Vendors, please standardize

[TFM110]

Vendors, please work together and follow 1 standard

First of all, before I start getting bashed on, as my profile says, “Just to clarify, I’m not dumb, I am just learning”, but also I love what all vendors in FIRST have done and have been doing for the community.

However, I have started to see some components become unstandardized which is causing this xkcd meme to keep coming true, and I hate it.

Standards

There are a few things that I would like to say but I feel like there will be others in the community that would also like to say something, as I could miss some things, they can happily post in this thread.

First: MAXSpline vs SplineXL
Yes, MAXSpline was released last year while SplineXL was this year but there are compatibility issues with both

WCP 2023-24 | Product Release & Updates

It appears REV ION Products with the Max Spline will work on the SplineXL tube, But WCP products with the SplineXL will not work on Max Spline tube. [image]

WCP 2023-24 | Product Release & Updates

I was about to say this. Another picture for context: [image] The blue is WCP, the red is REV

WCP 2023-24 | Product Release & Updates

This is on REV: I understand protecting your IP, but I’m disappointed they felt the need to give themselves a monopoly on a tube and a hole the tube fits in.

There will certainly be compatibility issues that may come across but if there was just 1 standard to the spline, this would be solved. I really, really love what @R.C and WCP are doing with SplineXL by keeping it open and allowing other companies to build/design around SplineXL.

@R.C

Designing for Compatibility:

• Any team or company may build off SplineXL, royalty free for competition or commercial use.
• If vendors would like to build around the SplineXL Interface, please shoot us an email and we can assist with tolerances/gauges, etc.

Second: Motor Shafts (Not really that big of a deal)
When the NEO was first released, it was marketed as a “Drop-in replacement for CIM-style motors” as they have the same shafts, which is amazing!
But then the Falcon 500 came out with their specialized shaft. Okay, fine, we just have to deal with it. It had its pros and cons but we went on. Now, the Kraken X60 has a completely different shaft (SplineXS), basically making all Falcon accessories useless, especially since the Falcons haven’t been in stock since the end of the 2023 season. It would have been nice if the Kraken X60 and the Falcon 500 shared the same spline but I know there was some stuff behind the scenes that possibly led to having to change the shafts, but still. We just have Falcon 500 accessories sitting, waiting to be used, which they probably won’t ever since we can’t order new Falcon 500s.

Third: Wires
For the longest time, we have had motor controllers have 12awg power wires and 22awg CAN wires, but now, the Kraken X60 comes with and is recommended to be used with 10awg for power and the SPARK Flex comes with 26awg CAN wires.

TDH, I don’t mind the 10awg wires for the Kraken X60 as the Anderson Powerpoles 1327 series (Which we all have been using in FRC for the longest time) PP45 connector supports up to 10awg. My only ick is that, as most teams slowly transition to inline wagos, they only support up to 12awg. I am glad WCP sells the PowerPole Adapter Board to use 10awg(or 12awg if you want) to just make your own long cable from the Kraken X60 to the PDP/PDH without having to use the cable that is provided and have a connector in the middle to extend which is an extra point of failure.

Now with the 26awg wires for CAN, I very much do mind. Almost all wire strippers that you can find only go down to 22awg, which includes automatic and manual wire stripers. If you want to use manual stripers, it probably won’t even strip the insulation, but if you use automatic strippers, that can break almost all the strands. I understand if it’s because of mechanical constraints but still.

REV Robotics - SPARK Flex and NEO Vortex

Due to mechanical constraints related to the comolded strain relief part and the internal clearances, we needed to go with 26 AWG wire. 26 AWG wire does in fact have less tensile strength than 24 or 22 AWG wire. However, we don’t believe it affects the usability or durability in a significant way, especially if following good wiring and strain relief practices.

You are giving teams compatibility issues as certain connectors teams use don’t go up to 26awg, so now, they have to buy new connectors.

REV Robotics - SPARK Flex and NEO Vortex

This is really unfortunate. Since most other devices use 22AWG or 24AWG for CAN, we stock 22-24AWG terminals (Molex 0016020102 and 0016020114). Now we’re going to have to stock a different terminal just for these motors (since there isn’t one that spans 22-26AWG as far as I can find), and train students carefully as to when they need to use which. It’s an opportunity for error that didn’t exist before. And if they do accidentally use the wrong terminal, it can result in the kind of subtle f…

Molex SL crimper in kop

Is the 26awg wire the CAN cables or a different cable? I haven’t gotten my hands on a Neo Flex yet to see myself. If 26awg is required, unfortunately a different crimper is needed then the one we have been providing. If you try to crimp 26awg wire with the 22-24awg terminals and crimp tool from the kit, it will not provide enough compaction and result in a loose crimp. To crimp to 26awg wire, you would use the terminals you linked (you have the correct parts linked) but would need the 24-30awg …

Fourth: Connectors
I mentioned about Andersons and Wagos in the wires section of this post. While both have their pros and cons, they are consumables and these are not the connectors I am worried about. I love using them. I would only recommend, as I mentioned earlier, to use 12awg to be able to use both interchangeably.
The connector I would like standardized is the DuPont connector on the CAN wires that are pre-crimped. Let me explain why.
For a while, I have been the type of person to use what is given to me. When I see a motor controller with DuPont connectors pre-crimped, and the wires are long enough to reach the other motor controllers with the connector, I will take advantage of that and use the retention clips, but every time I suggest just using the pre-crimped DuPont connectors to other people like on CD or the unofficial FRC discord, I get told its a bad connector and you should use a latching connector, like Molex SL. I am not going to waste precious build season time cutting the end of each CAN wire, then strip, and crimp a new terminal hoping they are good enough, especially with how small they are, mistakes are going to happen which will require students to cut and try to crimp again. But, if companies, like Molex, are willing to work with FIRST to get us a Molex SL crimper, or even work with companies like WCP to have the connector on the PowerPole Adapter Board and provide links in CD to the respective connector, I don’t see why companies can’t move to have Molex SL connectors for CAN.

Announcing Kraken X60 powered by Talon FX!

As @asid61 said, connector part numbers aren’t simple. We are working on the FIRST portal for our website though and there will be a short list of parts recommended for FRC there which should make things much easier. That’ll be up before the season starts. Design guide for FRC is also in the works but won’t be ready until after the season. In the meantime, here are the relevant parts for the SL connectors on the Kraken Adapter Board: SL Receptacle Housing with Latch, 2 circuit: 50579402 SL Re…

The only downside I see with using the Molex SL connector is if you want to use PWM control other than CAN cause the RoboRIO uses DuPont ends.

Overall, in my opinion, FIRST should be about getting students to learn, not having to be stuck in an ecosystem. Yes, each ecosystem has its benefits, but just imagine if everyone came together to make 1 universal product. Things would be so much easier, but of course, we, unfortunately, don’t live in that society.

[FrankJ]

@TFM110

The only downside I see with using the Molex SL connector is if you want to use PWM control other than CAN cause the RoboRIO uses DuPont ends.

Maybe we could get the rest of the world to switch back to the tried and true imperial system. /S

[Max_Morehead]

@TFM110

The only downside I see with using the Molex SL connector is if you want to use PWM control other than CAN cause the RoboRIO uses DuPont ends.

Aren’t Molex SL and “Dupont” connectors compatible? I can understand why some vendors would avoid Molex SL given their cost.

[Akash_Rastogi]

@TFM110

My only ick is that, as most teams slowly transition to inline wagos

I don’t know if this is accurate. We definitely don’t plan on switching over exclusively to inline Wagos. I definitely still like the PP45 way more for these size wires. The Wago are a nice to have for quick repairs so we do still keep them around. As you said though, 10awg isn’t required by anyone.

[cadandcookies]

I would additionally really appreciate it if there was a FIRST-enforced API specification that vendors were required to maintain compatibility with in order to have legal motor controllers. This seems to be a consistent complaint from our programming students whenever we try to mix and match motor controllers on a robot.

[calcmogul]

@cadandcookies

I would additionally really appreciate it if there was a FIRST-enforced API specification that vendors were required to maintain compatibility with in order to have legal motor controllers.

I’d like to see that too, but I don’t think it’s realistic.

API standardization

We used to have a common CAN API in the form of CANSpeedController (removed in 2017). The spec was way too big and got vendor complaints. In my opinion, the real problem is vendor motor controllers trying to do too much.

The API ends up being as big as WPILib’s current API surface area (aka huge) because vendors are basically repackaging a bunch of WPILib functionality (DIO, AIO, PID, quadrature encoders, duty cycle encoders, motion profiling, motor control).

You can either organize the API as a kitchen sink interface (CTRE), or as separate WPILib-esque objects for each feature (REV). The former has discoverability issues due to the sheer size of the API, and the latter has discoverability issues because features are spread out without a usage map.

For-profit parties also see APIs as a product differentiator, so they don’t have an incentive to standardize.

CAN frame standardization with open specification

This could have a smaller scope and let vendors innovate in the API space instead. Vendors probably wouldn’t like the limitation though, and it doesn’t address teams’s API divergence complaints at all.

Furthermore, this change would require significant firmware re-engineering for every CAN device. I could only see it happening for the 2027 control system refresh.

[AgentSmith451]

@Akash_Rastogi

I definitely still like the PP45 way more for these size wires

Agreed, in the last 2 seasons that our team has actually built a robot (2023 and 2020), we finally started switching TO the PP45’s from various other automotive style connectors (like Metri-pack). The Metri-packs are bulletproof when done correctly, but way overkill and easy to do incorrectly.

[EricH]

@TFM110

Overall, in my opinion, FIRST should be about getting students to learn, not having to be stuck in an ecosystem. Yes, each ecosystem has its benefits, but just imagine if everyone came together to make 1 universal product. Things would be so much easier, but of course, we, unfortunately, don’t live in that society.

You know, there’s kind of a funny thing in industry in general. Ecosystems exist, and are generally incompatible with each other except via carefully defined standards. Part of engineering is finding ways for a particular non-ecosystem part to interface with a particular ecosystem. And it is painful sometimes!

[bobbysq]

@TFM110

Now, the Kraken X60 has a completely different shaft (SplineXS), basically making all Falcon accessories useless, especially since the Falcons haven’t been in stock since the end of the 2023 season. It would have been nice if the Kraken X60 and the Falcon 500 shared the same spline but I know there was some stuff behind the scenes that possibly led to having to change the shafts, but still.

For what it’s worth, I expect SplineXS to be the definitive spline shaft going forward, unless Vex somehow puts Falcons back into production and manages to recapture the market. With the Neo Vortex having a swappable shaft, REV should be able to put out a SplineXS shaft and ensure its usage across vendors.

[TFM110]

@bobbysq

With the Neo Vortex having a swappable shaft, REV should be able to put out a SplineXS shaft and ensure its usage across vendors.

Ya. I forgot to mention about the NEO Vortex. I really love this feature and that REV is actively adding new shafts!

[kingc95]

Can we talk about Hex axles? I like 1/2 inch hex greatly, but dislike that it’s all not “Rounded Hex” by default. When we buy bulk hex axle stock at Andymark it’s a gamble if the bearings we have in shop were hex or “thunderhex” (we’ve gone back and forth or students order the wrong ones by mistake). Then 1/2 inch rounds won’t work for regular hex either for dead axles unless we machine them on the lathe. Can all hex axle stock be rounded by default. We don’t need the “tip”.

In a pinch Andymark churro stock also works when you run out of regular hex axles, still isn’t rounded enough though.

[TFM110]

@Akash_Rastogi

I definitely still like the PP45 way more for these size wires.

Me too! My team uses Wagos on our V1 robot and we use Andersons on our comp bot. We do use the WCP CAN inline wagos for CAN, but other than that, I try to make sure we use Andersons

[Tom_Line]

It’s a nice dream.

In some cases, the change was made to improve. The falcon spline allows for smaller gears to be used on the shaft as compared to the 8mm cim key, along with better torque transmission and no washing out or losing of keys. We love it.

In some cases, (and I’m guessing here…), the products are meant to be competing products on the market but have intentionally not used an identical feature due to patent/copyright/trademark, etc. I’m looking at the splines on the kraken, the spline tube shapes from WCP and REV, etc.

In some cases, the designs were engineering decisions for packaging etc (sparkmax JST connectors).

In every case, they have a valid reason for existing.

We can’t design a robot with the same size fasteners everywhere, even though it’d be great if we could keep everything common. But in some cases, there just isn’t a choice.

shrug. Engineering.

[Skyehawk]

@kingc95

We don’t need the “tip”

Could go back to 1/2" keyed shafts…

In reality I think there is still a place for regular 1/2" hex.

There is no appreciable amount of rounded hex in our shop. Rounded hex doesn’t interface with 3Dprinted items as well (all but requires an insert for any torque over a few ft*lbs)

Rounded hex definitely has its place, but I have found more than a few instances where I prefer the standard profile over the rounded (and where the rounded was used in place of the standard I have run into issues)

[Sidoti]

@TFM110

Overall, in my opinion, FIRST should be about getting students to learn, not having to be stuck in an ecosystem.

@TFM110

just imagine if everyone came together to make 1 universal product. Things would be so much easier, but of course, we, unfortunately, don’t live in that society.

Emphasis mine.

These two statements are vastly contradictory.
Do you want things to be easier or do you want students to learn?

When working as an industry professional, you may be tasked with using available products that are not in a nice and easy-to-assemble ecosystem. There are very few professions where you get to build your products or solutions out of LEGO.

While working at a robotics start up outside of Boston, nearly my entire job was to make non-standard parts coexist with each other. I was trained and prepared for these tasks because of my time in FIRST. Had I only done FLL and never gotten the open ended experience that FTC and FRC taught me, I would have been ill prepared for the ‘real world’. So FIRST served me well in my learning.

Post my time as a student I have made many projects. 3D printers, combat robots, small vehicles, rockets, drones, machine tools, etc. Not a single one of them was I greeted with a standard issued catalog of cohesive parts. Part of my success in those projects was my ability to adapt and react to these non-standard parts.

@TFM110

as my profile says, “Just to clarify, I’m not dumb, I am just learning”

You are not dumb. You are learning. Just the same as the rest of us. This is a learning opportunity and I do hope that this thread going forward does not bash you. As well, I hope you do not see my post as bashing as I am only trying to aid in the learning of anyone that decides to read this post.

@TFM110

unstandardized which is causing this xkcd meme to keep coming true

Different companies and suppliers not being standardized does not make this XKCD comic come true. If we saw REV have 4 different standards then we would see truth inside REV and REV alone. Same as if AM had multiple standards. Or WCP.

You also have to remember that all FRC suppliers are trying to make building robots easier while simultaneously COMPETING against each other. They are businesses. They need to make money to continue their goal of making more accessible robot parts. They do not make money if you buy other companies components. We live in a capitalist society. Regardless of our feelings on that, it is the truth.

We, the FRC community as a whole, would collectively suffer if every supplier was 100% compatible and standardized with each other. We would lose many of the great and advantageous products that we are now blessed with because these companies have made enough money to be able to design and sell these products.

Would it be nice if all of FRC was standardized? Sure. Would it be easier? Yes, absolutely. But I will end this rant with a quote.

“Nothing in the world is worth having or worth doing unless it means effort, pain, difficulty… I have never in my life envied a human being who led an easy life. I have envied a great many people who led difficult lives and led them well.”
― Theodore Roosevelt

This is, afterall, “the hardest fun you’ll ever have.”

[TFM110]

@Sidoti

You are not dumb. You are learning. Just the same as the rest of us. This is a learning opportunity and I do hope that this thread going forward does not bash you. As well, I hope you do not see my post as bashing as I am only trying to aid in the learning of anyone that decides to read this post.

Thank you for your insight! I really appreciate!

[asid61]

I desperately hope that FIRST does not turn into the barren landscape that is industrial automation devices in the name of “teaching students”. Students should be learning about engineering principles, not wasting weeks debugging vendor migration issues. Migrating from one vendor to another should be as easy as possible. If a vendor can only make money by locking teams into an ecosystem, perhaps they should focus on improving their products instead.

This kind of change, however, has to be driven by FIRST. Without the threat of making devices illegal, motors and other devices can come with any connector and API vendors choose. FIRST only has the legality lever to pull to control this right now.

[Dominick_Ferone]

We used to order hex through McMaster which is was nice and cheap but always got concerned with it not matching up with our bearings. One year we had to sand a ton of bearings to fit, not sure if the bearing was badly tolerances or the hex. But was something to cause concern as which standard was right.

[Jsteel]

On the rounded hex shaft - that is an item u pique to FIRST. The hex shaft with “tip” is what you get from any standard non-FIRST vendor (such as online metals, McMaster, etc). Standardizing on rounded hex just makes things more expensive and specialized.

Keep the tips!

[Thor5090]

@Sidoti

Post my time as a student I have made many projects. 3D printers, combat robots, small vehicles, rockets, drones, machine tools, etc. Not a single one of them was I greeted with a standard issued catalog of cohesive parts. Part of my success in those projects was my ability to adapt and react to these non-standard parts

If its anything electrical (in America at least)it is standardized under NEMA. You also used a standard of measurements. So trying to say you had to start from scratch with no standard for everything isn’t correct. What op is getting at is standardize some aspects like motor shafts and electrical connectors used but not everything. Low resource teams do not have the funds or equipment to make or have a stockpile of parts on hand for so many different standards.