Our team is playing with a practice drive, and we want to try the Grayhill encoders recommended by Vex on the ball shifter product page. DigiKey has them, but they don’t carry versions with wires coming off. So we need connectors as well. The spec sheet mentions a Molex part# 14-56-3056 connector. That connector uses 26 AWG wire.
What wire gauge do we need to use for encoder wires? The 2013 rules, to the best of my ability to interpret them, don’t specify a required wire gauge between sensors and a digital sidecar. Does that mean we’d be okay with 26 AWG under last year’s rules, or did I miss a rule?
The Molex connector is not in stock at DigiKey, so I’m looking at this connector to connect to the encoder’s pins.
I can’t think of any rules specifying a gauge for use with sensors like this. The only place wire gauges are specified is in R48 (2013), which is for wires hooked up to the PDB, not the DSC. Additionally, there is a stipulation stating:
Wires that **are recommended by the device manufacturer **or originally attached to legal devices are considered part of the device and by default legal. Such wires are exempt from R48.
Emphasis mine. So per the 2013 rules, I think you would be fine. Besides, have you seen the tiny wires that come with the KoP encoders from USDigital?
The answer might be different if you needed 12V to power them and had them hooked up to a breaker in the PDB.
And for what it’s worth, I’ve ordered a lot of stuff from newark.com when it hasn’t been available at digikey, and never had a problem. It looks like they have some of those connectors in stock, too.
I couldn’t look at the connector part you linked (get a “server error”), but the pins off these are standard 0.1" pitch headers. The same spacing that you have on the DSC or PWM connections on motor controllers.
We have used these encoders for the past 3 years. Making your own cables is easy, and I would recommend it. You’re not going to find a pre-made cable that’s broken out correctly to plug into the encoder and the DSC. So you’ll have to do some splicing at a minimum. If you have a crimper available to make your own PWM cables, you’re all set though.
In the past we’ve used the 20 or 22 gauge wire from Hansen Hobbies, but it’s a little large for the standard housings and it makes it difficult to insert the cables into the jackets. This year we’re trying out some 24 and 26 awg cable from them to see what works out. The only concern I have with the 26awg would be noise getting introduces on the signal cables, but this could be unfounded concern.
Note that the encoder has 5 pins, but only 4 are used, so you’ll need a 4 conductor cable. You only need to wire up all 4 if you need to determine the direction something is spinning. If you don’t need that information you can forgo wiring the “B” channel, and use a 3 conductor wire.
If you find a good source for 4+ conductor wires at 24-26awg, please share.
I missed that rule. Thanks for replying and pointing it out!
James,
Thanks for pointing me in the direction of PWM connectors and Hansen Hobbies. We already have stuff from them for making PWM cables, so using those 5 pin housings will be easy. (Apart from the fact that we need to do some training now that our resident PWM cable maker has graduated)
I ordered that 28 AWG 4 strand flat wire. That’s easy and inexpensive. My assumption is that I’ll have to find something in 24 or 26 AWG for a competition robot, but it’s surprisingly tricky to find 4 conductor wire.
Dan,
Jon has pointed to the correct rule. You can buy ribbon cable in higher conductor counts and simply peel off the unused conductors. i.e. buy 10 conductor and split in half for two five conductor cables.
I know I can get multi-conductor ribon cable from numerous sources. I’ve ordered a few different kinds in the past, but they’ve been less than ideal for one reason or another (e.g. insulation is too thick, not flexible enough, poor quality, too expensive).
I was hoping other teams who have gone through similar experiences and settled on specific piece parts that they routinely order, would be able to share their knowledge and save the rest of us from having to repeat the process.
I would expect a large number of teams order 24-26awg, 4+ conductor wire annually for sensor wiring. A reliable, inexpensive, source and part # would be a great piece of information for many teams (mine included).
You would be surprised… I’ve stuck my head into a lot of robots over the years, and I very rarely see multi-conductor cable used for anything other than the DSC ribbon cable and possibly CAN connections.
Our team has a stock of single-conductor wires (red, white, black, orange, green), and we simply twist them together to make whatever number of conductors we need. It’s pretty simply - get the wires cut to the correct length, clamp one end in a drill and hold on to the other. Run the drill until the wires are twisted together nicely. Twist a little past what you want, as they’ll untwist a little when you’re done. This also allows for some easy color coding for different sensors (especially limit switches, which only require 2 wires!), but usually not enough to rely solely on that for identification.
That’s certainly an option. I had considered it in the past and thought it would be tedious and difficult to keep the conductors together. I might give that a shot this year though on your recommendation. Thanks for sharing.
We use phone cable (the old flat kind) for most of our sensor wires. The stuff we have is 4 conductor, 26 gauge stranded. It’s cheap and readily available. The 4P4C phone connectors are easy to install if you have the right crimper, and you can use phone couplings to connect cables together (just make sure to get the straight through ones). It works for us anyway .
Stranded, shielded (STP) Ethernet cables](http://www.monoprice.com/Search?keyword=stp+stranded) aren’t a bad investment for sensors. Chop up a few patch cables, or buy it by the box to last you for many years.
Most of the custom circuit systems teams run in FRC don’t really care too much about signal integrity, but it’s a good habit as you implement more and more complex designs, especially if they operate at high frequencies, or can pick up stray RF energy over analogue lines.
By the way, with regard to the 2013 and prior custom circuit rules, I’m of the opinion that a custom circuit can (at the team’s option) include the wiring directly attached to the sensor or other device, all the way back to the non-custom robot part (like the sidecar). So the wire gauge rules aren’t necessarily applicable, unless there is a specific wire gauge called out for the particular type of device that you’re integrating through a custom circuit.
.