PDP Limitations, workarounds, challenges

Sparked by this thread, i want to learn more about the PDP and it’s infamous 16 motor slot limit.

The only robot I’ve heard of being 100% motors is 148’s Uppercut (2018). Now, if we even tried to go only motors, I know my team would fail horribly. We would either run out of slots, or fail to achieve the goal(s).

How have teams who face this 16 slot challenge worked around it year-to-year? What smart designs are there that teams can greatly benefit from while avoiding the limit?

Many swerve teams tend to max out on motors, since they only have 8 slots to work with instead of 10 or 12 after the drivetrain. That being said, 100 also ran out of PDP slots this year. This was mostly because we didn’t want to use pneumatics on a robot which barely had room for a battery, much less a compressor.

Generally, to work around few motors lots requires that you get clever. Most teams running swerve also have pneumatics on the robot, simply to give them a few more degrees of freedom. If not, you also have to be willing to design around a severely restricted amount of power, which means gearing and motor count for every subsystem requires much deeper and longer thought that it might be if one could just slap 2 Falcons at every problem.

1 Like

1899 also ran out of PDP slots this year. We used bag motors to take advantage of 2 motors in 1 PDP slot. We ended up using 17 motors in 15 PDP slots and the last slot was for the limelight.

This year 4099 used no pneumatics and 15/16 PDP slots. We built a high shooter with a Limelight. We also didn’t have a control panel mechanism or turret, but would have had room to add one of those. If we wanted to add more degrees of freedom beyond that, that would have needed pneumatics to either take over for our motorized wrist or to be used for that mechanism. It also forced us to solve the problem of staying up after T=0 by assisting our elevator with a constant force spring, instead of using a brake which would have required pneumatics.

Overall, it was fairly easy to design a robot which avoided pneumatics, and only required a couple of unconventional choices. I think most teams, regardless of if they used or didn’t use pneumatics, ended up near the limit this year if they made a shooter of some sort. Generally, avoiding pneumatics does make you think more about adding another mechanism and figuring out if it’s really worth another motor and PDP slot, instead of just throwing on another solenoid.

Also, this year didn’t lend itself to having pneumatics solve the challenge significantly easier than motorized mechanisms, like hatch panels in 2019 ended up being. Additionally, this game is abnormal in recent years for requiring “do everything” and even mid-tier robots to run up against the limit, which is likely shifting the replies in the other thread somewhat.

1 Like

I cannot stress enough how much i have seen of this. after stronghold it seems your robot needs be good at EVERY task to even be considered for a higher pick.

Sure sometimes teams go off of helping tasks like triple climb robots in 2018 power up or ramp/triple climb bots in 2019 but the fact that teams would have to base their robot ENTIRELY off of helping rather than scoring kinda takes some fun out of the game

1 Like

This years game made it almost impossible to put out major parts with pneumatic and the need for a high scoring robots space makes it hard for anyone to find place for a compressor (there are much smaller compressors out there, why cant first make them legal yet?)(https://www.nature.com/articles/sj.bdj.2017.137) is a good example which hits around a decent 115 PSI

because of this teams use an abnormally high amount of PDP slots

I don’t know if I agree with this, and you’ve actually made the exact opposite point my post made. There’s very good examples of robots which didn’t do everything and still were very successful in the past few years. In Steamworks, robots that didn’t even touch fuel were successful at even the highest levels of play. For Power Up, it was important to climb and do the scale sure, but those could both be accomplished with a single elevator and even at champs I saw a lot of robots who didn’t have a buddy climb still double climbing and getting the RP with good placement. In Deep Space, some of the most dominant robots didn’t go high, including 973, who physically couldn’t get the rocket RP and still managed to seed first on Newton and win a championship, as well as 2910, who put up some of the highest scores in the game. This year was unique because you needed a lot of degrees of freedom to effectively index the balls, which any team who wanted to shoot couldn’t ignore, and the imbalanced weighting of climbing made that a necessity, which was yet another degree of freedom.

4 Likes

So, we had an issue with running out of PDP slots this year as well because we had 15 motors and a REV servo expansion module. One big thing to look into are those Ozzyboard Super Servos. The REV Servo expansion hub takes one PDP slot and can support up to six servos, so if you can replace more than one motor with servos, then definitely go for it. We used 2 REV SRS Servos for hood adjust on our shooter and it worked wonderfully.
Another thing we discovered is that the three 90 deg. motors (Snowblower, DENSO Window, and Bosch Seat Motor) can be run off of relays so you can have two motors on one MC without having to run the two motors at once.
We have a pretty intensive vision programme this year that required a Raspberry Pi 4 and a lot of CPU power… and then three additional Pi 3s for regular drive cameras (I think the T265 camera we used for vSLAM w/ Sensor Fusion fit in there somewhere, but don’t quote me on that) and we found out that we could have as many 5V 20,000 mAh batteries on the robot as we needed.
One big thing that we encountered were power draw issues which was one of the other reasons we had to move our Pis off of the PDP since we were getting brown-outs. Something you can do to help that is you can ramp the power going to each motor a little bit, even over 1/100th of a second, you will save power. It’s sort of like how flooring your car uses more gas than slowly accelerating.
One final thing: Ask yourself if you really need all of the motors you’re using. For example this year we had 1 NEO on our climber and it worked great because we had a pnumatics-less shifter (it used two roller clutches and shifted by changing the direction the motor spins, I’m going to make a post about it pretty soon), our shooter only has one NEO on a 6" flywheel and we were able to shoot 70’ and we can shoot all five balls in just under a second. Then our extendable intake folds down by driving forwards and stopping because it barely sticks out from the bumpers and it’s easy to replace parts. Also, in a game like this one, not having an extendable intake is not the end of the world.

No-pnumatics on a small robot gang

Is this saying all slots are filled including or excluding pneumatics? 4513 hit up against the limit this year even with pneumatics, which led to us cutting our Control Panel Manipulator last minute.

Brushless motors can help a lot. There was a time where 3 minicims (or sometimes 775’s)/side on the drivetrain was the play, now its 2 NEO’s/falcons, which saves weight and slots.

Making full use of motors and things like pneumatics/custom circuits (servos or pneumatics for shooter hoods, using 2 bags on one motor controller somewhere (hopper agitator maybe?)) and understanding just how much power a mechanism needs. Our climber this year was a bit overdone with 2 CIM’s, but there were teams this year who climbed with a single 775. Understanding how to get the most out of motors (or how to augment it with things like pneumatics) will go a long way in preserving PDP slots.

1 Like

4 drive
1 intake
2 “Serializer” (v-hopper)
1 ball feeder
1 azimuth
1 hood angle
2 flywheel
1 booster (wheel before flywheel)
1 climber
1 color wheel spinner
1 limelight
__
16 total

And we had plans for more! We had workarounds to free up some slots, but they would either reduce performance (reducing the number of flywheel motors), or increased mechanical complexity (trying to mechanically link the Serializer and/or feeder motors).

With new motors getting lighter and smaller (allowing each mechanism to get it’s own motor), and FRC games continuing to favor “side quests” requiring their own mechanism, I don’t see the need for more PDP slots going away.

Don’t need to waste motor slots on indexing if you don’t index :slight_smile:

1 Like

The teams that I’ve been a part of regularly don’t use pneumatics.

501 this year:
4 drive
1 intake
1 serializer (one or two BAGs)
1 ‘ballevator’
1 turret azimuth
2 shooter flywheel
1 climber
1 limelight
1 WOF spinner (that never got used)
1 intake retractor (that never got used)
__
14 total; only 12 used to win an event

1758
In 2018 we used 10 and pneumatics
8-swerve
2-elevator

2019- 6
4-drive
1 arm
1 intake

2020- 8
4 drive
1 intake
1 feeder
2 shooter (only one at competition)

1 Like

This year we were closer than ever to the limit.

Breakdown was:
4x DT
2x Climber
2x Shooter
2x Hopper
2x Ball storage
1x Intake

Plus a motor controller lying around for an eventual Control Panel spinner, pneumatics, a limelight… yeah, we were getting close.

That being said, I can’t recall the last time we were this close to the limit.

Last year we were:
4x DT
2x Arm
1x Intake

Plus pneumatics and a limelight.
So pretty light.

I don’t really mind the 16 slot “limit”.

We pushed the limit this year!
6 motor drive train,
2 motor shooter,
2 motor climber,
2 motors for ball handling,
1 motor for shooter azimuth,
1 motor for shooter hood,
combined 2 motors for wheel of fortune with intake,
still needed to service sensors, a limelight, and an LED driver.
So we designed and built a custom breakout board.

Sometimes limitations create opportunities!

We’ve run into the limit the past two years.

In 2019 we had:
6x drivetrain
2x elevator
1x wrist
1x intake
5x climber
1x electronics (Jetson, 2 cameras, targetting LEDs)

In 2020 we have:
4x drivetrain
1x intake
2x conveyors
3x shooter
1x turret
1x hood
1x color panel
2x climber
1x electronics (rpi, 2 cameras, targetting LEDs, decorative LEDs)

That being said, I’m happy with the limit staying where it is now. I think this allows a good balance of being able to do a lot of tasks without teams being able to use truly dangerous amounts of power (limited by the battery) or being able to just throw more power at problems willy-nilly without thinking about the repercussions. It also leads to some amazingly cool designs like 1690’s three position shifter. As I said in the other thread though, I’d like to see an extra port or two added that can only be used for CUSTOM CIRCUITS so teams aren’t penalized by losing a motor for adding more complicated vision systems and other cool electronics.

4 Likes

In 2020 we used:
4 drivetrain motors
1 intake motor
1 hopper motor
1 feeder motor
2 shooter motors
2 climber motors

…For a total of 11 motors, no where close to the limit. If you are running low on slots, think about how you can simplify your design or combine mechanisms (like shifting gearboxes or using PTOs). We could have used a PTO on the drivetrain and dropped the climber motors, but we were no where near the limit so we didn’t need to make the robot more complex.

In 2017 we used 5 motors (4 drivetrain and 1 climber… had a passive gear mechanism) and placed 1st at our regional (as rookies) before playoffs.

In 2018 we used 8 (4 drivetrain, 1 elevator, 2 manipulator, 1 climber)

In 2019 we used 7 (4 drivetrain, 1 ball scorer, 2 climber)

Simplicity in design is something we strive for every season, with long discussions about how we can do things in the easiest way possible.

We had this issue last year since we didn’t use pneumatics and had to use some creative thinking to get around it.

Initial 2019 Robot:
6x Drive (MiniCIMs)
2x Arm Rotation (NEOs)
1x Arm Extension (NEO)
1x Arm Wrist (775pro)
2x Cargo Intake (775pros)
2x Front Spider Legs (755pros)
1x Rear Spider Legs (775pro)
1x Custom Circuits (2x REV Servo on REV Power Module for Hatch Intake, 2x Limelight 2s, network hub, 12v sensors)

Mid-2019 Robot:
6x Drive (MiniCIMs)
2x Arm Rotation (NEOs)
1x Arm Extension (NEO)
1x Arm Wrist (775pro)
1x Hatch Intake (Window Motor)
1x Cargo Intake (2 motors 1 controller AndyMark PG motors)
2x Front Spider Legs (755pros)
1x Rear Spider Legs (775pro)
1x Custom Circuits (2x Limelight 2s, network hub, 12v sensors)

Final 2019 Robot:
6x Drive (MiniCIMs)
2x Arm Rotation (NEOs)
1x Arm Extension (NEO)
1x Arm Wrist (775pro)
1x Hatch Intake (window motor)
1x Cargo Intake (775pro)
2x Front Spider Legs (775pro)
1x Rear Spider Legs (775pro)
1x Custom Circuits (2x Limelight 2s, network hub, 12v sensors)

Are you saying that you cascaded a pair of relays and small motors behind a motor controller? I always read the first sentence of R29 as preventing that. Even if legal, it seems like putting a relay behind a motor controller would still only allow each motor to rotate in one direction so as to not give reverse polarity to the relay.

R29. With the exception of servos, fans, or motors integral to sensors of COTS computing devices permitted in R27, each actuator must be controlled by a power regulating device. …