Battery life?

[Al Skierkiewicz]

Kari,
Batteries have a finite life. The batteries we use are speced at 400 charge/discharge cycles under normal use. Our use is pretty heavy with long periods of no use in between. I estimate a team can get realistically no more than 300 if their robots use a lot of current or 350 if they are more efficient. If you are still practicing with 2007 batteries you are getting some decent life.

[fsgond]

Our team uses 5-6 batteries changed out after EVERY match. We want our robot to have top performance during every match. We usually have a lot of motors and an air compressor that runs a lot of the match, so our battery is pretty low by the end of the match.

Also we buy all new batteries every year. Batteries from the one year previous are used for practice in the shop and programming and testing the bot in the pit.

Now to the question that you originally asked, I think that you could extend the matches a bit, but no more than 1 or 2 minutes.

Honestly from a field coach position, the matches are very fast, and done before you know it. Also I am not sure that I could take the stress for another minute or two.

[Warlord]

When I drive the practice bot we usually go for about 10-15 minutes before I see a noticeable drop in the performance of the robot (as in, it stops driving top speed). Sometimes a little bit longer.

Past robots haven't been quite as efficient, sometimes they have only lasted like 5 minutes or so.

No matter what though we always change the battery after every match, just to be safe.


I'd say that if you are just barely making it through a 2:15 second match with a full battery, there is a serious problem. Really anyone should be able to go at least double that without absolutely having to change the battery.

[Dale(294engr]]

Here is a 50A load battery characterization on our 2006-1 SLA conducted Jan 19, 2006 (note: typo on graph)

DMM is directly connected to battery terminals (so excludes all connector & wiring resistance voltage drops)

Vterm recorded & plotted every 10 seconds until Vbatt drops to 7.0v

started at -3 minutes with .5A fan preload to illustrate 'soft' voltage area between ~13.3 - 12.6v becoming 'hard at ~12.6v at high discharge rates

'hard' useful robot current occurs at .018 ohm avg internal battery resistance
then plummets quickly

note: Battery Terminal voltage drops 1v for each 50A of load

4 CIM drives with friction belt contact to floor (or pushing opponent)
start up currents can hit 400-500A for few hundred milliseconds resulting in
an 8 to 10 v drop from the 'hard' Vterm!! (excludes connector wiring Vdrops

(dynamic current: use Oscope across .001ohm shunt to verify!! scale 1A/mV
or use 120A CB as shunt in pinch for this, its R closely controlled ~.00096ohm
also useful for rough robot DC current: DMM 200mV scale across 120A CB)

400A: 12.6v-8v = 4.6v to keep CRIO, Radio etc alive else it resets!!
500A: 12.6v-10v = 2.6v CRIO, Radio etc reset!!

Advice:
Charging: do it ASAP after every use, float charge ALL other times
abide by mfr spec max fast charge, (guaranteed to do no harm)

Matches: new battery every time
....... used batteries =Increased Rint reduces max bot fps + pushing power
Fast charge (~5A) gets you to only ~85% upon completion
..remaining 15% float mode takes >=3 hours!! use your Auto dual rate charger and leave it on. have one or two batteries on charge overnight for critical matches for speed power edge needed.

Load curve reveals much about State of Health of a Battery
..expected life, speed and motor torque loss as time passes due to Vdrop resulting from higher internal battery resistance as it gives up its energy

Note upswing of voltage after heavy load depression is Normal
.. internal I^R generates heat: chemical reaction result: increasing voltage for a short period! isn't science wonderful?

if robot consumption expected closer to 100A average run Batt test at 100A!

main challenge: find Load Resistor(s) to take heat for about 10minute for 50A .25ohm test, or 5 min for .125ohm 100A test

(swapmeets often sell 200W AL heat sink 1% WW resistors inexpensively
series parallel as required P=I^2R size for 12.6v initial voltage-1v / 50A)

Dale.. 2 attachments data, graph + imbedded calcs as Rint changes

note: my email has changed from that on the attachments
now DaleScience@aol.com

[Al Skierkiewicz]

Dale,
Nice charts. I would like to caution teams on a few things. Charging a battery while warm or hot will disrupt the charge cycle. Our batteries are better added to the charger after several minutes of cool down. While our batteries will regularly undergo several hundred amp loads, it is not wise to continuously draw high currents. As you point out, there is sufficient current drawn through the internal resistance. At 100 amps, the internal resistance is dissipating 110 watts at manufacturer's spec. This continuous high temperature can cause permanent damage. I believe that many of the early deaths we see in the AGM types are due to this mishandling. I have recommended the West Mountain Radio CBA II or CBA III battery analyzers for many years. While drawing currents in the 7-10 amp range, these analyzers will produce curves similar to yours, calculate AH ratings and allow curve overlays of the battery as tested over it's lifetime. The one volt drop in your curves is indicative of the surface charge being lost. This is one of the most confusing phenomena for teams who only measure their batteries, right off the charger, with a VOM. For those teams that wish to duplicate your method, a drop in 2 volts indicates a cell failure. Those batteries should be recycled or relegated to a 'practice only' status.

[Dale(294engr]]

clarification...

Al's comment on 2v drop refers to a rested battery after charge under NO load..

Graph of data taken show TWO quick declines:

.. the first is surface charge and exists from 13.3 down to 12.7 or so under low load dissipates in 3 minutes.. under no load can exist for hours

.. the second upon 50A load declines immediately and is due to Batt internal resistance.. Vdrop = .02*50A =1v

.. then holds relatively unchanged until the battery is exhausted and again drops off quickly (decline#3) as the internal resistance rises to & >.1 ohm

... de K3MNN originally from Pittsburgh PA --... ...--

[Al Skierkiewicz]

Dale,
The 2 volt drop is not the 1 volts surface drop in your graph. Frequently, a damaged battery or one nearing it end life will have one cell drop out causing a 2 volt drop in the load curve. You can see this in my posts with the CBA II graphs. Please forgive any confusion.

[Cory]

This thread has strayed from the OP's intent, but we buy 10 new batteries every year. We put so many hours on them throughout the season that they inevitably end up not being as good as new ones. We practice during the current year with the previous year's batteries. We never use a battery in a match more than once. We won't even use a battery that's been on the bot in the pit for system checks/testing.

You can never have too many batteries or be too careful with them. It is the most critical part on your robot.

On topic, matches will not get any longer unless more than 6 teams are on the field at once. Just won't happen from a scheduling standpoint.

[Don Wright]

We tell Gail at the beginning of the season "We need enough batteries for the season." and so she orders 110 so we never have to use a battery twice...

Just kidding...

Like Cory, we order a new set of 12 every season and use them only at competitions, never back to back (usually only once per day or twice at the most for each competition).

We use the previous year's batteries for practice and still swap after each practice match...

The 2 year old and older batteries are for the scrapbots that we use for defense practice...

As for how long...depends on a lot of things... Our 2006 robot used 10 motors I think including several that ran almost all the time (two big CIMs, two FP's), 4 little CIMs in the drive, and two bane bots plus the pump and we would almost kill the battery in 2 minutes...

[bduddy]

Originally Posted by Cory

On topic, matches will not get any longer unless more than 6 teams are on the field at once. Just won't happen from a scheduling standpoint.

I was actually thinking about 4v4 with one robot off the field in a "pit area", and something like :30 + 4:00 . From this topic, it sounds like the batteries could likely last for that long...