Care to elaborate on this point?
Has CTRE ever left anyone wanting or released an incomplete product?
I took a few years off FRC but can’t think of an example where they weren’t my preferred vendor.
The big software change two years ago comes to mind.
On paper performance the falcons have alot more going for them than the Neos. And seems vastly superior.
Do they, though? You’re not sustaining, and probably not even hitting, peak power. The efficiency gains exist, but aren’t nearly as big.
But then we would have had to completely ditch the last three years of programming experience our team has spent working with Talon SRXs and everything they can do. That’s the reason we (and a lot of other teams) didn’t jump ship to using Neos, because we didn’t want to have to also go to using SparkMaxes. It was too steep a hill to climb for us.
We used CIMs very effectively during an offseason event this year…
… as ballast in our defense-only NEO swervedrive bot. 70 lbs of CIMs were loaded in our Pushing Power™ cube bolted to the frame of the robot. Quite effective.
In all seriousness, CIMs are definitely still going to be a valuable motor for a lot of teams. We are not planning to use them any time soon (but will will be hanging on to our miniCIMs for the time being). But they are definitely an affordable durable fool-proof motor option that I believe will still be a competitive solution for many teams for a long time to come.
The graphs above show a nominal gain in time / throttle response, and a nominal (at best) gain in power efficiency for a standard short sprint. Even when geared for the same exact speed for a 2/3 field sprint the Falcon only comes out ahead by less than a robot length across the entire length of the sprint. Why isn’t stupidly, astoundingly faster? The Falcon is current-limited for nearly a full second on a 40-foot sprint.
There are some major advantages that the Falcon has over the NEO for a drive train, but power/efficiency just isn’t one of them when it comes to the constraints on FRC (even when using NEO data provided by VEX).
The efficiency helps when it comes to single-motor mechanisms. Yet when scaling to multi-motor mechanisms (for power density) even the CIMs can come out ahead on everything but weight depending on the mechanism constraints and cost. For understanding who to trust when it comes to which considerations matter, you really have to understand who is financially connected to IFI/VEX, REV, and who is not financially connected at all. The Falcon release thread is like night and day in that respect .
Yes, CIMs can still compete. For >80% of teams, CIMs are still a great drivetrain workhorse.
Can I heart a post twice?
They do beat the NEO on power/efficiency specs within the FRC design space, but by narrow margins that don’t matter much. Very well done engineering to achieve that, and I don’t mind them trumpeting it in their marketing.
(My excitement for the product comes from having NEO-level performance come from a single integrated package implemented on the CTRE programming framework.)
For what it’s worth, we did last year with a programming team of two and it really only took them a week or two to figure out the NEO and SparkMax. You definitely could have put NEOs on your drivetrain, and it would probably have been less painful than you think. Lots of teams made the switch halfway through the 2019 season and went through the same learning curve. Not a cake-walk, but not impossible.
The ~780W peak power is a crazy number, but it’s not practically attainable. (I assume that’s what you’re talking about with the on-paper performance comment.) Current draw at that point is 120A. The practical differences between the Falcon and the NEO (as I see it) are:
- A nominal improvement in efficiency. This means a little more power output for each watt put into it. They’re pretty close though.
- Built-in controller. This means fewer points of failure, but the failures you do have will be more expensive.
- Higher free speed.
- Cooling port, for those who will use it. (Can the Talon FX read the motor temperature? It would be cool to have a solenoid that can trigger to automatically cool your motor.)
- Spline shaft. Incredible feature or expensive gimmick? Time will tell.
Arguably the more interesting feature is a user replaceable shaft… I would love to see multiple options here.
Some of the Sparks were SparkMax running with PWM, so it’s even higher then 18.6%.
I believe I counted those (25 teams), since the data had a SparkMax PWM column. However if the PWM teams created things differently in code, they’d be missed.
Supposedly the lack of a key is one of the reasons an 8T pinion is offered for this shaft. That 8T pinion allows for a single stage drive gearbox for 6" wheels (if I did the math right) with an 8:84 reduction that pulls about as much current as CIMs do when geared to 13FPS free. I posted the JVN calc results here:
Honestly this may be the differentiator compared to NEOs for us if this is reasonable to run.
In the spirit of Car Nack’s predictions - I’ll very confidently say that every single alliance on Einstein at both post season expos (or as FIRST calls them, championship events) will contain robots that use CIM or MiniCIM motors as the primary motors in their drivetrains.
The winning alliance at both expos will contain CIM driven robots.
These motors are cool, but for the vast majority of teams that extra couple percent efficiency or the lighter weight is not the primary constraint on their robot performance.
The Mini CIM line is a cop-out (even we prefer Mini CIMs, and nobody would dare declare the Pandamaniacs an elite team). How many are putting the big ones on to cross the endless sands, O great and wise one?
I can get behind this and hope that it is true!
I don’t see them as substantially different given they can be fairly easily swapped with a minor cost compared to swapping to Falcons or Neos.
(I don’t think this is equivalent but included for completeness)
4 CIM => 4 mCIM : $120
4 mCIM @ 29.99 => $120
4 CIM => 6 mCIM : $280
6 mCIM @ 29.99 => $180
2 VictorSPX @ $49.99 => $100
4 CIM => 4 Neo: $460
4 Neos @ $39.99 => $160
4 Spark Maxes @ $75.00 => $300
4 CIM => 4 Falcon : $560
4 Falcon 500 @ $140 => $560
(I don’t think this is equivalent but included for completeness)
4CIM => 2 Falcon : $280
2 Falcon 500 @ $140 => $280
3452 hasnt used CIMs since 2017, and I dont think has any intentions on going back. 775s have been replacements for us, lighter, and as long designed for properly, can be better than CIMs. The Falcon motors I think are gonna be huge, with more power and most likely better control than NEOs. I dont see any reason to go back to CIMs really.
Maybe some team just used the normal Spark class in code and never bothered to change it. I wonder what team did that.