775 vs CIM

Yeah, we opted against shifting at the time given that we had never done swerve or even custom gearboxes before (it was a crazy experience for us designers, builders and programmers). We ended up with a fine precision mode, sprint mode, and normal speed. All three were used every match, and torque was not deemed a necessary thing if we could simply outmaneuver every other robot.

We are presently developing a two speed 3 motor module and both a single and a 2-speed two motor module. We love swerve. I’ll post pics when they r close to done.

I would list RPM as advantage CIM, because most FRC applications require speeds of 5K rpm or less, many MUCH less.

Due to gear losses, power is at best a wash for the 775pro, excluding only those occasions when you need half a horsepower at over 5K for less than ten seconds.

This idea that “a 775 is only more powerful than a CIM for 10 seconds” doesn’t make any sense to me. Is the basis for this the idea that the output power of the 775 drops when placed under enough load to put it in the max power point on its speed / torque curve? I don’t know why it’s useful to say this.

That said, yes power is essentially a wash between the two motors factoring in the lower efficiency of additional reduction (if needed). The difference certainly isn’t super notable.

And this is coming from the guy who is pretty much to blame for the CIM right? Or am I getting you confused with someone else? I remember reading about the introduction of the CIM to FRC recently in a post but I cannot remember all the details.

In a thousand words, I think it’s based on this chart. Based on their website, I think it’s set to a speed and the torque is measured. Either way, this has implications for designing anything that has to be substantially loaded for periods longer than a couple seconds. So basically drivetrains and maybe bulk shooters (2006/09/17).

The implication here is that the 775pro only has more power than a CIM at max power, when it has (slightly) more power than the CIM all along its power curve. This is pedantic and hair splitting, I know, but so is saying “the 775pro is only better than a CIM for 10 seconds”.

Loading either motor at max power at 12V for an extended period isn’t a fantastic idea due to the risk of tripping the 40A breaker. For mechanisms that need to be loaded for an extended time, it is a good design practice to aim to load that mechanism for a point on the power curve where current draw is 40 amps or less. This is the design theory behind a lot of low gear selections in drivetrains (“traction limited at 40 amps”)

You have the right person.

The CIM is still a great motor for FRC. There is a saying I have heard in the DC motor world where Power is cheap, but torque is expensive. The reason being a low speed high torque DC motor tends to be extremely expensive compared to a high speed high power. What do you do? Add some geartrain… which last I checked is very expensive and relatively high loss in FRC (compared to high efficiency industrial or transportation drives).

775 Pro is a great motor, but you have to be very smart about using it. Much like the 6 motor drives:
this season I have watched 2 3XXX teams have a successful first half of their season, already played in 1 district, and then at the second, start popping Main Breakers during elims. Why? Teams were getting good enough that defense starting coming into play, which caused higher draws, more low speed maneuvering, which in turn pulled high currents. Later they implemented some current control SW and were back into good shape (Thanks CTRE for the capability).
I then saw at Wolrds a sub 1000 number team pop their main with their 6 CIM drive for the first time (after 2 districts, a regional, and a State championship). Why? they played harder defensive than ever before trying to shut down a scoring powerhouse. We talked and they replaced the breaker and added current limiting controls and I believe they resolved their issue for the following day, but they had the World Watching when they died in a match…

With a bot-load of 775s, not only do you have to be very careful with your Main Breaker, but playing defense, there is a pretty decent chance you might frag a motor as the motor fans do not move much air at low speed, and the currents are really high. If you are planning on this set-up, better have a solid solution for helping your motors breathe better.

Now, if you have the capability to solve these and a few more differences, then I think doing a 775 drive is an excellent idea. Just remember, all of these new efforts result in new work, which without new resources will pull resources from programming Auto routines and state machines, pull mechanical design of a manipulator for coming up with different ways of gearing and controlling motor heat. $$ you would normally spend on scoring buying additional gearboxes to make sub-assemblies so that when you do fry your motors, you have a whole set of spares to swap in…

I still think it is a really cool and potentially amazing set-up to pursue. I just hope teams realize, it is a more difficult set-up to get right. Remember, that when intentionally traveling a harder trail to follow, it will require more effort to get you from A to B. If you have the resources and preparation, you may experience a richer, fully trip with beautiful views, but you could also twist an ankle, get lost, show up late to the rendezvous, or end up with poison ivy.

IKE,

Great segway …

My team hasn’t used a CIM motor since 2014! However, we have replaced them with miniCIMs. The one thing most people are missing in this comparison is the total available electrical power in the FRC robot power system.

The CIM is a great motor that was not designed for FRC applications. The 775pro is a great motor that was adapted as best it could be for FRC applications (changed the windings, upgraded to a better ball bearing on the rotor shaft). The miniCIM was designed specifically for the FRC application.

This is very important to understand. Look at the power curves to get an understanding what total power the breakers really allow. Add to that the current draw and voltage drop when you are using 12 - 16 motors and the CIM and 775pro should not be used everywhere.

Now Dr. Joe likes them for the drive. We used them for the drive last year and didn’t have problems with the motors. However, there weren’t any gears on the market that could take the direct power from 3 775pro motors (I’m looking at you 60T 32DP aluminum gears). Using the 3mm belts from WCP would probably solve that but that is a discussion for another day.

I much prefer the 6 miniCIM drive we used this year. No breaker trips; acceleration for days; could practice for hours at a time and the motors just didn’t care. As a matter of fact, our limiting factor for practice was the main breaker getting so hot it would just trip itself (this took hours of continuous practice before it would happen).

Our robot had the following motors:

6 miniCIM for drive
4 775pro w/ 3mm belt for shooter
2 775pro w/ 10:1 Versaplanetaries for feeder roller / climber
2 775pro for hopper feeder system (thanks Barker, BTW)
1 775pro w/ 20:1 VP for gear intake roller

I am having my engineering team at VEX do some comparison testing with an actual FRC power system to see what you really get with 775pro vs. miniCIM vs. CIM as I really need to see some quantitative data. I can tell you our motors.vex.com testing was quite revealing to me about the CIM motor, specifically. The time studies and time to cool down were pretty eye opening.

Paul

two things:

  1. Everyone seems to be discussing this like its’ a dichotomy–CIM or 775pro. Is anyone using anything else on drive? I’ve considered using six MiniCIMS might be a good alternative…

2.16 lbs, 215watts, 5840 rpm, 65mm x 91mm, $24.99

https://www.vexrobotics.com/217-3371.html

  1. If the only drawback of a 775pro is that it overheats if the internal fan stops, this seems it could be easily fixed by designing a manifold and an external blower that would force air (lots) through the openings whether it was turning or not…
    And this is FIRST, so surely someone’s tried that…

My guess is that it would still fail when stalled because the current draw is just too high then regardless of air movement.

Edit: Funny that I was sniped by the President of Vex by two minutes…

Here is exactly what we do for that test:

3 Minute Peak Power Test

Peak power data was acquired experimentally using the following test:

  1. A motor is spun at free speed
  2. A brake is quickly applied to bring the motor to 1/2 free speed (theoretical max power)
  3. The motor is held at this RPM for 3 minutes and power output is recorded

Note: If you know the efficiency of the motor, you can also determine how much heat is being absorbed by the motor during this test. This is also useful for learning how fast a motor conducts heat to the exterior.

This is found on this page: http://motors.vex.com/#testing

Click on the “How we test …” button to see all of our test methods.

So we apply an output torque that brings the motor to its theoretical 1/2 free speed and hold it there for 3 minutes (or until the motor gives up).

Something worth noting is that a lot of discussion of using 775Pro motors is by using them capped at 8-10V power, due to their increased robustness at these voltages. If you do the math and derate the motor specs for these lower voltages, you’ll find the 775Pro’s power output at 10V is only slightly higher than a mini-CIM. You are still saving a little weight, depending on how you gear your mechanism, but a lot less weight than you would save vs full CIMs.

Of course, this doesn’t precisely apply if you are controlling current instead of voltage, YMMV

Paul,

I’m having trouble rectifying something here…

If MiniCIM=2/3 CIM, wouldn’t a 6 MiniCim drive = 4 CIM drive performance-wise?

The 4x CIMs would be lighter than 6x MiniCIMs
The CIM is more efficient than a MiniCIM by a reasonable amount

Is the big advantage of 6x miniCIM drive being able to pump current through more wires? Or is it something else that I’m missing?

I would think one reason is that 4x CIMs @ 40A gives 1100 Watts, whereas 6x miniCIMs @ 30A is 1218 Watts (according to Ether’s motor calculator). The miniCIMs would pull 180A total vs 160A, but each auto-resetting breaker is less likely to trip.

Two main benefits

  1. Spreading the load over multiple individual motor breakers.
  2. The increased thermal mass is better from a robustness point of view.

I can’t speak for Paul, but the reason we went with a 6 miniCIM drive is because you can exert the same torque and spread roughly the same current across more 40a breakers. You get to gear more aggressively (or just operate more safely at the same gearing).

That chart shows that the power output drops after several seconds. Since the speed is being held constant, that means the torque output drops after several seconds. So I’d say the statement is pretty accurate. Also, as a sidebar, gearing for 40a is a bit conservative. I’m perfectly comfortable with 50a or 60a. Peak power is 70a, which will hold for 2 to 9 seconds. These days I’m more concerned about drivetrains causing brownouts than blowing 40a breakers (because we’re using 6 breakers, of course).

I wasn’t referring to operating at different parts of the power curve, but the target free speed of the geared system. If that speed is 5000 rpm or less, the gearing losses for the 775pro will more than eat up the small difference in shaft output power between the two.

And yes, the 10 seconds is based on the peak power vs time curves for the two motors from Vex. They cross at about 7 seconds duration.

Edit: In general, when designing a drivetrain with CIMs, your primary limitation to avoid system dropout is the 40A breaker, followed by brownout. In designing a drivetrain with 775pros on 40A breakers, your primary limitation is the motor. I haven’t dug deeply enough to see whether a 775pro motor or a 30A breaker is more likely to go first.

For those who did not follow our other posts/threads for our experience with running a 8X 775Pro drive train this year, here is our actual experience with this setup. I know I would be very skeptical also if someone told me that 8x 775Pros would not burn up under almost any circumstances (we did not encounter this so I cannot say what would cause this to happen). I appreciate all the theoretical discuss going on here, but we have actual data from three regional events and a Championship run.

We ran the same motors on the practice bot and the competition bot the entire season without burning up/replacing a single motor. This is omitting the fact that we burned up two motors the first day we incorrectly programmed the prototype chassis during the build season. No fault of the drive train itself. We also did not see a foreign object jam up the gearbox. I am curious about what effect this would have on the motors. This is why I would like to have safeguards that monitor the encoder speed and limit motor power if the encoders are not turning while the motors are powered on. Something to play around with in the off season.

We had NO issues with motors this year when running 8X 775Pros on our drive train. We did trip the main breaker once at Miami Valley, due to the voltage limiting only being programmed in forward and not reverse, we were pushing a bot while driving backwards at the time.
We did trip the main breaker twice at Seven Rivers from heavy defense.
We then added code to reduce motor voltage when the current was too high for too long for Champs and did not trip the main breaker during Champs even while heavy defense was being played. We were able to push our airship across carpet without tripping the main breaker using this new code.
We plan on playing with current limiting the motors instead of voltage limiting. Unfortunately our code/time did not allow us to integrate advanced Talon controls on the drive motors during the season.

I do not recommend running one, two or even possibly three motors on their own wheel. For example, you run a single 775Pro to each of 6 wheels on a WCD or you run a single 775Pro to a single swerve module. What happens when you get into a pushing match and your front/middle wheels come slightly off the ground/unload? Your two rear motors get all the load on them and could easily overheat them. This happened to other teams this year and I cringe at the CAD models I have seen on CD recently. I believe that 6x 775Pros would be able to work on a drive train, but I believe their operating temperature would increase to a potentially damaging level. Voltage limiting 6x 775Pros and running them with expectations close to a 4x CIM drive train could work. I do not recommend running a 6x 775Pro drive train with the motors running full out, this will in my opinion cause issues. At the least you will be more likely to burn up motors, whether you do or not. We originally were going to run 10X 775Pros for safety but I believe that is overkill now.

Your spur gear that the motor pinion gears drive needs to be strong. Period. All four motor pinions will be wearing on this gear and therefore the wear will be 4X higher than usual. We used a .1875" wide custom EDM’d 4140 gear. We wore out 303 stainless steel gears in our first regional event. The 4140 gears showed no wear after two regionals and Champs. I like the Andy Mark 85 tooth gear I saw on CD the other day, and I hear that WCP is going to offer steel 32DP 60/80/100 tooth gears soon.

The battery can only provide so much power and the motors share that available power. If you directly tie together 4X 775Pros on each gearbox AND limit them to 80% voltage (9.6volts) they will barely get warm and never get to hot. I can still see the serrations in the brushes on the motors from an entire season of use. This is the key to not smoking the motors. Running 8X 775Pros at full voltage will cause the system to brown out. Before we added any voltage limiting code, the motors were running at 100% voltage and if the drive joysticks were moved to quickly, the bot would stutter during acceleration, I believe we were drawing the battery voltage down too far and the system was browning out.

Design the correct gear ratios into the gearbox. All bets are off if your theoretical free speed is 40 feet/second. Our low gear theoretical free speed was 5.2 ft/sec, ultra-low to reduce the chance of burning up the motors in a pushing match. High gear was 19.12 ft/sec. We were traction limited in low gear but could stall the motors in high gear, so we had months to burn up motors in high gear and did not.

I’m very interested to see a comparison of the power output between 6x CIMs or 6 Mini-Cims verses 8x 775Pros, AFTER running them full out for the equivalent of a eliminations run (quarterfinals-finals). When limiting the voltage to around 9.6volts, we found that after a 6-9 match elimination run the 775pros were only lukewarm. I know in the past we had 4X CIM drive trains that were too hot to touch. Looking at the motor power output verse temperature of the motors that I saw last year the power drops off with an increase in motor temperature. I would have to believe that the 775Pros being lukewarm would be putting out more power after back to back matches then the CIMs/MiniCIMs due to the heat retention of those motors. We have never ran 6X CIMs or MiniCIMs so that might keep the temperatures down also.

Can you please elaborate on the gearing/game for 4x CIM drives that were too hot to touch?

Not the OP, but ours got too hot to touch in 2016. We had 12.75:1 gearing on 8" pneumatic wheels (~13 ft/s).

2016, Pneumatic 8WD, Low Gear=6.15 theoretical free speed (ft/sec), High Gear=15.76 theoretical free speed (ft/sec)
I know last year during practice, CIM motor temperature was the limiting factor for the # of practice matches we could run at the shop, not the usual charged battery issue. Granted, the pneumatic tires give a higher friction on the carpet then other wheels do.

2013, Nitrile 8WD, Low Gear=6.06, High Gear=17.69
We use this chassis a lot for defensive practices
Very fast bot, very low rolling resistance
Still heats up the motors fairly quickly