Is 4 neos enough for a west coat drive?

This (and traction of different wheels in general) is probably one of the most requested experiments of all time and I think a search on this forum may yield some people’s findings. Notably, I think there’s some CIM vs Neo pushing tests from when the Neo just came out, which would more or less be the same test as what you’re proposing in that it’s the same chassis just more power.

A possible thing that people overlook sometimes is that while the total friction force is given by F = N*mu where mu is the coefficient of friction, mu is not always constant. We normally assume that mu = 0.9-1.1 is a constant value (this is usually the case for static coefficient of friction), however in reality dynamic coefficient of friction mu_k is often a function of the speed between the two surfaces, which would be the amount of wheel slip. Thus, the more wheelslip, the greater the amount of traction.

tldr: this is a very technical way of explaining what you’re suggesting and why it’s often overlooked in robot design

2 Likes

Yes. We ran a 4 neo WCD in 2019 and it took us to semis in our world champs division.

5 Likes

IMO the best WCD of 2019 was build by 1796. It uses six 6" Colsons and two 6" dual omnis in an 8WD west coast set up, and is powered by four Neos + two miniCIMs.

Hands down the best acceleration of any robot in the Carson division (Detroit). First overall draft pick.

Without the need for clearance, due to high centering risk at the corners of HAB level 1, 6" wheels would not have been required and 1796 could have been just as zippy without the miniCIMs.

3 Likes

We actually ran 4 Neos in 2019 and we could push pretty well. I can’t back this up with any math, but 4 seemed like plenty for that robot. It also may have helped that is was heavy, and had shifting gearboxes.

1 Like

Has anyone built a 4 NEO single speed and thought “you know, this acceleration isn’t zippy enough”? I don’t think so. NEOs scream. They accelerate fast enough to tip over top heavy robots sometimes. I really would want to see specific math to support a drive with more than 4 NEOs in it; I think if you see it it’s basically just out of habit or designer rationalization.

2 Likes

I really do suggest that, no matter what motors your team chooses to run, the drivetrain is tested extensively.

This season, my team tried to run NEOs with a custom drive train for the very first time. Our drive train was ‘a little bit untested’ going into our first event, although it did run fine once or twice. At the event, we were having issues with electrical from the start. One SPARK Max was blinking the wrong colours, and we ended up redoing almost the entire electrical during lunch on one of the days. With enormous help from FTAs and teams, 1114 and 5024 specifically, we were diagnosing and fixing the electrical system. Omar Zrien from CTRE was there and with his scientific equipment it was certain that one of the NEOs was broken. Whether it was a problem with the electrical system or a factory defect, I am not entirely sure. But this shows how crucial it is to test your drivetrain well before the event. It is worth noting that the robot’s electrical system was also put together rather quickly.

Keep in mind that NEOs are effectively “NEO can be a drop-in replacement for CIM-style motors” (REV). If they just do not work, you can always replace them with a CIM or a mini-CIM and still use the same motor controller, the SPARK MAX.

So design your WCD with possibility of replacing NEOs for CIMs later on. This is something that my team did not really do, we just left enough clearance for the gearbox, the battery, and the NEOs themselves. When we decided to replace them all for CIMs, there wasn’t enough space and so the battery was moved a bit higher up, and we actually ran a CIM and a mini-CIM per gearbox in a puzzle-like pattern. Gladly, it worked amazing in the end!

As a side note, for drivetrain design, or the design of any other mechanism, there are very useful resources that will help you. Start with the JVN Calculator. It will tell you that, with reasonable gearing, 4 NEOs will be enough for a WCD!

2 Likes

While I haven’t seen personally worked with Neo’s, just Falcon500’s, I can definitely say 4 is enough. If not careful, 6 can be more detrimental than 4. We found running 6 at full power could easily pull more than the rating for standard battery connectors, and we physically melted a few before realizing this. You definitely don’t want to use the mentality, “well, everyone else is doing it, so we should” in this case without doing plenty of testing first.

That’s a bit surprising. The SB-50 battery connector is rated for about 175A 120A [edit - less than I thought!]. While 6 Neo/Falcon500/Cims will definitely draw more than 120/6 ~= 20A ea at stall, once you pass ~200A total, one of two things will usually happen:

  • the battery voltage will drop below 6.3V and you’ll brown out (assuming battery + system resistance ~ 0.03 ohms)
  • the 120A main breaker will (eventually) trip (depending on heat up time).

If your battery connectors were melting, I suspect you might have had an issue with the way contact was made inside. Perhaps the crimp wasn’t great, or the contact wasn’t fully installed.

Regardless, current limiting is mandatory when using such high power motors!

3 Likes

One thing I will note is that while a NEO can be a drop-in replacement for a CIM, a CIM is not always a drop-in replacement for a NEO. The NEO and Falcon 500 are 60mm in diameter, whereas the CIM and miniCIM are 64mm diameter. Our 2020 robot uses an AM14U3 kitbot with some structural pieces that would intersect a CIM-style motor but had just a bit of clearance with the Falcon 500s we had.

3 Likes

Could you provide a source for the rating being 175A? I’m not trying to call you out in anyway, but I seem to remember us finding it was 50A. The SB-120 sold by Andymark specifically states on their website that it’s for higher current systems. If the mainbreaker fails before the connector, I find it odd Andymark would provide a product like that.

2 Likes

I’m wrong. The connector is rated for 50A for a 30deg C temperature rise, and 120A to “not to exceed the maximum operating temperature”. FRC robots clearly exceed that limit regularly, but usually for short enough periods to let it cool again - especially with current limiting!

Still, it makes me reconsider joining teams that switch to SB120 connectors.

3 Likes

Our driver had to keep the throttle below 0.5 at all times on our 2020 robot because it would accelerate too fast and tip over… This was with 4 neos…

3 Likes

We used the KOP chassis with one NEO each side for programming practice. Worked great, granted it pretty light with only the chassis, electronics board, and battery.

1 Like

We probably need to do some real world 4 vs 6 NEO tests. We’ve only used 6 since 1720 implemented them before last year’s IRI when we played defense primarily. They really could push just about anyone and we were really impressed by their performance during that event.

We stuck with the same setup this season and had equally good success pushing defenders out of the way as needed with a single speed kitbot drive. With the Thriftybot 3 motor plate it’s just almost too easy not to throw a 3rd motor on. We’ve never had current issues either even in heavy pushing matches.

2 Likes

One thing about Neos that frankly sucks are the little screws, make tiny little Screws, scratch that make that crappy little screws that hold on the outer housing. That is really not a great solution. Very disappointing for an otherwise massive revolution in motors.

4 is a fine but careful with the housing mount screws!

1 Like

Did your gearbox require removing it to press on a pinion, or was this an issue with NEOs that never had them removed?

1 Like

Got it, the screws are not fun. I agree. Especially the ones that are secured with Loctite.

To their credit, REV did publish guidance on removing the tiny screws. If they get a chance, perhaps the next iteration will use larger ones? M3 would be much easier to handle.

We have found that the M2.5 screws can be removed reliably using a new, sharp 1.5 mm Allen wrench, and some careful attention.

2 Likes

What ratios and wheel size were you using in that set-up? Pure kit-bot, or did you opt for a different wheel or ratio kit? Were you traction limited with 6 NEOs? Would you have been traction limited with 4?

1 Like

Granted, guidance was provided; however, when the proper tool is used carefully and the tool yields under torque, does this still sound like a wise setup?

A better method might have been to avoid any OEM thread locker and caution the teams that they need to apply thread locker prior to finally assembly.

2 Likes

This would have almost certainly led to far more failures than the failures caused by careless students stripping a tiny screw. A huge chunk of teams would miss this advice or not do it.

3 Likes