I have considered using all NiMH batteries in my current robotics project. But, I came up with an idea. Why not use the type of battery that is best suited to the type of load? In my mind, this would make better sense as I would use the type of battery that is best suited for the specific type of load. In my design, the electrical systems are galvametrically isolated from each other using relays and opto-isolators. In other words, the main drive motors are powered by their own batteries, the servos may or may not be powered from their own batteries, and use the main drive motor batteries if not. The electronics are powered by their own batteries. Now that I'm thining about this, the following questions come to mind:
What battery types are designed for intermittent, high current loads like main drive motors? Or even servo motors for that matter...
What battery types are best suited for the more or less constant discharge rate of electronics?
Or, what battery technologies are out there, and what are they best suited for?
I'm trying to start a discussion here, so comments, suggestions, and
Assuming the batteries are isolated, there's no real problem with this; however, it DOES complicate your charging scheme, since different battery chemistries will likely require different chargers. This alone can make your proposition a non-starter.
Batteries with low internal resistances. These include nicads, nimh and slas. There may be others.
See above -- alkalines are well-suited to electronics, have great power density (for batteries) and have long shelf life. Unfortunately, recharge capacity is highly limited, even for "rechargable" alkalines. The batteries mentioned in 1) above will also work fine for electronics.
Commonly used in robotics are NiCads, NiMh and sealed lead acid (sla) (for higher current requirements). These batteries are well-suited for both drives and electronics, and good chargers are easy to come by or build.
Other technologies include lithium ion -- reasonably low internal resistance, great energy density, but expensive, limited availability and difficult to charge properly, rechargable alkaline, even solar cells and large capacitors.
This topic gets covered a lot in most of the standard hobbyist books. "Robot Builder's Bonanza" (Gordon McComb) and "Mobile Robots" (Flynn et. al.) are both good starting points.
At about the time of 11/16/2004 2:19 PM, The Artist Formerly Known as Kap'n Salty stated the following:
I've already considered the charging problem. The charger for each type of battery will be onboard the robot. The chargers are designed to use a filtered, unregulated DC voltage of about 30V or so at a max current of 10A in a common bus arangement. All the chargers tap off that supply which is externally connected. Only during charging will the grounds be more or less directly connected. I plan on using relays to accomplish this. When external power is applied, the MCU is signalled and it will not allow anything to run except the most critical components, like the MCU. All sensors, all motors, and their control circuits will be powered off using relays.
So I guess a small wet-cell deep cycle lead battery is out of the question then? ;-)
I'm not going for the alkaline batteries. They are great for power density, but are for most intents and pourposes not rechargable. They are cheaper, but after replacing a few sets of those, the cost of battery replacement far exceeds the cost of the rechargable types. Not to mention the additional landfill material because they are a throw away item. So even in all my portable equipment like my pager, discman, the TV remote, etc all use rechargables. I've been using the NiMH in these things quite reliably myself.
Because the charger(s) are built into the robot, I'm going to have to build them.
I'm staying away from anything that has the work 'alkaline' in it for the reasons that I mentioned above. Now I have though of Li-ion, but I don't know much about it. The cell phone uses Li-poly which is another one that I don't know much about. I have thought about using strings of large capacitors, something on the order of ten 47F at 2.5V and using a switching regulator to convert the voltage, but that seems to be more trouble than it's worth...But the recharge is on the order of a few minutes....Which reminds me....I read somewhere someone recently invented a battery technology where the battery can be recharged in something like 30 seconds or so...Li-Si I think it was... Now that would be something.
I have both editions of Gordon's book, and I think I have the other book as well (extensive library). Time to do some more googling...
Li-ion and Li-polymer batteries are not hard to charge. If you find the cell end voltage, usually 4.2 volts for more recent batteries and 4.1 for older batteries. If you have a lab/benchtop powersupply that you can set the a current and voltage limit it is easy. Just set the voltage to 4.2 (check this with a good multimeter at the end of the leads you use to charge, there can be some small loses in leads/connectors) Then set the current limit to something sensible like
0.1 C. I have been charging batteries for years for my mountain bike lights at .33 C with no problems. The batteries are charged when the current meter is showing only a small current flowing, 10s of milliamps. I used to have an exact figure for this but I can't remember it it was in an elector magazine.
Extra tech details: The voltage end point needs to be set very accuratley. For every .1 V under you will lose ~ 7% of capacity. MEGA DISCLAIMER *************************************** OVER CHARGING SETTING THE END VOLTAGE TOO HIGH CAN CAUSE THE CELLS TO CATCH ON FIRE. I have never had this prob and I have left batteries on the charger(PSU) for nearly a week. The current just drops to near 0.