For outreach events to younger kids, our team will frequently bring a robot and let kids drive it. So, it can be a bit of an extended run and it’s sort of a pain to have to change the battery in the middle of that.
There are two solutions: (1) Just swap in a larger 12v battery, and (2) actually hook up two battery connectors (in parallel). On the second option, you could theoretically just plug in a fresh battery and then unplug the old battery indefinitely – basically it’s just a hot-swap.
Has anybody done this? If so, did you run into any gotchas?
We have a Demo Bot/T-shirt Cannon that runs on two batteries in parallel, though it wasn’t intended for hot-swapability, it does effectively work that way as an added bonus. Granted, in the case of our demo bot, it’s power efficient enough that we generally don’t have to swap batteries mid-event (It also uses “cheap and dirty” controls, so it boots basically instantly).
In general if you want to use two batteries, I would recommend using “competition batteries” if you can avoid it, but if you must use competition batteries, try to use batteries with similar charges and internal resistances (measured with a battery beak). That said, if you use the other battery just for hot-swapping (IE, you don’t run the robot 2 batteries on it at once), I don’t expect it would cause any issues, regardless of voltage/resistance, since the time the batteries spend actually connected to each other would be minimal.
If all you’re doing is allowing a quick swap, do you really need two robot connectors, or do you really mean that you won’t have to re-boot?
If you do decide to run two batteries simultaneously, try to match them as closely as possible - two brand new batteries, or (more likely) two “demo” batteries so you don’t end up discharging one battery into the other.
Another thing 3946 has done is implement a “nerf” mode for demo driving, which essentially turns off half the motor controllers on each side of the drive train. The first time we thought of it we literally unplugged them, but now it’s a standard programming thing usually added right after competition season ends. It was put in place to help reduce the likelihood of damage to people and equipment, but it also extends battery life.
If you attempt to hot swap two batteries connected via a wye cable to prevent a reboot, you are going to have a problem during the time interval when you have one fully charged battery and one depleted battery connected to your system at the same time.
The fully charged battery will attempt to charge the depleted battery as quickly as quickly as the pixies can flow.
If your aim is to reduce downtime, you could power the radio off of its own battery / power source to speed up the rebooting process after a battery swap. Since radio reboot is such a long part of the comms process, this may do enough of the trick for you to get by.
These currents can end up being very high since the only thing limiting the current is the internal resistance of the two batteries and the resistance of the wire between them. This can easily be much higher than you get when running a robot. Also, your fully charged battery quickly becomes not fully charged.
This is actually a really good idea. An FRC battery could power the base control system for many hours without issue. Though I would be somewhat curious as to what would happen with the CAN network if a bunch of devices suddenly loose power.
As an alternate to the solutions you present in this thread, Adam Heard from 973 put out a RAMP video a few months ago about using a 12V UPS (universal power supply) on practice bots to power the radio (and roborio?) to cut down on downtime for practicing. Your use-case is basically the same. Maybe you could try it.
Art and Philso,
The fresh battery will try to charge the depleted battery but actually at a relatively low current. The difference in voltage between the two batteries under similar low load will be small and therefore the current will be small. (Remember that lead acid cells require a higher charge voltage be supplied to force charge current into the battery.)
However, another solution would be to run down to your local RV center and buy a dual battery isolator. This would allow both batteries to supply current to the robot but not to each other. It might be a little pricey but would prevent the batteries supplying current to each other.
I like the second battery to keep up the Rio and radio though. I have to add…That requires a little wiring that is not legal for FRC competition. Of course a “y” adapter or battery isolator is not legal either.
A good strat that we’ve used in the past is to wire everything to the lead acid battery, and run just the radio on another battery. RIO bootup is quick, radio bootup takes forever, so keeping the radio on its own power source helps quicken the battery swap process.
This is what we did for our parade robot as well as bumped up the size of the battery for the drive system. The radio and cRio (yes cRio) are powered from a normal robot battery.
I’m not sure of what the battery is but it’s physically nearly two standard robot batteries in size. If it’s not available we can place two normal robot batteries in the mount and swap as needed. A parade would normally consume 3-5 batteries. We now rarely ever need a swap.
There will be no ill affects of hot swapping the batteries using a Y connector. As Al pointed out to get a battery to charge you need to provide it with more than resting voltage. So only a very small current will flow from the fully charged battery to the discharged battery. I would take many hours if not days for a set of mismatched batteries to deplete the charge from being connected in parallel.
A typical isolator as used in RVs won’t work in this situation. They are designed to allow one alternator to charge two batteries through the use of diodes. So you have one input and two outputs that are isolated from each other. Power can only flow from the Alt terminal to the Bat terminals not between the Bat terminals nor to the Alt terminal.
Either way it is a lot of money to spend to fix an issue that doesn’t exist in this application. The only down side to a Y cable is that if you leave both connected and not in use for long periods of time they will slowly discharge each other.
