I am trying to construct a capacity tester for 12 volt batteries. The theory is that a 24 amp hour battery should be able to provide 24 amps before dropping to a discharged state of 10 to 10.5 volts. Or in my plan, provide 6 amps for four hours. I anticipate hooking up the battery to be tested, as per the attached circuit, set the clock to 12:00 and close the start switch momentarily. This should energize the relay and the battery operated clock. When the testing battery drops to the 10-10.5 volt range, the zener diode should switch off, and the clock should read the elapsed time. My problem is that I do not know the values for the resistors and possibly the zener diode. Intuitively, the zener should to be 10 to 10.5 v, but I don't know this.
Incidentally, this device is to used for aircraft batteries, and the Federal Aviation Regulations decree that when a battery cannot provide 85% of its rated capacity, it must be replaced. Thus, other methods or types of


battery testers are not in consideration.

 

Apparently my attachment was cut off and does not show that the bulb shown on the lower right is a 6 amp load.
All help appreciated.
Don

 

The actual capacity of a battery is generally very dependent on the rate of discharge -- it is not nearly as simple as the nominal amp-hour rating would have you believe. Generally the lower the current draw the higher the observed capacity will be.

Since this is for aircraft batteries, I would be extremely hesitant to do anything along the lines of what you are trying to do. If your homebrew tester says that a battery is good to continue using and that aircraft is subsequently involved in an incident or accident, do not be surprised to be named in the resulting lawsuits regardless of whether the condition of the battery had anything to do with the event -- the lawyers will throw together some legal mumbo-jumbo that will paint a murky chain of cause-effect that is intended to sway some idiot-filled jury in the personal injury case that the poor widow deserves compensation and that you should be held liable because you were too cheap to use a "real" tester.

 

I agree with WBahn.
When it comes to critical aircraft equipment you don't want to mess with testing using an unauthorized piece of test gear, unless you don't mind being sued for everything you have.

 

WBahn & Zapper,
Thanks for the replies. I appreciate both the technical information (about the discharge rate) and the legal ramifications. The guys who do the annual inspections are supposed to do the capacity testing each year. I wanted a way to check during the year to see if I need to worry about the state of the battery before that time.
Are any of you guys who are smarter than me (that means everyone on the site) able to tell me what those resistor/zener values need to be. I would ask how to calculate them, but I suspect even if I was told the formulas, I would probably still not be able to figure it out.
Don

 

First check the current specified for the rated battery capacity. It's typically 1/10 the Ah rating or 2.4A for a 24Ah battery.

Below is the LTspice simulation of a circuit to do what you want.
I took the liberty of modifying your circuit to improve its operation.
It uses an inexpensive TL431 adjustable, shunt reference (configured here as a comparator with adjustable trip point) to give a more precise and adjustable voltage cutoff for the relay. (Zener and other circuit tolerances might otherwise make it difficult to meet your 10 to 10.5V desired trip point).
Since that device works best with a ground reference I changed the transistor to a PNP high-side driver to power the grounded relay coil.
I also added a diode across the relay coil to suppress any inductive transients which could zap the transistor.
(I didn't show the relay contacts or connections but you can add those).
As you can see the nominal cutoff is 10.25V (relay current goes to zero) with the pot wiper set very near to 50% (midpoint). (The voltage divider action of R5 and U3 reduce the battery voltage to the 2.5V trip voltage at the Tl431's input).

 

Wow! Let me mull over that. This is awfully far above my expertise level, but if you don't mind me bugging you for additional information, I would like to give it a try.
Thanks
Don

 

but if you don't mind me bugging you for additional information, I would like to give it a try.

Glad to help.
Just post any questions you have.
If you look at the data sheet for the TL431 it might help you understand its operation some.
Basically the 431 conducts from cathode to anode when the control input is above 2.5V and stops conducting when the control input drops below 2.5v.
Here they are for as low as US$0.15 ea.

 

You have no idea how electronically inept a person you are talking to. But here goes;
How fussy are the resistor values. I haven't shopped yet, but wonder if I'm going to find 15.5 k ohm resistors, for instance. (My primary source for such stuff is Fry's)
Same question about the 1N4148 diode.
Is the relay going to be a problem. How do I make sure it matches the specifications R=80, 10mH. Or are most of my 12v relays going to work.
Same question for the Rtot=10K wiper=.5 potentiometer. How will I figure out the necessary specification. Will I need to adjust it by testing to match the battery voltage at cut-off.
You mention that 2.5v cut-off for the TL431a. The data sheet says it is adjustable, but the picture depiction does not indicate that to me. Is it preset for the 2.5v.
I have yet to check on the current specified for the battery capacity, but I'm assuming you mean that if it is 2.4 A, I need to get pretty close to a 2.4 A load to properly do the test. I'm also assuming that the expensive testing devices hold a constant amp draw. But my thinking is that between the fully charged 12v battery and the cut-off at 10.25 v, a 2.4 amp load should not vary too much to keep me from getting a reasonably accurate result.
Again, thanks.
Don

 

A 15k will work for R5 because that's 400 ohms off spec, and the adjuster has 10,000 ohms. You can easily dial out the difference.
R3 and R4 could be 470 or 520 ohms and work just as well.
The millihenries are just a feature of the simulator. They don't matter in real life.
The 80 ohms are a good estimate of any small 12V relay that will connect a load of a few amps.
The 10K pot is the adjustment for the TL431. You just turn the knob to the necessary specification, which seems to be 10V to 10.5V
The 1N4148 is just a cheap diode. You could use 1N4001 to 1N4007 just as well.

Did I miss anything?

 

wonder if I'm going to find 15.5 k ohm resistors, for instance. (My primary source for such stuff is Fry's)
Same question about the 1N4148 diode.

The 15.5kΩ is a common 1% value.
The 1N4148 is a common diode, but just about any diode will work.
You can readily buy them at any online electronics supplier such as Digikey, Jameco, Mouser, etc.

Is the relay going to be a problem. How do I make sure it matches the specifications R=80, 10mH. Or are most of my 12v relays going to work.

The values I used were just typical for testing.
Any 12Vdc relay should work.

Same question for the Rtot=10K wiper=.5 potentiometer. How will I figure out the necessary specification. Will I need to adjust it by testing to match the battery voltage at cut-off.

You just buy a 10kΩ pot with a configuration that you like.
It can be a small trimpot or a larger panel-mount pot, your choice.

Edit: If you have a variable power supply and a meter you can adjust it to the desired cutoff voltage.
If not, I can tell you how to adjust it for any given battery voltage.

You mention that 2.5v cut-off for the TL431a. The data sheet says it is adjustable, but the picture depiction does not indicate that to me. Is it preset for the 2.5v.

It does have an internal fixed reference voltage of 2.5V.
The adjustment comes from the external resistors connected to the input terminal.

I have yet to check on the current specified for the battery capacity, but I'm assuming you mean that if it is 2.4 A, I need to get pretty close to a 2.4 A load to properly do the test. I'm also assuming that the expensive testing devices hold a constant amp draw. But my thinking is that between the fully charged 12v battery and the cut-off at 10.25 v, a 2.4 amp load should not vary too much to keep me from getting a reasonably accurate result.

You should check the battery specification to see at what current the capacity is tested but I think it's typically a 10 hour rate (1/10th its capacity).
If you can find a bulb that gives near that current you want it will help keep the current constant as the voltage drops, due to the large positive resistance versus voltage of an incandescent bulb.

 

Thanks guys. I have to go shopping. I will probably be back.
Don

 

Crutschow,
I built the circuit as per your schematic (I think).
I don't have sophisticated equipment, but when I apply 12v to the circuit, it doesn't pull in the relay. (12v applied directly to the relay operates it.)
Adjusting the potentiometer all the way from one extreme to the other makes no difference. (The pot I have goes from virtually 0 to 7.2K.)
The closest I was able to get was 460 ohms for the resistors which should be 500.
The closest I was able to get was 15K ohms for the resistor which should be 15.5K (Once in the circuit it only reads 3.9K)
I have attached the schematic depictions of the TL431a and 2N2907.
The emitter on the 2N2907 is at the top of the schematic.
The cathode (K) on the TL431a is at the top.
Any idea on where the problem may be, or how to figure it out.
Thanks,
Don

 

It probably doesn't matter, but I forgot to attach the pics of the devices.

 

The component values you used should be close enough.
Most of the time when a circuit doesn't work it's a wiring error.
Double check all you connections using an ohmmeter and mark off each connection as you do on the schematic with a highlighter marker.
Make very sure that the transistors and the TL431 pinouts are correct.

Post a picture of your circuit with all the connections.

Measure the voltage at the pot wiper output as you adjust the pot with the battery connected. What is the maximum voltage you measure..

 

I suspected my wiring as well, and have checked it several times, both as to component placement and solder joint continuity.
I have attached a photo of the circuit, but since it is on a board, you can't see the components and connections. I have labeled all the component locations.
I feel pretty confident of everything except the transistor and the TL431.
You can see the tab on the transistor, (on the lower left side on the board.) The on-line depictions of the 2N2907 show that the tab is adjacent to the emitter, where I have it connected to the (+) input. If that one is correct, the base and collector, are necessarily correct.
On the TL431, I have drawn a depiction of its shape (as it appears under the board). The internet data shows the "K" cathode at the lower end of my depiction, the center "A" anode (going to the ground), and the "R" on the upper end.
I'm not certain how to measure the potentiometer voltage as you ask. With 12.55 v applied to the circuit. from + to the "top" of the pot, I get 9.99v at one extreme and 12.55 at the other. Naturally, from + to the ground side, it reads the full 12.55. Across the pot, at one extreme it is 2.5v and millivolts at the other extreme.
You have no idea how much I appreciate the help.
Let me know what to check next.
Don

 

If anyone is keeping up with this thread, the problem apparently was that I installed the PNP transistor backwards, due to my ignorance of how the emitter, collector and base are depicted in the original schematic.
I do not have a variable voltage power source, so I will have to test the device by discharging a battery a few times.
Again thanks to the guys who jumped in to help me.

 

To bring anyone up to date, without having to go back and read the entire thread; I'm trying to build a circuit to test capacity on a 12 battery, by applying a load and measuring the time to discharge to 10 to 10.5 volts. Crutschow suggested the attached circuit. I built it, but with the potentiometer at one extreme (1 ohm), the unit allowed the battery to drop to about 7.5 volts (at which time, I jiggled the leads and it shut down. At the other extreme (about 5K v), the battery dropped to 3.7 v when it shut off (again on jiggling the lead).
Any ideas.

 

Did you replace the transistor when you put it in backwards?

The relay drop out voltage should always dominate. This http://www.te.com/commerce/Document...=V23076-X0000-A001&DocType=DS&DocLang=English relay has a 1.2 drop-out and a 6.9 V pull-in.

So, the battery has to be greater than about 6.9+0.6V for the relay to pull-in once.

I uploaded your doc file in .pdf format.

 

KISS,
Yes, I replaced the transistor. (And installed it with the collector at the top of the schematic, ie. on the + side) This looks wrong to me, but the other way, it did not pull in the relay at all.
If I'm understanding what you are saying, I have to make sure I have in excess of approx. 7 volts to expect it to pull in. In between tests, I am recharging the test battery, so it always has in excess of 12 volts to start.
Thanks for the post, and any other thoughts.
Don