bought two of them for testing and comparing solar panels. highly recommended.

Thanks Tony. Do they keep a record of the time alongside the voltage, etc. That is, could I load the battery, and just make sure I didn't pull it down too low; then go back and see at what amp hours it reached my cut off voltage?

- - - Updated - - -

Srizbf,
The transistor in the original schematic is a 2N4401.
Any suggestions on zener or resistor values appreciated.
I'm working on getting the schematics converted to a different format. I don't have a converter program and I am fearful about loading a dangerous one. (I am as ignorant of computer stuff as I am of electronics.)

What happens when you first power it up, it starts off with everything reset to zero.
Time, amp hours, and watt hours all go to zero.

It immediately shows voltage, current, and watts, and time starts to increment, amp hours, and watt hours start to count up too, assuming there is some current to cause that.

If the current then falls to zero at some point, as it will when your load disconnects, as long as there is still voltage to power the unit, it will continue to display the final accumulated total amp hours and watt hours which will hold at the final values.

So you can start a battery discharge test, and go away and just leave it.
Come back much later, and the final total readings will still be there on the screen for you.

Time starts when the instrument first powers up, and it keeps going until you power it down.
Time does not stop when current falls to zero.

The instrument is normally powered from the circuit being monitored, but in some circumstances that might fall to zero, in which case the instrument can get its power from a second source, there being a small power socket on the side to do that.

One last thought.
When the voltage falls to the point where the load disconnects, the battery will then be unloaded, and the voltage might start to rise slightly. You don't really want the load to switch back in, so it may need to have a system that ensures that cannot happen. Once it turns off, it stays off.

Tony,
Thanks, so yes, I will still need the circuit I am working on, to drop the power to the unit, so I could then read the accumulated amp hours.
Naturally, my ultimate goal is to determine the number of amp hours consumed until the battery drops out of the "functional" range. Generally considered to be 10 or 10.5 volts (depending on who you are talking to). I had developed my system to draw 6 amps at 12 volts. I recognize that the amperage varies as the voltage drops, but I figured I could check the amperage when the battery gets to about 11 volts, and use that as an average. It ought to be pretty close. But if I spring for the analyzer and my circuit cut off system works properly, it should be dead on.

- - - Updated - - -

One last thought.
When the voltage falls to the point where the load disconnects, the battery will then be unloaded, and the voltage might start to rise slightly. You don't really want the load to switch back in, so it may need to have a system that ensures that cannot happen. Once it turns off, it stays off.[/QUOTE]

Yes, that is the beauty of my original circuit. You have to energize it manually. Now all I have to do is get my machine to cut off at the right voltage; either by trial and error, or someone on this forum is smart enough to calculate it for me.

Another idea that might spark your interest.
Being lazy, I purchased a programmable voltmeter, which has proven to be a handy little gadget.

**broken link removed**

This is basically a four digit voltmeter, but it has two independent 1 amp rated changeover relays. The exact pull in and drop out voltages can be set digitally in 10mV steps (on the 0-50v range). Four digit voltage readout would be 12.34v for example.

I am using one of these to keep a standby flooded lead acid engine starting battery fully charged.
It trickle charges the battery up to 14.20 volts and then disconnects the charger. Once the battery falls to 12.60 volts through self discharge, it reconnects the charger.
That is much more healthy for the battery than constant floating it at 13.8v.

A similar system might make a very nice automatic battery discharge tester that could connect and disconnect a load for capacity testing.

One relay could be programmed for discharge testing, the other relay for battery charging.
So you can use the same voltmeter for both functions without having to fiddle around reprogramming it every time, which is a bit of a chore.

Warpspeed said:

Another idea that might spark your interest.
Being lazy, I purchased a programmable voltmeter, which has proven to be a handy little gadget.

**broken link removed**

This is basically a four digit voltmeter, but it has two independent 1 amp rated changeover relays. The exact pull in and drop out voltages can be set digitally in 10mV steps (on the 0-50v range). Four digit voltage readout would be 12.34v for example.

I am using one of these to keep a standby flooded lead acid engine starting battery fully charged.
It trickle charges the battery up to 14.20 volts and then disconnects the charger. Once the battery falls to 12.60 volts through self discharge, it reconnects the charger.
That is much more healthy for the battery than constant floating it at 13.8v.

A similar system might make a very nice automatic battery discharge tester that could connect and disconnect a load for capacity testing.

One relay could be programmed for discharge testing, the other relay for battery charging.
So you can use the same voltmeter for both functions without having to fiddle around reprogramming it every time, which is a bit of a chore.

Click to expand...

Tony,
Unless I am misunderstanding, this device will do exactly what I am trying to build. I could program it to kick in at 10.25 volts to operate another two pole relay, one feeding a battery powered clock and the other connecting my 6 amp load to the battery. I assume it will not restart it the voltage rises like you cautioned in your immediately prior post.
Am I reading that right?

Yes I believe it will be perfect for your purpose having used one myself.

The exact relay pull in voltage is specified, and the exact relay drop out voltage is specified.
You can set either quite independently. And there are two separate relays.

If you set relay drop out at 10.25v and set the pull in to anything more than slightly above that, it should work fine.

You can always push a few buttons an reprogram it any way you want.

I had thought of using a battery wall clock as an elapsed discharge timer, but you are way ahead of me there.

The main advantage of buying one of these is it will work perfectly without trying to cobble together some mess of components that may be unreliable or get out of adjustment.

Thanks Tony. I think I will get one.
Don

I have done this sort of thing for testing 12&6v batteries, there's a lot you have to take into an account. The size of your battery and what the discharge current you want to load test. If you discharge your battery higher than the 1amper hour then you have to use what's called the puekurts factor. Example if you have an 100AH rated at 20hr rate so 100/20=5 so it should be able to supply 5amps for 20 hours, lets now say we want to discharge at 10amps you think it should be able to supply 10amps for 10 hours but your thinking wrong it will only last for 7.49 hours to 100% discharged or 6.35 hours for 80% discharge rated (Manufacturer's 80% Discharge Time) providing 79.49AMP HOURS in total. I used a micro controller to the battery voltage and current with a timer counting the hours. Example if we had 100AH discharging at 10amps once the test had started it would monitor the voltage and time. If the battery reached 10.5v before 6.35hrs the test would be aborted and battery disconnected. If it went past the 6.35hrs the test would also be stopped had its past its test. The pass mark was set to 60-70% and deemed to be serviceable. If you want to do it without a micro it would be better to use a Comparator circuit to use for a cutoff.

Yes you are quite right.

The industry standard for rating larger batteries is for a ten hour discharge rate, but there are many exceptions to that.
Fork lift batteries are often rated for discharge over an eight hour working shift for example.

He is using an accurate voltage comparator for cut off, and a microprocessor instrument to accumulate amp hours and watt hour readings, so it should all be very accurate and very repeatable.

Its definitely an interesting and very worthwhile exercise.

wizpic,
Thanks. You are obviously much more well informed on this subject that I am. I was pretty amazed when I learned that the battery manufacturer dictates a one hour time to test its 24 ah battery, with a "pass rate" of 85%. I did not plan on being absolutely accurate with the testing, but wanted to get fairly close. The testing machine the manufacturer recommends costs about $1200. Could you give me an estimate of how far off I would be if I discharge it at an average 6 amp rate.

Warpspeed said:

You might want to look at one of these Turnigy power analysers.
It monitors volts, amps, watts, and records watt hours and amp hours with very fine resolution.
Perfect for monitoring and recording the results of a battery discharge test.

[.

Click to expand...

Amazing!
It is far cheaper than purchasing the raw components.

schmitt trigger said:

Amazing!
It is far cheaper than purchasing the raw components.

Click to expand...

Its just amazing what you can buy these days, and the low cost of many things is very discouraging towards developing small do it yourself projects.

Warpspeed said:

Its just amazing what you can buy these days, and the low cost of many things is very discouraging towards developing small do it yourself projects.

Click to expand...

But by doing this it takes all the fun out of the hobby and there for makes it less interesting and pointless, well that my thoughts.

Battery manafactures usally give there ratings in either 20 or 5hr rating which is used in most standards the 5hr raiting gives less capacity compared to the 20hr rate. Forklifts use the 5hr rated amphour using slightly higher amphour battlers to make sure the forklift lasts an 8hr shift.
In my design I used mosfets and Pwm to give a constant 30amps 16bit adc to measure the voltage and current and the micro controller took care of the rest, the unit tested 4 batteries at once individually.
Yes this is a very interesting project
Highflyer what is the battery capacity ?
Give me some more details of the battery your testing, manufactures don't really recommend discharging batteries to 100% to often it shortens the battery life 80% is better. To carry out this test the batteries must be fully charged first otherwise it would result in the battery failing.
I can offer to help in creating a smaller tester based on my design for a singe battery because this subject is very interesting and just for the fun of it

The batteries I'm dealing with are aircraft batteries, 12 v, 24 ah, recombinant gas, made by Concorde. The FAA regulations require that batteries be capacity tested during an aircraft's annual inspection. I am interested in checking my battery between those annual inspections, to make sure it is in good shape. Just FYI, I keep a "Battery Minder" attached full time. (Approved by the manufacturer of the battery). This works fine, but it will "mask" a deficient battery, and you only find out when you fly out of town for a few days.
The FAA and Concorde state that the battery capacity test should discharge at a rate equal to the amp hour rating (ie. 24 ah battery discharged at 24 amps), to a voltage of 10 v, and if it doesn't make 80% of the rated time, it is considered non-airworthy.
I don't mind spending some time building the test mechanism, but I'm doing it partially because it is fun, and I don't want to invest a lot of money. I am not in it commercially.

To me discharging at 24amps will not be very good or healthy for the battery. The test should take .56 hrs to reach it's 80% discharge state. The problem with batteries is you can test today and pass and next day it could fail there is no guarantee in any method you test. We tested a £2000 unit to measure the internal resistance of the batteries(which they do not recommend) for ups batreries. The idea is you test the batteries each week individual cells and build a picture up once that one cell changes in a higher resistance this is deemed to give ad early indication that cell may fail at some point.

I will see if I can find a data sheet on them I've never heard of batteries using a 1amperhour rate I.e 24AH @ 24amps

There are all kinds of funny ways to rate batteries. Automotive batteries are sometimes rated in CCA (cold cranking amps) which amazingly is a thirty second discharge rating.
**broken link removed**

There is no reason why a special purpose battery could not be rated for a one hour discharge. Lithium iron phosphate batteries LiFePo4 can be safely either charged or discharged at a one hour rate in electric vehicles, so its not unheard of.

Warpspeed said:

There are all kinds of funny ways to rate batteries. Automotive batteries are sometimes rated in CCA (cold cranking amps) which amazingly is a thirty second discharge rating.
**broken link removed**

There is no reason why a special purpose battery could not be rated for a one hour discharge. Lithium iron phosphate batteries LiFePo4 can be safely either charged or discharged at a one hour rate in electric vehicles, so its not unheard of.

Click to expand...

Yeah auto are rated in CCA but they still have an AH rating and these are not designed for supplying current continuous like deep cycle. I understand that there are some batteries like lifepo4 etc these are designed for high discharge currents unlike AGM batteries which the Op is talking about. To discharge at 24 amps comes with a lot of heat to get rid off.
If the op was to discharge at 24amps he will have to take the temperature of the battery into consideration looking at the data sheet for those batteries they state .5 to 120hrs discharge rates not fully gone through it though