Lithion-Ion Batteries

[akoscielski3]

I've had this question in my head for the past couple of yearss, but still dont have a real answer. Why don't we use Lithion-Ion batteries for our FRC bots?

My team has used a Drill since 2006 that uses a Lthion-Ion battery, we love it because it doesn't stop working until the battery has appsolutely no more power in its battery.

I know that they're illegal, but why are they illegal? And won't it be better to run at 100% power all the time? You would always have the motors spinning at the same speed, there would be no more variables.

Thanks

[Jon Stratis]

I can speculate on a couple of reasons:
- Cost. Those batteries would be more expensive, and FIRST includes batteries in the KoP for every team.
- Safety. We've all heard stories about laptops or cell phones with Lithium-ion batteries catching fire or exploding, and that's the last thing we want in a robot! This is most often caused by abuse and physical stresses far less than our robots go through. It can happen during charging (while many people are next to it in the pits) or due to damage, such as being accidentally punctured by a robotic arm. While our current batteries can suffer some of the same issues, it's much more rare and generally speaking has a safer failure mode.

[artdutra04]

Besides cost and safety already mentioned, the engineering reason behind Li-Ion batteries not being suitable for FRC are the high current spikes we see.

All batteries are rated with a "C" rating that is a multiple of their rated current capacity. The "C" rating is what determines what is the highest amount of current that can be used to either charge or discharge the cell. For example, a 800 mAh 3.7v Li-Ion cell rated at max charge 1C and max discharge 2C can be charged at a rate no faster than 800 mA and discharged at no more than 1600 mA.

In general, Li-Ion batteries are never rated for high discharge capacities. Lithium Polymer batteries are better, and can have ratings in the double-digit C range, but these are not cheap. For FRC, a single Li-Po battery capable of meeting our needs would likely be at least $150.

Lithium Iron Phosphate chemistry shows potential, but is still too new and costly. LiFePO4 has lower capacities per kilogram than Li-Po, but is inherently a much safer technology. It will not explode or catch fire if mishandled or abused.

[JesseK]

If one read the safety literature on LiPo batteries and then observed the types of practices put into use at competitions from the bottom 5% of least-safe teams, it would be easy to make a substantial argument against LiPo batteries for competition use. Simply put, it's very hard to trust all of our neighbors when a small mistake will be very costly (or painful).

Plus the GDC would have to amp up the requirements on battery placement/protection on the robot itself, and we all know the last thing we want are more design constraints.

[MrForbes]

We got a new Lithium battery powered cordless drill in 2008, and it worked ok for a couple years, then quit charging. Last week I figured I'd give it a try...and it showed that it was charging. Huh. Having a mysterious battery system that sometimes works, and sometimes doesn't work, is probably not the best thing for our robots.

Lead acid batteries (like we use) are pretty forgiving, very powerful, relatively inexpensive, etc. They're not high tech and not modern, but they are the best solution for the problem of powering our robots.

[Mk.32]

So with "Lithion-Ion" batteries the OP mentioned I thought I would try to clear some things up.
The three types of Lithion batteries that are commonish are:

Lithium-ion
Lithium-ion polymer
Lithium iron phosphate

The first two are often used mixed up, LiPo [Lithium-ion polymer] are an evolved version of Lit-ion [Lithium-ion], and LiFePO4 [Lithium iron phosphate] are an even more evolved version.

As stated above Lit-ion batteries would not be able to provide the amp draw that is required on an FRC robot, also they are pretty dangerous [remember exploding laptops anyone] and expensive. And they are out dated compared to the new LiPo/LiFePO4 in both safety, price and power capacity.

Now I run a set of LiPo packs to power my quadcopter, these are small ones at 2.2ap/hr rated for up to 40C draw current and they costed about 17$ each, and can be charged at 8C [in 7.5mins]. Personally I love them.

However for FRC usage, LiPo packs big enough [17ap/hr] would cost easily $150+ and chargers would cost like $120 each [ones that can change them fast...] as well as the PSU to power the chargers. Also most importantly LiPos are dangerous when treated incorrectly, I have heard of people burning down houses due to charging issues/mishandling. Also they are also not nearly as tough as the lead acid packs and require more thought on mounting and protection. LiPos do however have the highest power density of most commercial batteries which is why they are popular in the RC world but with the safety risks alone they are unsuitable for FRC usage [sorry but I don't trust rookies with high explosive batteries]. However a pack that big could easily provide the power to run an FRC robot while being lighter then the lead acid battery, but the life span would also be about the same [500 cycles] making them even more price ineffective.

Lithium Iron Phosphate chemistry shows potential, but is still too new and costly. LiFePO4 has lower capacities per kilogram than Li-Po, but is inherently a much safer technology. It will not explode or catch fire if mishandled or abused.

I don't necessary agree with this, LiFEPO4 cells are price effective, very solid/safe, and rated for over 1000 cycles [Apple uses them]. But they are hard to get, not common place and have a lower power density compared to LiPo but still higher then Lead Acid. The cheapest method to get a A123 [well known brand of LiPoFE4] pack is to actually to take a part certain models of the Dewalt drill battery packs and savage the LiFePO4 cells from them and solder them up to make your own pack Click. However these are still expensive compared to lead acid, about $150 for 10ah/hr click and require an more complex charger to charge them [more $$$]. Price is the the main reason I see FRC not using LiPoFE4 for awhile.

And won't it be better to run at 100% power all the time? You would always have the motors spinning at the same speed, there would be no more variables.

I am not sure what this means, however even with LiPO/LiFePO4 batteries there is voltage drop when you draw large currents, and the voltages do flux as you use them. So you would still have to use PID/etc to account for different voltages when running things like a shooter wheel. And I doubt anyone in FRC would run there motors 100% all the time.

Also not all drill batteries are the same, some are great and last forever and some are crap and work for 5 secs [HF..] and not all Lit-ion cells are the same.

[AlexH]

i've seen LiFePO4 (A123 type cells) batteries sold as SLA replacements sold by chinese companies and comparable SLA batteries are more than 5x less.

you can also buy PCM boards that make LiFePO4 batteries act like SLA batteries

[ToddF]

Hijacking your thread: If FIRST were to take any technology from the RC world, I'd kill to be allowed to use brushless motors and ESCs in our robots. We could keep the expensive Jags and Victors to drive the CIMs for driving, if we had to. But, we would get tons better performance at a fraction of the cost if we could use brushless motors and ESCs. There is a good reason why nearly the entire RC world has made the transition to brushless.

[JVN]

Originally Posted by ToddF

Hijacking your thread: If FIRST were to take any technology from the RC world, I'd kill to be allowed to use brushless motors and ESCs in our robots. We could keep the expensive Jags and Victors to drive the CIMs for driving, if we had to. But, we would get tons better performance at a fraction of the cost if we could use brushless motors and ESCs. There is a good reason why nearly the entire RC world has made the transition to brushless.

This is the first time I've seen brushless motors suggested as a cost saving measure. Do you have a rough BOM of what you'd put in the kit spec'd out with pricing?

-John

[EricH]

Originally Posted by JVN

This is the first time I've seen brushless motors suggested as a cost saving measure.

Yeah... The same project that I worked on that used LiFePO4 batteries (a pair of 12V ones) debated brushed versus brushless, and quickly decided that the brushless ones were budget-busters (and we'd have to take even more anti-dust precautions than we did with brushed, which was a factor in that project).

[ToddF]

Originally Posted by JVN

This is the first time I've seen brushless motors suggested as a cost saving measure. Do you have a rough BOM of what you'd put in the kit spec'd out with pricing?

-John

I'll take our shooter from last year as an example. We used a pair of RS-550 motors ($7.25 ea) driven by a pair of Jaguar drives ($119 ea), for a total of $252.50 system price.

That motor is 36mm in diameter, has a Kv of 1608 rpm/V, max current draw of 85A and max power of 254 Watts.

For an apples to apples comparison, Hobby King has a 36mm diameter brushless inrunner (S3674-1660) for $39.20 each. This motor has a Kv of 1660 rpm/V, a max current draw of 60A, and max power of 1600 Watts.

To drive it, we would need a 60A ESC. The HobbyKing Red Brick 60A ESC is $13.20 each. This ESC is a fraction of the size and weight of a jaguar, though it is controllable via PWM only.

So, using the motors and power drivers mandated by last years rules, we had a total of 508 Watts of power available at our shooter wheel for $252.50 total cost. If we were allowed to use brushless motor technology, we could have had 3200 Watts of power at our wheel for $104.80 total.

More power, less weight, less than half the cost. Note that the real driver of the cost difference is the drastically lower cost of the drive electronics. They can get that cost so low by removing all the smarts from the controller and making a bazillion of them. I wouldn't want to eliminate Jags and CIMS altogether, just allow these better technologies as an option for use where appropriate.

[Jon Stratis]

Lets take a look at that S3674-1660 motor a bit more...

For an apples to apples comparison, Hobby King has a 36mm diameter brushless inrunner (S3674-1660) for $39.20 each. This motor has a Kv of 1660 rpm/V, a max current draw of 60A, and max power of 1600 Watts.

power = current x voltage. Running at 12V and 60A, the maximum power input into the motor is 720W. Where's the other 880W coming from? It's a perpetual motion machine!

Realistically, that 1600W comes from running it at higher current/voltages. That motor is supposedly rated to 22.2V, according to the website. But even at that voltage, 60A only provides 1332W of input power!

So, all of these numbers need to be taken with a huge grain of salt. Without actually testing the motor with our system (in other words, take your proposed set up and hook it up to an FRC robot) and measuring the output, the comparison can't really be Apples to Apples.

Your numbers significantly overestimate the power this motor/controller combination would provide you.

[AdamHeard]

Originally Posted by ToddF

I'll take our shooter from last year as an example. We used a pair of RS-550 motors ($7.25 ea) driven by a pair of Jaguar drives ($119 ea), for a total of $252.50 system price.

That motor is 36mm in diameter, has a Kv of 1608 rpm/V, max current draw of 85A and max power of 254 Watts.

For an apples to apples comparison, Hobby King has a 36mm diameter brushless inrunner (S3674-1660) for $39.20 each. This motor has a Kv of 1660 rpm/V, a max current draw of 60A, and max power of 1600 Watts.

To drive it, we would need a 60A ESC. The HobbyKing Red Brick 60A ESC is $13.20 each. This ESC is a fraction of the size and weight of a jaguar, though it is controllable via PWM only.

So, using the motors and power drivers mandated by last years rules, we had a total of 508 Watts of power available at our shooter wheel for $252.50 total cost. If we were allowed to use brushless motor technology, we could have had 3200 Watts of power at our wheel for $104.80 total.

More power, less weight, less than half the cost. Note that the real driver of the cost difference is the drastically lower cost of the drive electronics. They can get that cost so low by removing all the smarts from the controller and making a bazillion of them. I wouldn't want to eliminate Jags and CIMS altogether, just allow these better technologies as an option for use where appropriate.

Your comparison is completely unfair.

You are comparing US made versus Chinese made, you are also comparing an esc optimally sized for operation versus one far oversized for operation.

I love RC cars, and spend way too much money on hobbyking and can say with certainty I have used more motors and ESCs from them than anyone on chief. I love them for personal use, and a lot have surprised me with their performance and durability, but I would not put them on an FRC robot.

[Andy A.]

Originally Posted by JVN

This is the first time I've seen brushless motors suggested as a cost saving measure. Do you have a rough BOM of what you'd put in the kit spec'd out with pricing?

-John

I couldn't even begin to price that kind of thing, but it occurred to me I couldn't really price out the motors currently in the kit either. I know many moons ago nearly everything in the kit, motors included, were donated.

These days the number of kits has obviously grown and motor selection is a little more elaborate. Does anyone know how much of the in-kit motors are donated, purchased or ___? What about motors made available for purchase by teams? Are those discounted somehow?

Now that I think about it, I miss some of the old motors. The Globe motor was awesome.

[JesseK]

Originally Posted by ToddF

Hijacking your thread: If FIRST were to take any technology from the RC world, I'd kill to be allowed to use brushless motors and ESCs in our robots. We could keep the expensive Jags and Victors to drive the CIMs for driving, if we had to. But, we would get tons better performance at a fraction of the cost if we could use brushless motors and ESCs. There is a good reason why nearly the entire RC world has made the transition to brushless.

Keep in mind that the RC world deals with vehicles that are a few pounds or less, thus the motors do not draw as much current overall. This alone leads to a less expensive motor controller.