The Best Encoders for Drivetrain

[btcshields6]

Hi all!

I'm sure if this is the best place to post this, but as you can, I'd like to try and find the best encoders for drivetrain motors. Please, give me any thoughts and opinions you have on what to use, or tell me what your team does on their drivetrain.

Thanks,

Benjamin

[eddie12390]

We prefer Grayhill 63R over any of the other offerings that we've tried. We've had no issues with them skipping counts or anything like that and they are a lot more rugged than the ones that we've used from US Digital Encoders.

[Gregor]

Second the Greyhill 63R. We use them everywhere.

Worth the price. You can't scratch them and you can use them year after year.

[Lemon]

We've had great success with the analog versions of the US Digital MA3 encoders

[ice.berg]

We are using the Grayhill 63R encoders this year as we are using the vexpro gearboxes as well. So far after a ton of testing and one regional we have had no issues and are very pleased. They are so much easier to use compared the the common US digital E4P. They are available in a few different CPR options as well.

[Munchskull]

My favorite so far have been the AMT103-V Kit (http://www.digikey.com/product-detail/en/AMT103-V/102-1308-ND/827016) encoders. That said they new to us this year. however they are cheaper than most encoders, in addition they fit most shafts and are less temperamental than optical and mechanical encoders.

[z_beeblebrox]

My favorite so far have been the AMT103-V Kit (http://www.digikey.com/product-detail/en/AMT103-V/102-1308-ND/827016) encoders. That said they new to us this year. however they are cheaper than most encoders, in addition they fit most shafts and are less temperamental than optical and mechanical encoders.

We're also having success with those (and the 102 variant). 6 on our robot.

[jman4747]

My favorite so far have been the AMT103-V Kit (http://www.digikey.com/product-detail/en/AMT103-V/102-1308-ND/827016) encoders. That said they new to us this year. however they are cheaper than most encoders, in addition they fit most shafts and are less temperamental than optical and mechanical encoders.

Third. Very good encoders.

[Owen Busler]

For those of you using the Grayhill 63R, how do you connect your shafts to the encoder?

Just making sure im looking at the right one

http://www.futureelectronics.com/en/technologies/electromechanical/switches/encoder-switches/Pages/1613116-63R256.aspx?IM=0

[Oblarg]

For those of you using the Grayhill 63R, how do you connect your shafts to the encoder?

Just making sure im looking at the right one

http://www.futureelectronics.com/en/technologies/electromechanical/switches/encoder-switches/Pages/1613116-63R256.aspx?IM=0

Surgical tubing and zipties actually works very well.

[audietron]

For those of you using the Grayhill 63R, how do you connect your shafts to the encoder?

Just making sure im looking at the right one

http://www.futureelectronics.com/en/technologies/electromechanical/switches/encoder-switches/Pages/1613116-63R256.aspx?IM=0

I usually design the gear plate with a 1/2" hole in it for mounting. Then 3d print a pinion to press fit on to the .25" shaft and another to slide on to the 3/8 hex shaft (usually the first stage).

[SJaladi]

For those of you using the Grayhill 63R, how do you connect your shafts to the encoder?

Just making sure im looking at the right one

http://www.futureelectronics.com/en/technologies/electromechanical/switches/encoder-switches/Pages/1613116-63R256.aspx?IM=0

I can second the surgical tubing and zipties method. We used this method to successfully mate the encoder shaft to the drive gearbox output shaft. We also found a way to use surgical tubing to mate the encoder shaft to a 1/2" hex shaft. We inserted thinner surgical tubing which snugly fit over the encoder shaft inside of a thicker piece of surgical tubing which snugly fit over the 1/2" hex shaft. Then we just used a bunch of zipties to hold the whole thing together and so far its been working perfectly.

[BSV]

I can second the surgical tubing and zipties method. We used this method to successfully mate the encoder shaft to the drive gearbox output shaft. We also found a way to use surgical tubing to mate the encoder shaft to a 1/2" hex shaft. We inserted thinner surgical tubing which snugly fit over the encoder shaft inside of a thicker piece of surgical tubing which snugly fit over the 1/2" hex shaft. Then we just used a bunch of zipties to hold the whole thing together and so far its been working perfectly.

Dumb newbie question: So what would be the best way to add one (okay, actually two) of these to a kit bot? Any special mounting needed? Thanks!

[SJaladi]

Dumb newbie question: So what would be the best way to add one (okay, actually two) of these to a kit bot? Any special mounting needed? Thanks!

Looking at this drawing (http://files.andymark.com/am-0654-Toughbox%20Mini.pdf) you can see that there is a portion of the main output shaft sticking out of the back of the gearbox. You can slide a piece of surgical tubing over that portion of the shaft and then use a small ziptie to cinch it so that it stays on the shaft. Then you can fabricate some kind of simple plate to secure the Grayhill encoder to. It can be really simple we made ours with a small scrap of polycarbonate and a hand drill. then once the encoder is secure you can simple slide the other end of the surgical tubing over the encoder shaft and secure with a ziptie and you should be all set. It's also interesting to note that by using the surgical tubing method you don't have to make the shaft of the encoder and the gearbox exactly concentric. A few mm either way won't hurt the function of the encoder.

[BSV]

Looking at this drawing (http://files.andymark.com/am-0654-Toughbox%20Mini.pdf) you can see that there is a portion of the main output shaft sticking out of the back of the gearbox. You can slide a piece of surgical tubing over that portion of the shaft and then use a small ziptie to cinch it so that it stays on the shaft. Then you can fabricate some kind of simple plate to secure the Grayhill encoder to. It can be really simple we made ours with a small scrap of polycarbonate and a hand drill. then once the encoder is secure you can simple slide the other end of the surgical tubing over the encoder shaft and secure with a ziptie and you should be all set. It's also interesting to note that by using the surgical tubing method you don't have to make the shaft of the encoder and the gearbox exactly concentric. A few mm either way won't hurt the function of the encoder.

Thank you, yes, I can see how that would work. I will share this with our team.

[ice.berg]

I can second the surgical tubing and zipties method.

Third on the surgical tubing method.

Only instead of zipties we use spray paint.
Spray some inside the surgical tubing and slide it on. Acts like a lubricant while its still wet only then when it dries its like a glue. One of our mentors uses the same technique to put handlebar grips on his dirt bike. So far we haven't had any problems with the system.

[tr6scott]

We had some issues with surgical tubing this year on our elevator, when doing quick direction changes, we had some issues with the surgical tubing twisting and this small error would give rise to a noticeable "bounce" in the elevator as the pid corrected it.

I picked up a foot of 3/16" id fuel line from local auto parts store for a $1.26 and all of the bounce went away.

[btcshields6]

Thank you all for your responses and advice! I'll look over this information with my team and figure out what will work best for us.

Benjamin

[Bruceb]

you can also use pneumatic tubing. Just heat it a bit with a heat gun and slip it on. It is stiff enough to eliminate that twisting as well.

[Batterink]

Can someone post a picture of this surgical tubing method? I think I understand how it works, but I'd like to confirm my understanding.

[Chris is me]

Surgical tubing and zipties actually works very well.

Don't even need the zip ties. Get the .25" ID tubing, or even slightly smaller. Muscle it on and you have a rugged, flexible connection that doesn't twist or otherwise muck up your sensor readings.

Alternately, you can press fit the encoder into a .25" hole in a driveshaft, provided you have some LOOSE way to mechanically constrain the encoder from spinning.

If it wasn't clear, I'm also a fan of the Greyhills.

[marshall]

We use a total of 8 encoders on our drivetrain this year. We've got a mix of the CUI quadrature and the US Digital analog encoders. It's a pretty solid setup. We use the Helical (I think that's the brand name) shaft couplers on the US Digitals to couple them to the outputs of some versa planetary gearboxes. The outputs on the versa planetary gearboxes are not always perfectly straight from our observations so the couplers help a lot with the alignment issues. I suspect the surgical tubing will also help with alignment issues. It's clever. Surgical tubing is definitely more cost effective than the couplers.

We also had to switch out the set screws on the couplers to something with a cone tip and higher grip.

This has been our first year with the CUI encoders and we're pretty happy with them but the grounding issue is a serious problem for FRC teams. We had to get creative with some nylon 4-40 bolts to get everything nice and isolated but it worked. It's just something to be mindful of when mounting them.

EDIT: Also worth mentioning is Pololu's pre-terminated .1" cables. They are awesome and saved us a ton of fiddly crimping this year for wiring all of our encoders.

[SJaladi]

Just a side note from drivetrain encoders; when using encoders for a mechanism like an elevator where there is a finite end to the travel (i.e. the top and the bottom) it is good practice to have a limit switch at either each end or at least one end of the travel. This encoder can serve two purposes, an electronic limit on the travel as well as a way of resetting the encoder count at a fixed point. For example on our elevator this year we have a limit switch at the bottom of the elevator that acts as an electronic limit if the operator is manually lowering the elevator, but it also serves as a zero point for if the operator wants to "recalibrate" the elevator encoder to zero.

[Joe Ross]

Can someone post a picture of this surgical tubing method? I think I understand how it works, but I'd like to confirm my understanding.

There's a picture in this whitepaper: http://www.chiefdelphi.com/media/papers/1743

[Dunngeon]

Don't even need the zip ties. Get the .25" ID tubing, or even slightly smaller. Muscle it on and you have a rugged, flexible connection that doesn't twist or otherwise muck up your sensor readings.

Alternately, you can press fit the encoder into a .25" hole in a driveshaft, provided you have some LOOSE way to mechanically constrain the encoder from spinning.

If it wasn't clear, I'm also a fan of the Greyhills.

We've had great success mounting the 63r directly into the hex shaft.

We drill a .25 hole in our shaft with the lathe, then using a bit of hot glue push the encoder into the hole. Zip tie the wires to something rigid nearby and you're in business. The only thing to be wary of is ensuring that your tail stock and chuck head are aligned correctly, if they aren't the encoder will wobble quite a bit when the shaft is spinning.

[GeeTwo]

For coupling, we've used 1/4" shaft couplers. On our air cannon (in which we made a mini-cim and TB-mini into a servo), we coupled the shaft on the mini to the stub on the back of the gearbox and a potentiometer. This provided plenty of support for the pot - we just had to keep it from rotating. We found that some 10GA steel wire fits those encoder mounting holes nicely. We bent about 8" of this wire into a sufficiently rigid structure to hold the pot in place.

[Munchskull]

Question for the people using grey hill encoders. Why don't you use a more structurely sound mounting method or an encoder that let's a shaft in to it?

[SJaladi]

Question for the people using grey hill encoders. Why don't you use a more structurely sound mounting method or an encoder that let's a shaft in to it?

In the experience of Team 1923, because we lack the machining resources to mount an encoder perfectly concentric to the shaft that it needs to connect to, we prefer to use a mounting/coupling system that allows for some misalignment. In the past we've used shaft couplers that are flexible and allow for some misalignment, but once we discovered that surgical tubing gives us the same degree of precision that the flexible coupler did, we chose to use surgical tubing as it is much cheaper and easier to service.

As for the structurally sound part of your question, as far as my experience with encoders goes, we've never had an application which required a lot of rigidity. We generally just make a small Lexan tab that we bolt the encoder to and then we mount the tab so that the encoder is as close to concentric with the shaft that we want to couple the encoder to as possible.

Hope that answers your questions!

[ice.berg]

Question for the people using grey hill encoders. Why don't you use a more structurely sound mounting method or an encoder that let's a shaft in to it?

The encoder is still mounted to our gearboxes soundly, its just the coupling between the gearbox shaft and the encoder shaft that the surgical tubing is used.

A. Cost - for four corners its cents for surgical tubing compared to the 15+ dollars per corner for precision couplers.

B. Vibration - the surgical tubing spins the shaft but at the same time limits a lot of the vibration transfer into the encoder generated by the gears and robot moving around.

C. Misalignment - Goes hand 'n hand with vibration. The surgical tubing allows for slight misalignment of the encoder and the output shaft. Where as a precision coupler is not as forgiving. The grayhill encoders are quite durable compared to others but still we want to limit the extra load caused from misalignment

[asid61]

We drill the end of the shaft with 1/4" hole and add a nylon-tip set screw to keep the encoder from spinning. Encoder mounts are made from 1/16" thick clear polycarbonate. Works okay as long as the shaft remains reasonably concentric with the encoder.

[Dunngeon]

Question for the people using grey hill encoders. Why don't you use a more structurely sound mounting method or an encoder that let's a shaft in to it?

Because the ball bearings in the 63r give enough support to cantilever the remainder of the encoder in space. As far as I've seen, this method of mounting is the simplest, and consumes the least amount of weight.

Obviously you wouldn't do this on a machine seeing 100's ( or even 10's) of hours of run time because it wouldn't be sufficient. On FRC bots, it's acceptable.