Motors/Battery life

[Manoel]

Originally Posted by eugenebrooks

Al, I beg to differ...

We compared the capacity of the then new EX battery to year old ES batteries last year, and found the year old batteries to have 50% more capacity at a constant current 20 amp load. This year's milage may vary, but the batteries were quite different in performance when checked last year.

Eugene

Well, FIRST disagrees with you. Use that to your advantage

[KenWittlief]

there are definate variations in capacity from battery to battery from the same year. Every once in a while we got one that was so poor at holding a charge we wrote "BAD" on both sides with a sharpie. They would only last about 3 or 4 minutes, while a good battery would last 12-15m, and they were new!

[Al Skierkiewicz]

Originally Posted by eugenebrooks

Al, I beg to differ...

We compared the capacity of the then new EX battery to year old ES batteries last year, and found the year old batteries to have 50% more capacity at a constant current 20 amp load. This year's milage may vary, but the batteries were quite different in performance when checked last year.

Eugene

Gene,
I will test our batteries, I have a CBA battery analyzer from Mountain Radio. I know that there was a manufacturing defect in some batteries last year. When discharged, at some point in time the output voltage suddenly dropped by 2.1 volts. Then later there was another sudden drop. Indication was that cells were shorting or of vastly reduced capacity. I will post when I know more. There should be no reason for a 50% difference between production runs.
Dale, I have to take issue with your methodology. The load testers are for short term testing. When you load to capacity over a long period of time the resistance changes with temperature. It is possible the load testers were changing value to a lower resistance.

[KenWittlief]

ah yes! battery temp is a hugh factor for lead-acid/jel cells.

They can loose as much as 50% capacity from 0F to 150F (across their temp curve)

usually the warmer the battery, the more capacity. Im not sure if it drops off at higher temps?

[eugenebrooks]

Originally Posted by Joe Johnson

There is another possibility. When we did some tests a few years ago, we found significant variation from battery to battery even among those that are nominally the same model number.

While I don't think that it is likely to be the factor that moves a robot from good to great, using the "good" batteries over the "medium" ones can't hurt.

Joe J.

Last year when we did these tests, we checked two S ones from the prior year and two X ones from the current year. The two S ones were very close to each other in capacity, and the two X ones were very close to each other (within a few percent). One can't draw conclusions from a sample of two, but the moral of the story is that you need to test the capacity of your batteries and qualify them for competition use. A good battery does not make good robot great, but a bad battery will definitely make a great robot marginal.

[eugenebrooks]

Originally Posted by Al Skierkiewicz

Gene,
I will test our batteries, I have a CBA battery analyzer from Mountain Radio. I know that there was a manufacturing defect in some batteries last year. When discharged, at some point in time the output voltage suddenly dropped by 1.2 volts. Then later there was another sudden drop. Indication was that cells were shorting or of vastly reduced capacity. I will post when I know more. There should be no reason for a 50% difference between production runs.
Dale, I have to take issue with your methodology. The load testers are for short term testing. When you load to capacity over a long period of time the resistance changes with temperature. It is possible the load testers were changing value to a lower resistance.

We were not seeing sudden drops when we did our testing and our load was a constant current. We will pull the tester out of the attic in the next week or so and check all of our batteries for this year. It will be interesting to compare notes...

[Al Skierkiewicz]

Well we checked one of our 2006 batteries and compared it to graphs from previous batteries. This one battery is very similar to the previous years batteries. I am still working on getting test data locked down between a very old battery and the fresh ones this year.
In most cases a battery appears to produce better when warm but there is a diminishing return. The chemical reaction progresses better at higher temperature but the internal resistance changes with temperature as well. It is for this reason that some folks turn their headlights on for 20-30 seconds before trying to start the car on a cold day. For our use, batteries at room temp are just fine, they will heat up during competition but will cool off in a short period of time. The temperature changes also affect charging and the way the charger senses the state of charge on the battery. For this reason I don't recommend putting a battery on charge immediately following a match. Give it 10 or 15 minutes and the charger will more accurately assess the condition.

[Ianworld]

I know one reason some engineers might think that the batteries weren't capable of supporting the current draw of a first robot. A battery of the form factor that FIRST uses cannot put out the amperage that the FIRST robots use. While robots have a 120 amp breaker they will often draw in excess of(i'm guessing here based on how often my team's robot stalls out) 200 to 300 amps for short periods of time. I checked the battery spec sheets a while ago and they were labeled by exide as being a special type of "Deep Discharge" battery. That basically meant that they were able to under heavy loads be able to output more amps than a regularly configured battery. I can't rememeber what the written max load was, I think it was around 240 amps, but I am not sure. The batteries as other people have stated have quite a bit varience between them so I'm sure that some of the batteries can put out 300 amps easily with less resistance(and thus for longer periods of time) than other batteries. Anyhow my 2 cents.

[KenWittlief]

Originally Posted by Ianworld

I checked the battery spec sheets a while ago and they were labeled by exide as being a special type of "Deep Discharge" battery...

Now I have this image stuck in my head, of a white haired guru teaching the batteries at Exide

"when you are sent out into the world there may come a day when you are called upon to give all that you possibly can. Do not draw your power from up here, by your terminals. You must reach down deep, all the way to the foundation of your plates, and pull the energy from your entire being!

You must prove yourself worthy. He who offers up his deep energy freely shall be recharged 100 fold, but he who keeps his energy to himself shall be cast aside

torn asunder, and recycled, according to the law!"

(sorry! lost my mind just for a moment :^)

[eugenebrooks]

A tale of 7 batteries:

I finally got around to firing up the load tester on our batteries.
I list the results below:

Constant Load Current: 20.3 amps (checked with fluke current clamp)
Cutoff Votage: 11 volts

All of the batteries were charged with a
three stage battery charger just prior to testing.

The test is to record the time in "EDU main loop ticks"
that the battery takes to drop below 11 volts while under the
constant current load of 20.3 amps.

"Bat Num" Model Year "EDU Main loop ticks"
Battery 1 EX18-12 2006 77559
Battery 2 EX18-12 2005 62519
Battery 3 EX18-12 2005 35849
Battery 4 ES18-12 2004 79576
Battery 5 ES18-12 2005 82548
Battery 6 EX18-12 2006 95907
Battery 7 ES18-12 2004 109930

Battery 3 is one of the kit batteries from last year, as is battery 2.
The low performance of Battery 3 was double checked and a
second test produced 34377. This reflects the repeatability of
the measurement. It is unlikely that we would want to trust
Battery 3 for more than a single match before charging.

Battery 4 spent some time powering an electric fence
during the off season, which I think did it some harm.
It matched the capaciity of battery 7 when it was checked last year.

I have been digging for the time of an EDU main loop
clock tick, I remember it to be 0.0172 seconds, but
I have not found it in the legacy documentation yet.

2004 was, and still is, a very good year for batteries...

[Al Skierkiewicz]

Gene,
We are running tests on all of our batteries (some going back to 2002). Using the CBA II at 7.5 amps, we test down to 8 volts (the cutout for the RC). The computer then allows us to overlay the results and I will be printing those up shortly. It takes a few hours to run the test so we don't want to leave a battery on test when no one is present. Interestingly enough, we have another battery that has diminished cell capacity on one cell that obviously reduces by 2 volts during the test. The CBA also calculates amp hour ratings from the test.

[Gdeaver]

Unfortunately, many teams do not have constant current robots. The batteries are subjected to huge inrush currents constantly and a occasional 30 sec grudge push match. A more realistic test would be to cycle the batteries at 20 amp with a periodic surge. In 2003 our bot ruined the batteries from constant high loads. I use some gel cells in some portable lifting equipment that are designed to supply short term surges as opposed to constant load. They are of the wheel chair form factor. This battery design would be more appropriate for first uses but they are much more expensive than the battery design we use.

[Al Skierkiewicz]

G,
We are more interested in veryifying that the batteries we own match a published curve and can be compared to other batteries. We are not trying to predict run time, just looking for batteries that don't meet spec or are past their prime. What we have determined thus far, is most of the batteries fit the curve and that older batteries are showing reduced capacity.

[eugenebrooks]

The battery tester we built, controlled by an EDU RC controller, can test with varying loads. One can program any desired load profile from 0 to 63 amps, in steps of 1 amp, and simulate a match if you want (with current limited at 63 amps). One could upgrade to support higher currents if one wanted. We have only used the load tester for a fixed load, however, to certify the capacity of our batteries against the published curves. It pays to identify any batteries that are not measuring up and relegate them to practice use, or single match use.

[Al Skierkiewicz]

Well,
Here as promised is a few of the battery test runs. The CBA II can draw down 7.5 amps and gives a visual display of amp hour capacity. It at least will allow us to compare to published curves for the batteries. Please note that one of the old batteries is slumping quite a bit and will suffice for while longer in practice at home. Please also note that one of the newest and unused batteries exhibits a reduced cell (drop of 2 volts) but in general the curve still fits the other batteries. It is interesting to view the slightly reduced capacity of the older batteries which may have gone through several hundred charge discharge cycles in their lifetime.
Although this method does not represent real world current draw on your robot, it is representative of the health of your battery compared to new. I believe these tests can be further modified by using a shunt (a parallel load resistance) to accelerate the current draw. However, the tested curves do match the published curves fairly well, at 7.5 amps the amp hour rating is reduced to at least 12 AH and the batteries are testing at almost 16 AH when new.