Alternatives for Signal and Pneumatic Connectors?

[nuclearnerd]

Our team had a heck of a time with push-in pneumatic fittings and dupont 0.1" pwm connectors coming loose this year. We're looking for more robust solutions in 2017.

For Pneumatic fittings, we're considering using barbed fittings, similar to these https://www.mcmaster.com/#barbed-fittings/=151woyg, instead of push-to-connect fittings. Barbed fittings will be a pain to remove, but at least they shouldn't come off unintentionally. I don't know why we had more problems with the p-t-c fittings than other teams (it's not because we mixed 1/4" with 6mm I know), but at least we weren't the only ones: https://www.chiefdelphi.com/forums/s...hreadid=152305

For signal wire, the PWM connectors are a PITA to crimp, but the bigger problem was that they had a tendency to work their way out of the roborio headers. Next year we're considering having a printed circuit board made which would adapt the roborio header pins into a more robust connector, perhaps wago, screw terminals, or an RJ11 jack.

We had great success with Anderson connectors for power wiring, and will continue using them in 2017.

Has anyone here played with other pneumatic and signal connectors? Can you share any OpEx (Nuclear industry term, short for "Operational Experience") with us?

[adciv]

We use PWM from Hansen Hobbies. We've never had them come out of the roboRIO but we also secure them to structure nearby. When doing a PWM-PWM connection, we electrical tape the two together. Can you post a picture of your wiring?

Anderson Power Poles had two issues this past year. First, we had some of the SB50s wearing out and need replacement, leading to a loose battery connector which bit us. We also had one PP15/45 Power Pole come loose and are considering zipties for next year.

[jwfoss]

I am curious what brand and model of push-to-connector and tubing you are using that has given you problems. There are a number of different manufactures for each. Tubing also comes in different hardness and thicknesses.

Typically when we have a failed connection it is due to the following reasons:
1. The hose was ripped out by a non-normal action/interaction
2. The hose was not properly installed
3. The hose was not cut straight
4. The hose was not re-cut after several installations.

Number 4 is important because the tubing deforms with every disconnection and re-connection. It is standard practice for us to re-cut the hose before re-install.

[nuclearnerd]

Originally Posted by adciv

We use PWM from Hansen Hobbies. We've never had them come out of the roboRIO but we also secure them to structure nearby. When doing a PWM-PWM connection, we electrical tape the two together. Can you post a picture of your wiring?

Here's a mid-season picture of our wiring. It's pretty cluttered due to our narrow belly-pan and staggered gearboxes, and typical frc hurry. We also have to come up with a better solution for CanBus wire junctions

The PWM header connectors worked their way out fairly often (can't remember if the loose one in the picture was intentional or not). The 1 position housings were especially prone to coming off, and we stopped using them after our first competition, but the 3 position ones were still a problem.

[Nuttyman54]

It's not cheap, but the WCP Spartan Board is a great solution (IMHO) for the RoboRIO connectors. It offers a variety of nice features, but one of the best is that it changes all pinouts to Molex connectors with detents. We used it this season and never had any issues with wires coming lose on the RIO. It doesn't solve the PITA crimp problem, but it's an improvement.

I know you guys work closely with some super sweet 3D printing, I've seen a handful of solutions floating around of 3D printed attachments which help retain PWMs in the RIO. I don't have a link to any but I'm sure someone will post one. I seem to recall 111 Wildstang put a version for download up somwhere.

As for pneumatics, I've always used the push-to-connect fittings so I can't recommend any alternatives, but I can say that it matters a LOT if the tubes are cut flush 90*. This is the most frequent failure case for the push to connect fittings as far as leaks or staying engaged. If there's any angle at all on the end of the tubes they won't seal properly which can lead to them coming out.

[nuclearnerd]

Originally Posted by jwfoss

Typically when we have a failed connection it is due to the following reasons:
1. The hose was ripped out by a non-normal action/interaction
2. The hose was not properly installed
3. The hose was not cut straight
4. The hose was not re-cut after several installations.

Number 4 is important because the tubing deforms with every disconnection and re-connection. It is standard practice for us to re-cut the hose before re-install.

Our tubing and fittings are pretty much all the AndyMark stuff (not sure what brand). We have a few from McMaster Carr - I'll have to check which

I'm sure 1 through 3 were the causes of many failures, but I hadn't considered #4, thanks!

I was wondering if barbed fittings might be a more "fault tolerant" solution.

[nickbrickmaster]

For your PWM-RIO connections, I recommend WildStang's Cable Magic print (If you have a 3D printer of course.) They'll keep your wires connected and save your RIO from stray shavings. We've used them for both years with great success.

[FrankJ]

The trouble with barbed connectors and 1/4" tube is that you are restricting already small diameter tubing. Push to connect fittings are the least restrictive. Probably the best bet is death to all 6 mm tubing in your shop and strict quality control on cutting tubing. If you are a diverse shop and cannot bear to get rid of you 6 mm tubing at least make it a different color than the 1/4 inch.

The PWM cable are bigger problem. You are stuck with them at the roborio. Sticking with commercial cables will solve the crimping issue. Making sure there is no strain on the connection. Wire tie the cable down so the strain is pushing into the connection.

[adciv]

Originally Posted by nuclearnerd

Here's a mid-season picture of our wiring. It's pretty cluttered due to our narrow belly-pan and staggered gearboxes, and typical frc hurry. We also have to come up with a better solution for CanBus wire junctions

The PWM header connectors worked their way out fairly often (can't remember if the loose one in the picture was intentional or not). The 1 position housings were especially prone to coming off, and we stopped using them after our first competition, but the 3 position ones were still a problem.

Looking at your wiring, I'm going to take a guess here. Do you reuse pre-made wiring from year to year? Are you interested in being able to make your own wiring?

Related comment, we use 3x1 PWM for canbus connections and electrical tape them together. No issue yet. We've also not had an issue with PWM coming out of the RIO, I suspect its the wiring you're using. I also think the issue is related to unsecured wiring. The mass of cable is probably causing the PWM headers to come off the RIO.

[Wren Hensgen]

Good suggestions. I'd also make sure that you buy a tubing cutter, which will ensure that the cuts you make are nice and square. Don't try and do it with scissors, or a flush cutter.

Also, just make sure that the fittings are pressed all the way down. Some people don't get them all the way in. Unlike a lot of other connectors, they seat under pressure, which means that if they're leaking, you can lean down on them, and they should stop.

[Tom Line]

Match Marking and quality control can help you.

We've had problems with tubing in the past - some tubing is harder than others, some softer, and it all depends on what supplier you purchased it from. We've thrown out batches of tubing from certain suppliers.

Back to the marking. I would have someone be put in charge of checking ever single connection on the robot. Have them pull (HARD) on every pneumatic tubing connection. If it passes put a mark on it with a green paint pen.

Likewise, do the same with electrical connections. All of our electrical connections get hot glued to the robo-rio. Our biggest problems have come from in-line splices, especially where we have used 2 pin servo connectors. Even though we tape them, they can be finicky if you are doing the crimping and assembly of the connectors yourself.

[nuclearnerd]

I've heard a lot of these rules of thumb for getting push to connect fittings to hold reliably. We've worked at trying to cut tubing square, pushing the tube all the way into fittings, and throwing out any batch of tube we thought was failing repeatedly. There's probably more we can do for quality assurance, but at some point I have to step back and think "there has to be a better way".

If barbed fittings reduce the id to much, another option is compression fittings with support inserts. https://www.mcmaster.com/#compressio...ubing/=151zg0i These should be more fault tolerant than p-t-c fittings, and more easily removed than barbed fittings. But it would strip be a pain to work with two wrenches in the confines of a robot chassis.

[Wren Hensgen]

Originally Posted by nuclearnerd

I've heard a lot of these rules of thumb for getting push to connect fittings to hold reliably. We've worked at trying to cut tubing square, pushing the tube all the way into fittings, and throwing out any batch of tube we thought was failing repeatedly. There's probably more we can do for quality assurance, but at some point I have to step back and think "there has to be a better way".

If barbed fittings reduce the id to much, another option is compression fittings with support inserts. https://www.mcmaster.com/#compressio...ubing/=151zg0i These should be more fault tolerant than p-t-c fittings, and more easily removed than barbed fittings. But it would strip be a pain to work with two wrenches in the confines of a robot chassis.

I currently work for a pneumatic distributor a job I got from my involvement with FRC. I can honestly say that the push-to-connects leak about one time in a hundred, and giving them a shove solves the problem. I've never seen one fail, or break, and we assemble them constantly, even re-using them. We stock and sell Koganei, Pneumadyne, and Norgren fittings, and we used to distribute Parker. None of them have had this kind of issue. These really are the best, and cleanest, and usually, the most reliable attachment method.

Who made the tubing that you use? If it's too soft, it can start to deform, and refuse to seat. I can recommend Freelin-Wade very highly.

[MrBasse]

How often are you reusing your ptc fittings? I used to work for Parker in their Brass Products Division. The teeth that grab onto the tubing is stamped from an almost nonexistently thin piece of metal. They are designed to be used just a few times, but it typically isn't an issue to reuse a dozen times or so.

If with new tubing and new fittings you push the tube in and can pull it out by hand you have three choices: 1. Throw away the tubing because it's junk 2. Throw away the fittings because they are junk 3. Both 1 and 2.

If you write a nice email to Parker they often will donate some fittings in a random grab bag kind of way. I was in the design department, but got a tour of the quality assurance area a few times. Our leak rate was more like one in a thousand rather than one in a hundred. And those were typically just because the orings got sheared by a crooked tube.

Edit to add: PTC fittings were developed for tractor trailer braking systems, if hose barbs were more reliable they would use them for that industry where air is used to stop the biggest vehicle on the road. Stick with the fittings you have now, maybe check new suppliers or brands if needed.

[R.C.]

Originally Posted by Nuttyman54

It's not cheap, but the WCP Spartan Board is a great solution (IMHO) for the RoboRIO connectors. It offers a variety of nice features, but one of the best is that it changes all pinouts to Molex connectors with detents. We used it this season and never had any issues with wires coming lose on the RIO. It doesn't solve the PITA crimp problem, but it's an improvement.

As a heads up, the spartan board will be dropping in price and including a gyro for under $120, we'll also be selling the mating connectors this year.