What are considerations for battery management?
T Chap
Posts: 4,223
I have a product that uses a single 23 stepper driven by several controls and sensors. Currently, the system is ran by two series 12v 7ah gel cell batteries(24v) that are on a floating smart charger. The batteries are not switched, they are always in the circuit. The charger displays slow and fast modes, and is always in slow unless the motors are moving, then it may kick into red(fast) for the short time the motors are on. I have run three prototypes like this for over 6 months with no problems, as the motors and controls are constantly consuming power for holding torque, so it has balanced out about right. For commercial release I need to tell the micro to shut off the motors after a few seconds of use to conserve energy, and if needed, the motors can turn back on. If an attempt to move the motor happens by some chance, the encoders will read it at once and the micro will correct the movement back to where it was, so the effect wil be that ther was holding torque all a long.
Ok, so now that there will be a lot less energy wasted, I want to manage the batteries more seriously for a commercial version. I am looking for where to start. My thoughts are to get the temp from the batteries by a sensor, get the voltage, and regulate the charging to avoid damage and prolong the batteries life. Ideally, a single IC that would easily interface to the Prop would be nice, 1wire or serial.
Here s an idea, it is a TI part, but it doesn't read temp. It suggests it can accurately monitor ambient temp, not quite sure how that happens without a sensor.
"A typical standby supply current requirement of only
1.6mA allows these ICs to predictably monitor ambient temperatures."
Maybe I will add a visual/auditory alarm is such a case of battery failure or overheating.
rocky.digikey.com/WebLib/Texas%20Instruments/Web%20data/UC2906,3906.pdf
Any suggestions would be appreciated. Thanks
Post Edited (originator) : 12/3/2006 9:27:29 AM GMT
Ok, so now that there will be a lot less energy wasted, I want to manage the batteries more seriously for a commercial version. I am looking for where to start. My thoughts are to get the temp from the batteries by a sensor, get the voltage, and regulate the charging to avoid damage and prolong the batteries life. Ideally, a single IC that would easily interface to the Prop would be nice, 1wire or serial.
Here s an idea, it is a TI part, but it doesn't read temp. It suggests it can accurately monitor ambient temp, not quite sure how that happens without a sensor.
"A typical standby supply current requirement of only
1.6mA allows these ICs to predictably monitor ambient temperatures."
Maybe I will add a visual/auditory alarm is such a case of battery failure or overheating.
rocky.digikey.com/WebLib/Texas%20Instruments/Web%20data/UC2906,3906.pdf
Any suggestions would be appreciated. Thanks
Post Edited (originator) : 12/3/2006 9:27:29 AM GMT
Comments
Some time ago, Texas Instrument (TI) bought out the Benchmarq Company. Benchmarq had a large series of battery monitors, battery fuel gauges, and battery charger chips. TI has continued to expand the line and it now covers all battery chemistries, as well as single and multiple cells. Unfortunately, SLA (sealed lead acid) is considered a somewhat obsolete technology, and is not represented as well in their line as other battery chemistries. Regardless, it's still worth a look, as I think they have at least 5-6 SLA chips to offer.
Regards,
Bruce Bates
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<!--StartFragment -->
To do so, creates wear and tear on their physical makeup.
Lead Gel cells seem to be the best, least tempermental for trickle charges though they are not the most compact or energy dense.
I really don't agree that seal lead acid is obsolete technology. It is just been given a bad name by people that are promoting 'high tech advances'. Most, if not all of the newer technologies are more tempermental and have a shorter battery life at huge increases in per unit cost. In some cases, the energy density of new battery technology combined with the high recharge rate is exactly what makes the battery explosive.
In other words, if you put enough chemical energy in a smaller and smaller package at a higher and higher rate --- BOOM!
The beauty of the gel cell is that is doesn't leak and yet provides a very tough workhorse.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"If you want more fiber, eat the package.· Not enough?· Eat the manual."········
I have been looking at the temperature issue and you are right - most [noparse][[/noparse]maybe all] the smart chargers do not have a temperature sensor. Parallax has sold a very useful and in this case optimal temperature chip from Dallas Semiconductor. It is a programible thermostat that can trigger a shut down at high temp [noparse][[/noparse]say 45 degress centigrade] and a turn on at low temp [noparse][[/noparse]say 35 degrees centigrade]. That would keep you well below the boiling point of water and probably do significant good towards both conserving energy and useful battery life.
I really don't think you need to change your charger -- just add the thermostat to the circuit and attach to the side of the battery. If the whole device is not mobile, adding more gel cells may allow you to keep the charger off for longer periods of time. It seems to me that the charger is really sourcing power [noparse][[/noparse]maybe partially] during peak use. That may or may not be more economical in the whole scheme of things. Hard to say.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"If you want more fiber, eat the package.· Not enough?· Eat the manual."········
Are you refering to the 1620? I don't see anything else that looks like what you mentioned. I do already have an LM34 on the board which is contained in the battery box. I set it to around 110F to trip a comparator and do some other functions, maybe just have that same circuit turn off a relay to the charger. It gets trick though, as on power up how does the charger get juice to the batteries if the batteries happend to be dead and the board is off? If the processor kicks on the relay to the charger if the LM34 is below the threshold, maybe there has to be a physical switch to bypass the relay in order to charge the batteries initially. Seems like a messy solution but fire preventionis more important.
Post Edited (originator) : 12/4/2006 8:21:56 AM GMT
Good question about where you get the power if you have to deal with a dead battery.
I suppose you need a by-pass switch.
Don't forget, it has to touch the battery. So I don't think you can easily integrate it into a board.
I suppose there are other 'applicance store' type thermostats that can easily do the job and toggle 120VAC.
No relay, no 5VDC requirement either.
I suspect this is an area that 'smart electronics' just gets stupid.
A generic 120 VAC thermostatic device and a simple trickle charger really appeal to me.
I keep looking at my DS1620, but it takes a day to program it after you read the docs and write a routine for a BS2 to do so. The main appeal is that I can re-adjust the temperature triggers to whatever I find optimal.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"If you want more fiber, eat the package.· Not enough?· Eat the manual."········
I think an over temp condition would be extremely rare, but safety measures have to be redundant.
The processor can shut the SSR off too in case of over temp, triggering an alarm.
I would also do over and under voltage if ether case happen for any lenght of time
"="Shut Down
You do have some very good points
Let me know what you come up with
Sam