Just wondering what you guys like to shoot for in your DT calcs.
-LM
If we current limit our 8 775-pro drivetrain to 20A each, we brown out during accelerations. We get away with 25 amps each with a bit of voltage ramping. That’s ~200A (brownout happens at ~5v voltage drop, which corresponds to ~0.025 ohms system resistance)
That sounds like you have a wiring problem - in my experience, a well-wired robot should be able to pull over 200A total before brownout, usually closer to 250A.
We have the exact same experience with out robot. We don’t use the voltage ramping because it introduces a floaty feel in the drivetrain that the drivers don’t like. We limit to 18 amps and don’t see a brownout through a match, so your numbers are about right. We use a variety of batteries (new and old) that are all amperage tested over a span of 12 hours. We know they are good.
On a fresh battery with no other subsystems we can certainly push that amount higher, but our objective is to get through a full match without a noticeable degradation in performance. So the settings we have were empirically determined so that we need to not be browning out at the end of match.
Edit: Our practice robot shows the same drivetrain characteristics.
Edit 2: I realized that I never answered the original post. We determine our low gear based on what we believe our average “short” sprint distance is going to be. For instance, from the scale to get a cube. Then we find our what we’d like our high gear to be for our long distance sprints - for instance from the starting position to the far scale. We try to find gear ratios and spreads to match, then we tell our drivers not to push. If you’re pushing it’s likely that you’re doing it wrong. It’s cool and flashy to push other robots, but I think this year we were only doing it perhaps 3 or 4 different times.
Can you explain how you translate a sprint distance to a gear ratio?
Isn’t this game dependent?
Calculating the time it takes to travel a distance can actually be pretty complicated in practice. You can create a spreadsheet using the gear ratio to calculate torque in different time increments. Sort of like this:
First line of spreadsheet
Torque at given speed (start with zero) -> Use time step (like 5 ms) of applying that torque -> speed at end of time step
Repeat that line over and over. Your torque will decrease as your speed increases, etc. This is motor dependent, so you need the motor values too. I’ve seen a couple spreadsheets floating around that do this.
Pushing is game dependent, but the maxim is generally true. If an alliance can put a lesser-performing team defending you and you allow them to get in a pushing match with you, then you’re playing their game and they get the net advantage. For example, this year it was possible to get locked into one side of the field when going to the portal. You should be taking that into account when setting up your strategy, and your drivers should know that the second a robot tried to defend them in the portal they should be heading around the other side of switch to get away (assuming you set your robot up with a sprint speed to do that).
There will be occasions where you get pushed into a wall and held there. Wait for the count and then leave in another direction.
You don’t look for a certain current limit to prevent brownouts?
I’ll second the statement that, unless your strategy involves playing defense or pushing very heavy game pieces, you’re better off designing your drivetrain around a given optimal sprint distance than pushing current. Normally, doing this would be time consuming and involve complicated spreadsheets and guess-and-check work.
Lucky for you, I just published v2 of my design calculator, which includes a sprint distance drivetrain gear ratio optimizer (i.e. the Reverse calculation on the SprintDistance sheet). All you need to do is enter the details of your robot (number/type of motors, wheel size, weight, etc) and the distance you want to optimize for, and it will calculate the optimal gear ratio that will get the robot there in the shortest time possible. Let me know if you have any questions on how to use it.
We found experimentally that our transition from full forward to full reverse was actually just as severe as pushing.
For us we like to stay below 180 amps or so total robot draw. Remember that you need to account for total robot systems - for instance if you’re pushing another robot fighting to pick up a cube.
In 2017 with 4 CIM swerve at ~7:1 with 4 inch wheels, we browned out at full acceleration, imposing a a 50 Amp limit we were fine - holding above 7 V and did not feel the debit in handling. We used this in 2018 and also no issues. Our testing includes use of drained batteries.
In this off-season we have been running 775 drive trains - a couple variants of 4 x 775pro swerve and a 6 x 775 pro tank drive - using 3d printed gearboxes.
Based on experience so far a 40 amp limit is probably fine for a 4 x 775 swerve - maybe make peak at 40 and steady state at 35 and do what Tom said about avoiding extensive pushing etc In this case the 775’s go through a versa-planetary 3:1 or 10:1 gearbox - 2 different variants. Sound is a little higher pitched but overall acceptable.
In off-season we have dialed current down to 30 SS and 35 peak to allow run for 2 hours at a time with only breaks to change batteries.
Difference is that for CIM drives it seems the limit is RoboRio brownout - even for 4x cim let alone 6x CIM or miniCIM, for the 775 pro drives it may be the motor temp (an 8 x 775pro may well get around this as demonstrated and discussed this year)
We are just getting the 6 x 775 tank drive going - one observation is that the 12 tooth to 80 tooth gear mesh is really loud! - have others seen this?- the gear mesh is not too tight.
99 dbs!
We didn’t measure it but the announcer at our first event commented on it when the bot went by in auto. Sounded like a swarm of very angry aluminum bees.
Angry aluminum bees - great description of the sound!
BTW I measured using a free Iphone app.
We usually don’t plan to do much pushing. That said, we base our calculations on a max allowed draw of ~200A on the system, and each breaker at ~125% of the breaker rating, though we would likely raise that if it were really short term (5-10 sec per match); I’m thinking of a can grabber, not a climber.