Experience With 6 Cim Drive train

[Deke]

Originally Posted by Thad House

swiching between 10 fps in the sweet spot, where we never blew once, and 15+ fps for the sprints when we need to.

Bingo! This would optimize the 6 CIM drivetrain layout with the breakers allowed.

I think people are use to the 4 CIM drivetrain gears and their limits, the 6 CIMs can take a little more than what is expected. They are two different animals.

[Oblarg]

Originally Posted by Infinity2718

Bingo! This would optimize the 6 CIM drivetrain layout with the breakers allowed.

I think people are use to the 4 CIM drivetrain gears and their limits, the 6 CIMs can take a little more than what is expected. They are two different animals.

FWIW, last year with HiGrip wheels and a 120lb (including battery) robot, we were just on the very edge of traction-limited with 6 CIMs geared to 14fps (6.1:1, 4'' wheels) - i.e., we were traction limited if literally nothing else on the robot was drawing current, else we would stall the drive. Now, admittedly, we could probably have clawed back a bit of overhead with more efficient wiring, but...

Say a 150lb robot with plaction tread (COF of ~1.3 instead of ~1). So, 14 fps * (1/1.3) * (120/150) = ~8.5fps. For a comfortable amount of overhead, I don't think I'd set my low gear for anything faster than 7.5-8. Of course, you can adjust this upwards as appropriate for a lighter robot or less-frictioney wheels, but 10 fps would almost certainly be faster than I'd be comfortable with for a max-weight robot with high-traction wheels.

[jeremylee]

Originally Posted by Thad House

swiching between 10 fps in the sweet spot, where we never blew once, and 15+ fps for the sprints when we need to.

Which is why I recommended a 1.6 to 1.7 ratio from vex pro for thier 3 cim ballshifters as I reached same conclusion for this years game. I was thinking of it as a turbo button when that little boost is needed.

However, I think sweet spot 6 cim single speed may suffice for other years games.

Anyone push with blue nitrile geared around 12 ft/sec with 6 cims? I'm curious if increased COT impacts stall current enough to cause trips.

[Rob Stehlik]

Originally Posted by Andrew Lawrence

I've always read about how y'all prefer to gear more conservatively with your 6 CIM drives to abuse the acceleration boost, yet whenever I play around with acceleration and time to distance calculators I find that the extra motors become decreasingly more useful the closer you get to the "sweet spot" you have determined. At higher speeds - regardless of target distance - my calculators show that the extra motors can save a tenth of a second or so (nothing significant), and at lower speeds (especially around your sweet spot) the calculator shows improvements of a few hundredths of a second at best. Could you go into detail as to how you determine how much time you save with the extra motors to travel a set distance, and what tools you use? I feel I'm doing something wrong with the calculators I have due to the disagreements my data is showing compared to yours, but I cannot figure out what it is.

You are correct that the difference between 4 CIMs and 6 CIMs is not huge when geared lower. To be honest, we didn't compare the performance of different numbers of motors in great detail. The kids wanted 6 CIMs, so they got them

We use a spreadsheet that was originally written by Andrew Keisic on team 294. It should be here on CD, but I looked the other day and wasn't able to find it. Anyway, the spreadsheet allows you to input various drivetrain parameters, and outputs time to target distance as well as distance vs time graphs. Based on the assumption that we were using 6 CIMs, and choosing an average distance we think we would be driving during the game, we adjust the gear ratio until the time to drive that distance is minimized.

In answer to your question, if we pick a distance of 20 ft, the spreadsheet predicts it would take our robot (geared for 11 fps) 3.2 seconds to get there. Dropping down to 4 CIMs, the time increases to 4 seconds. As you said, it's not a huge difference, but perhaps more than you are seeing with your calculator.

A spreadsheet provides a method for relative comparison between drivetrain setups, but it's never going to be perfect. We also like to compare current designs with past robots. We were pretty happy with the drivetrain on our 2013 robot, so ultimately chose a gear ratio last year that would give us similar performance, even though the construction was completely different (gear drive vs chain drive). This is easy to see on the distance vs. time graphs.



One thing I would like to do is gather some real data from past robots to see how accurate the spreadsheet results are.

[jeremylee]

Originally Posted by Andrew Lawrence

At higher speeds - regardless of target distance - my calculators show that the extra motors can save a tenth of a second or so (nothing significant), and at lower speeds (especially around your sweet spot) the calculator shows improvements of a few hundredths of a second at best.

I see a difference of .15 secs to travel 20ft geared around 12ft/sec. Not significant like you said at these lower ratios.

But I also see a cim speed of 2850 rpm with 6 cims versus 2134 rpm with 4 cims to spin the tires of our 2014 robot in high gear in a model I built. Thus, it seems adding more cims decreases speed drop under load making the drivetrain "feel" stronger while consuming the same current? This may seem odd until you understand the cim motor efficiency curve...

http://www.usfirst.org/uploadedFiles...rves_Rev_A.pdf

Ideally, we'd always run our motors at peak efficiency instead of peak power. Only caveat is when you gear so aggressively you stall the motors where adding motors only makes the problem worse.

[timytamy]

We had 4 CIMs and 2 Mini CIMs, geared at 5.95:1 on a AM14U with 4" wheels.

Before our regional, we ran out of 6AWG wire, and the local supplier only had 8AWG. So for around a week we ran with 8AWG and a 70A breaker. During high acceleration/deceleration or turning, with the compressor running, we did actually trip the 70A breaker a few times.

However, when we got to Hawaii (with the correct wire and breaker) we got in a number of pushing matches against teams with shifters or 6 CIMS, but we never tripped the 120A breaker.

I should note that the 70A breaker wasn't up to the same quality of the coper busman ones we have in the KoP (felt light and cheap), I suspect that it was tripping on 70A of instantaneous current, while the 120A breaker will handle 120A for quite some time. I know this is a crude measure of current draw, but it is one none the less.

Hopefully that's useful

[jam2014]

On the upside, there should be huge opportunities with the new 2014 Power Distribution Board on board current monitors to detect and manage excessive power drain.

If the average of 6 cim currents is more than X amps for y seconds
turn off 2 cims ?
power back 30% ?
warn the driver ?
fire driver ?

Hope the PDB current monitoring works as well as I hope.
I could be wrong but last I heard, it was only on the 40amp outputs ?

[BJC]

Here is a tip for everyone in this thread who are popping their main breakers or looking to run 6 CIMs and gear fast:

1. Use lower CoF wheels.

2. Build lighter

Put down the Roughtop tread in favor of something like the VersaWheel DT or a Colson wheel. Consider constraining yourself to 100lbs instead of 120lbs. There is a reason that a team like 1625 could gear a single speed 12:64 (~17.3 fps free for those of you keeping track at home) and still T-bone like a monster this year. Lower CoF wheels and lighter weight mean pulling less current when accelerating and at stall.

Cheers, Bryan

[Oblarg]

Originally Posted by BJC

Lower CoF wheels and lighter weight mean pulling less current when accelerating and at stall.

This isn't strictly correct; a stalled motor will pull the same current regardless of the CoF of the wheels. Rather, lower CoF wheels make it harder to stall the motor in the first place.

[Thad House]

Originally Posted by BJC

Here is a tip for everyone in this thread who are popping their main breakers or looking to run 6 CIMs and gear fast:

1. Use lower CoF wheels.

2. Build lighter

Put down the Roughtop tread in favor of something like the VersaWheel DT or a Colson wheel. Consider constraining yourself to 100lbs instead of 120lbs. There is a reason that a team like 1625 could gear a single speed 12:64 (~17.3 fps free for those of you keeping track at home) and still T-bone like a monster this year. Lower CoF wheels and lighter weight mean pulling less current when accelerating and at stall.

Cheers, Bryan

We were 100 lbs and VersaWheel DT's at IRI, yet we still popped. If you are geared high enough to not be traction limited, Stall is Stall, so weight and CoF don't make much of a difference, other then change where traction limited is. We popped because we pushed, not during accelerating.

[BrennanB]

Originally Posted by Thad House

We were 100 lbs and VersaWheel DT's at IRI, yet we still popped. If you are geared high enough to not be traction limited, Stall is Stall, so weight and CoF don't make much of a difference, other then change where traction limited is. We popped because we pushed, not during accelerating.

I think the idea is it's harder to stall with lower CoF on the wheels. However if you do stall anyways, you have the same problem.

4476 ran a 6 CIM at a 13-14 ft/s theoretical. We never popped a breaker, and I suspect it was a combo of lower friction wheels, and we put a couple of breakers on the PD board for the drive CIMS at 30A instead of 40A. The idea being that the PD breaker would trip before the main breaker would.

I'm not really sure which made a difference as both were implemented prior to testing to make sure we never blew the main breaker, but it makes sense to me that both would have had a significant factor in preventing main breaker trips.

[BJC]

Originally Posted by Oblarg

This isn't strictly correct; a stalled motor will pull the same current regardless of the CoF of the wheels. Rather, lower CoF wheels make it harder to stall the motor in the first place.

Well yes. A stalled motor is a stalled motor. I was referring to stalling your drivetrain. I don't claim to be any sort of expert in this area and the below is oversimplified but the frictional force between the wheels and the ground = CoF x mass x gravity. If this force is greater than the force exerted by your motor on the wheel [torque at wheel / radius of wheel = force exerted] then the wheel will not turn and your drive motors will stall.

When I recommend decreasing wheel CoF and weight it is because this combination is decreasing the normal force on the robot's wheels. When you decrease the force resisting the wheel rotation such that it is less than the force attempting to turn the wheel at drivetrain stall then you are what people refer to as traction limited, meaning the drivetrain's wheels will lose traction and spin in place rather than stopping while the motors continue to try to turn.

Originally Posted by Thad House

We were 100 lbs and VersaWheel DT's at IRI, yet we still popped. If you are geared high enough to not be traction limited, Stall is Stall, so weight and CoF don't make much of a difference, other then change where traction limited is. We popped because we pushed, not during accelerating.

You are exactly correct. The tips I shared are not magical, they are only directionally correct. There are other factors to consider which will effect these equations: efficiency, gear ratio, bumpers, etc. For example, when pushing another robot sometimes bumpers deform such that their weight is shifted onto your robot increasing your robot's effective mass. Sometimes the robot rocks back onto two wheels splitting the weight of the robot over only 2 wheels rather than 4. These types of things can be difficult to anticipate. I don't know your drivetrain situation (or if you are normally traction limited) so I can't comment on what might have contributed to your breaker tripping.

I do want to clarify something though. Popping the main breaker is temperature based and the cumulative result of robot actions over the course of the match. Quickly accelerating from standstill and forward / reverse slams are normal robot actions that pull large amounts of current and cumulatively heat the main breaker. This is where lower CoF wheels can provide small cumulative temperature gains (the wheel CoF determines how easily the wheel spins in place during take-off acting as a mechanical dampener mitigating the otherwise extremely high current spikes.) Using something like Omni wheels to minimize frictional losses when turning can also make a big difference. An inefficient collector or generous use of motors for auxiliary non-drivetrain actions can also be a huge factor that is often overlooked. Of course, stalling 6 CIMs will heat the breakers faster than any other individual action, and is generally the last thing that was happening before the breaker trips, and so is most frequently looked at. Pushing can often be the last large rock on the camel's back, but you can't forget all the straw you heaped up their earlier.

Hopefully this clarifies some of my thoughts on the subject.
Cheers, Bryan

[Joe Ross]

Originally Posted by jam2014

On the upside, there should be huge opportunities with the new 2014 Power Distribution Board on board current monitors to detect and manage excessive power drain.

The 2014 PDB does not have built in current monitoring. The 2015 PDP does.

Originally Posted by jam2014

I could be wrong but last I heard, it was only on the 40amp outputs ?

Where did you hear that? I don't believe the PDP ever had that limitation, and it certainly doesn't now.

Originally Posted by jam2014

I am looking for some real world feedback on teams experiences using a 6 cim drive train with a 2 speed transmission designed for both low speed (4-5 ft/sec) and high speed (say 14ft/sec).

If your robot got in an extended pushing / blocking match with other 6 cim robots for a good part of a match...

- Would you be at risk of blowing the breaker ? If so, how often did you blow your breaker last year. Did you replace the breaker one or more times last year. Have you implemented any software protection (reduce power after X seconds) or hardware (fan) to defend against that ?

- how warm or hot would the following be by the end of the match
- the battery
- the terminals on the battery
- the contacts of the battery connector
- the breaker, its terminals etc

- do you guys do anything to cool the cims (fans, aerosol can cooling for example)

I have heard of a team that replaced all the 14 gauge cim wires with 10g wire. Did you guys do anything like that ?

We ran a 6 CIM drive train, 2 speed, with theoretical speeds of ~18 ft/s and 8 ft/s. We used 8 wheel drive with versawheels. We never blew the main breaker. We played plenty of defense towards the end of the season.

We used 10 awg wire on each of the drive motors, as well as kept the wires short (PDB was mounted very near the motors). We used 4 awg wire from the battery to the breaker and PDB, but used standard 6 awg on the battery. We did not implement any software magic, but the driver new to shift to low when pushing.

Using the 2015 control system, we collected current during a match at the SCRRF fall classic. This was with rookie drivers. See http://www.chiefdelphi.com/forums/sh...&postcount=153

[JesseK]

Man, this discussion .... if only it happened a few weeks from now rather than today. I might have had something that could show exactly (ish) where that fine line of traction limited vs breaker popping is.

EtherSim is coming...