I might be missing something with a 6 cim drive...
 

So I was thinking about main breaker trips with 6 cim drive trains, and I was thinking about how to avoid that.
But because torque is directly proportional to current, and the motors are all cims, for a given current (say, 120a) wouldn't the toruqe output be the same regardless of the number of cims?

Basically, if the amount of current was limited per cim, through the 40a breakers only, then 6 cim drive would have more max power. But if the power of the system is limited to 120a total, then how is it advantageous to use 6 cims instead of 4?

Right now I think it's because the breakers need a few seconds to trip and efficiency gains, but that seems like a small gain to me for the extra space and weight.

Re: I might be missing something with a 6 cim drive...
 

Why do you believe that the system is limited to 120A? The main breaker is guaranteed NOT to trip until the current is ABOVE 120A. Even at 240A that main breaker will not trip in under 45 seconds.

Also, a 6 CIM drivetrain pushing the same as a 4 CIM drivetrain will draw less current per CIM, thereby possibly preventing the 40A snap action breakers from tripping.

The real difference between 6 CIM and 4 CIM drivetrains is acceleration. Assuming both drivetrains are geared for the same max speed, the 6 CIM drivetrain will accelerate faster. If the both drivetrains are stalled, the 6 CIM drivetrain will pop the main breaker faster as well. Such is tradeoffs in engineering.

Re: I might be missing something with a 6 cim drive...
 

The 120a Breaker is capable of instantaneous currents of over 720 amps before tripping (see attachment), and 240 amps for a considerable amount of time (for a FRC match). While not breaking the easier-to-trip 40 amp fuses, you could run much higher currents than a 4 CIM drive, which would be limited to 160 amps, for some seconds.

Re: I might be missing something with a 6 cim drive...
 

What spec are you looking at ?

Re: I might be missing something with a 6 cim drive...
 

This is from a large PDF from the manufacturer. I clipped this image out a while ago for the breaker we use. I will try to find the source


EDIT: Page 34 of this pdf

Re: I might be missing something with a 6 cim drive...
 

That was 'off the top of my head'.
It's corrected now :eek:

Re: I might be missing something with a 6 cim drive...
 

Ah, okay, that makes sense. I thought the breaker would trip faster than that, but great! We're going to be designing for a 6 cim drive then.

Re: I might be missing something with a 6 cim drive...
 

Just a word of caution: If you stall a CIM it pulls ~133A. If you stall a 6 CIM drive, it'll pull somewhere between 400-500% of the rated current for the main breaker (current being limited to the batteries internal resistance, I don't have that number in front of me right now). That will trip the main breaker in 1-6 seconds (as per the chart above). The 40A snap action breakers will be at ~ 250% of rated current and may take longer than the main breaker to pop (again I don't have the spec sheet in front of me) and that's not accounting for any other electrical load you have at that moment.

Basically, ensure your 6 CIM drivetrain is never in stall (make sure it's traction limited, not torque limited).

Re: I might be missing something with a 6 cim drive...
 

If each CIM were drawing 133A, a 6 CIM drive would pull 800 amps.
In addition to frying all 6 motor controllers, the plastic PDB would be molten as well. Your wires would be scalding to the touch, and there would be a large quantity of gaseous substance emitting from the robot. Basically, yes it is a bad thing, but no, it will practically never happen. This is really not something to worry about.

We ran a 6 CIM VexPro ballshifter running 3" wheels, and geared to 5.5fps and 15.5fps. In addition to bulldozing everyone we faced, we never blew a breaker once. That being said, we did have a "ramp up function" in the code. Basically, if you jammed the joystick into full forward, the code would automatically and linearly increase the PWM "throttle" very quickly (read: like 1/2 second) from 0 to 100%. Even when throwing the robot from full forward into full reverse, we never had a breaker-blowing situation. The acceleration time is negligible compared to instant-100% because the ramp up is so quick and you have the comfort of knowing whatever you could possibly do to that joystick will not disable the drivetrain. The ramp up has the advantage of suppressing the instantaneous current draw into a slightly longer duration but lower peak draw.

Re: I might be missing something with a 6 cim drive...
 

When testing this year, we found that a new breaker can take 6 stalled CIMs and a compressor (~250 amps) for about 25 seconds before tripping. Note that this is to a large extent cumulative over the course of a match, because these breakers are thermal and they don't dump heat instantly (or even particularly quickly); that is, if you stall for 20 seconds early on, you're going to be in real trouble if you get into another pushing match 30 seconds later.

Re: I might be missing something with a 6 cim drive...
 

Actually, I heard 254 stores covers breaker in ice packs or something right before the match. Not sure how true that is.

Of course if we go 6 cim we would have a shfting gearbox. That way we would have pushing power as well.

Re: I might be missing something with a 6 cim drive...
 

Quick question, how would you stall a 6 cim drivetrain in normal usage? The wheels would lift at some point, even driving into a wall, right?
Of course, we would still build in current limitations and preferably shift.

Re: I might be missing something with a 6 cim drive...
 

Get into a pushing match with a robot, if the traction of your wheels exceeds the torque created by your robot then the motors will be in stall. (Assuming all wheels are on the ground)

Even with 2 or 4 wheels off the ground, it's possible for your wheel traction to still be greater than the torque of a 6CIM drivebase (However, unlikely)

Keep in mind that in a 6wd configuration, there are only 4 wheels in contact with the ground at any given time. It remains the same for 8wd, with only 4 wheels ever actively touching the ground (assuming center two wheels are dropped)

Re: I might be missing something with a 6 cim drive...
 

Fun fact: 6 CIMs at stall consume a total of roughly 9576 watts of electrical power, which is over 9000!


Seriously though, a robot with 6 CIMs under "normal load" output (excluding friction) a total of 2.6 Horsepower, while a 4 Cim drive does just 1.75. At exactly 40 Amps per motor the numbers are (theoretically) 3.86hp and 2.57hp, respectively.

And as everybody knows, power/weight ratio is the most important figure. :rolleyes:

Re: I might be missing something with a 6 cim drive...
 

That's why we ran 6 CIMs and 2 MiniCIMs this year. More power is always better. :)

Re: I might be missing something with a 6 cim drive...
 

So have people stalled a 6 cim drive before then?

Re: I might be missing something with a 6 cim drive...
 

Yes.

With a 125lb robot (including battery and bumpers) geared 6.1:1 with 4'' HiGrip wheels, we stalled 6CIMs when simultaneously running the compressor on a not-quite-full battery, resulting in a current draw of ~250 amps.

With the compressor off and the battery completely full, the same robot spun the wheels and pulled ~180 amps.

Re: I might be missing something with a 6 cim drive...
 

But CIMs stall at 133 amps apiece...? (theoretically as always) their manufacturing specs say so. that would be 800 amps, aka very bad...

Re: I might be missing something with a 6 cim drive...
 

They stall at significantly less than that when you're running them off of a real battery whose effective voltage drops significantly when you start drawing lots of current from it.

It's the same reason that you can stall them at all in the first place; if you naively just calculate from the stated stall torque of the motor you'll find that your robot would have to weigh an ungodly amount to stall them at any gearing you'd see on an FRC bot.

Re: I might be missing something with a 6 cim drive...
 

True true. I thought about that but did not account for it.

As the resistance outside the battery decreases and approaches near zero (v=ir so 12=133r so r= ~.1ohms) (the motor stalls), the internal resistance inside the battery becomes a significant percentage of the overall resistance, so the potential difference beyond the two terminals is proportionally less than it were when the internal resistance was a small fraction of the overall resistance. Another point for Ampere's law.

Re: I might be missing something with a 6 cim drive...
 

I, too, failed to account for it in our design this year, with the result being that we had to swap out our gearkits (which was an excruciating process due to the design of our drive) before worlds.

That's the way you learn best, though.

Re: I might be missing something with a 6 cim drive...
 

Ooooookay, we'd better be careful then!

Re: I might be missing something with a 6 cim drive...
 

http://www.chiefdelphi.com/forums/sh...ad.php?t=89768

I don't know. You can see some pretty interesting design choices in FRC.

Re: I might be missing something with a 6 cim drive...
 

How does what you described relate to Ampere's law?

Re: I might be missing something with a 6 cim drive...
 

On our practice bot we drove the robot up to a cement wall and had a mentor apply his weight to the top of the bot upping the traction big time. I had the driver lay on the stick full power into the wall effectively stalling all six CIMs at 100%. Fresh battery, new CIM motors, talon motor controllers.

The main tripped in ~8 seconds and for each successive test it tripped sooner and sooner, I stopped after it was tripping in ~5 seconds.

We are geared 5.95:1 in the box and 1:1 (belts) to the 4" HiGrip wheels. (I did some speed tests using encoder logs and we were seeing 12-13ft/sec max robot speed on slightly worn wheels).

We never tripped a main breaker during a match on the competition bot and to my knowledge never tripped a breaker during an estimated 200 hours of driving the practice bot (same breaker I tortured above).



YMMV

Re: I might be missing something with a 6 cim drive...
 

After building and running robots with 6 motor drive for the past few years, I had never experienced an issue with blown main breakers. This year the team had no issues through 3 regional competitions and champs seeding matches until they found themselves in eliminations on Curie. At this level of competition we had the main breaker kick out near the end of four of our elimination matches. Luckly it did not impact the outcomes of the match as they were very close to the end but it just goes to show the stress on the equipment at that level of competition.

Now you've gotta understand our machine is running 6 CIMs (2 speed 18fps/6.5fps) on drive, 2 RS775s on the reloader, an RS550 on intake and the compressor, so it is a very hungry little thing- drivetrain excluded. We would never dream of running a battery that hadn't come right off the charger fully charged. Even a battery that was fully charged and settled overnight was noticeably sluggish in our machine. It was typical for the battery to be warm after our matches as well as all the 6AWG wiring. I wouldn't use this experience to say that 6 CIM draws too much power. There were other inefficiencies which could have been improved to mitigate the breaker trips. First off, you've got to get the power from the battery to the PD board. Solder all your 6AWG terminals (torch, flux, solder, heatshrink)! The mentor in charge of wiring our machine refused to follow my advice and insisted crimped connections were sufficient. During elims on Currie, each of the terminals was a hotspot capable of burning you at the end of the match- any heat generation is a loss of power so here is improvement #1. Secondly, all the 6AWG wiring was warm- In the future it would be prudent to use 4AWG including higher capacity Anderson connectors (not sure if this is legal as per rules though since there is wording specifying a part number for these to be used). Thirdly, the main breaker was cooking hot. We did try swapping it out for a brand new unit but this did not affect the performance. Instead we borrowed some computer duster from the Robonauts #118 (Thank you so much!) and were able to delay the trips by cooling the breaker between matches. I am not sure if the heat buildup was being generated within the breaker or if the poor connections with the terminals were sinking heat into it through the lugs.

Basically, you just have to be very conscious that a 6CIM robot is going to be hungry. You need to design your gearboxes at the mercy of the main breaker and battery. If you are using 6 CIM single speed, make sure the machine breaks traction before stalling the motors. If you are using 2 speed, then make sure you actually use it! Next time around I would insist on some sort of current sensing device combined with the feedback from an encoder on the driveline which would automatically kick the gearboxes to low gear if there existed a current increase past a certain limit for a certain motor speed for a set amount of time. I can't count how many times I had to yell at our driver "LOW GEAR!" when he found himself in a shoving match. Outside of this, make sure your electrical system is robust as the additional strain is hard on these components. Oversize the wiring if you can afford the weight so that you eliminate your losses due to heat.

Re: I might be missing something with a 6 cim drive...
 

Wow, thank you for the informative post!
I talked to one of our electrical team, and he says that it would be simple to add current sensing for shifting in addition to a manual shift.
I'm not sure if he'll go for soldering connections (our electrical team doesn't like solder) but if it's just the big wires it might be okay.

Re: I might be missing something with a 6 cim drive...
 

Keep in mind if you oversize the wiring, solder the connections, shorten wires, etc. you're increasing current draw, not decreasing it, which will just make your main breaker get hotter. I'd think your 6AWG wiring and anything that was getting hot was saving your main breaker from tripping sooner! If you follow through with your plans to oversize the wiring to 4AWG, solder connections, etc. you may find you're tripping that main breaker more often.

Re: I might be missing something with a 6 cim drive...
 

That might not actually end up being the case. Several main breaker trips I saw this past season were quite definitely attributable to hot wires warming up the breaker rather than the breaker itself reacting to true overcurrent.

Re: I might be missing something with a 6 cim drive...
 

That's interesting Alan - thanks for pointing that out. I hadn't considered that. Towards the end of the season our robot was blowing the main too, but I can't say how hot the wires were. It's only a 4 cim drive and Andy mentioned he's never had a problem blowing the main before, but I think this year we had a shorter path from battery to breaker to PDB than we ever had (inches, with one cable just 2 connectors butted together with no wire showing) so we could potentially supply more current through the main breaker than ever before.

Hmm, I wonder if this would warrant throwing some thermocouples on and seeing just what the temp profiles are on the breaker and on the cables leading to it while we stall it against a wall.

What did you see that would make the wires heat the breaker? How did they solve this?

Re: I might be missing something with a 6 cim drive...
 

In all but one case I'm thinking of, the obvious problem was a loose nut on one of the connections to the main breaker. The resulting high-resistance connection turned a lot of power into heat right at the breaker.

The underlying cause of the other case wasn't clear. The red wire from the breaker to the Power Distribution Board was too hot to touch for more than a few seconds, and the wire from the battery to the breaker was very warm. Because the black wire was still cool and there was no obvious indication of a short from the battery negative terminal to anything, I don't think it was a simple high current issue.