Gear Tooth Sensors versus Encoders?
 

This year we decided to look into alternatives to using the gear tooth sensors - we've used the kit-privided gear tooth sensors and also other hall-effect type sensors for measuring robot speed and position for many years now, and wanted to check out our options. We've been looking into potentially using fixed optical encoders or shaft encoders in order to measure rotations on an axle, but they're just so darn expensive to get any that would hold up to a several-hundred RPM shaft.

I completely understand why we're using gear tooth sensors, but have other teams experimented with other mechanisms for reliably measuring wheel rotation that doesn't break the bank?

-Danny

Re: Gear Tooth Sensors versus Encoders?
 

We are going to try for the first time ever to use shaft encoders. We think we have a crack programming team of freshmen girls and we seem to have our ducks in a row so here we go.

We might even get to do some work in Labview.

Re: Gear Tooth Sensors versus Encoders?
 

Last year, we built our own encoders by using a pair of QRB1134 (http://www.fairchildsemi.com/pf/QR/QRB1134.html) sensors and a black/white disc attached to our wheel. We did it mainly because it's easier to mount for us than to try to attach an off-the-shelf encoder to our wheel module. The significant cost savings were a nice side benefit (the encoders we used to use were $50 a piece, the QRB1134s are $3/pair. You need to add a resistor or two to use the QRBs, but if you search for them on Google you'll find several websites of others who have used them in robotics applications (both for line following and making home-grown encoders).

It worked great and we're probably going to use those sensors again this year.

Re: Gear Tooth Sensors versus Encoders?
 

By "pair", do you mean 2 per wheel for quadrature output, or one for each wheel? That is certainly a cost effective solution.

We have used USDigital encoders on our bot for years. They are not the cheapest ones ($50+) out there but they reliable enough to be reusable (old robots don't run autonomous).

Re: Gear Tooth Sensors versus Encoders?
 

We've been messing around with these for a while now and they seem to work pretty well. Mounting them is a little fun, but last year we managed to make brackets and attach them directly to the output shaft. A simple PID loop for straight driving proved a success...

Only thing I'm not sure about is how much we paid for them!

Re: Gear Tooth Sensors versus Encoders?
 

This year team 40 is switching from the Grayhill to USDigital because we've had trouble with coupling. The USDigial encoders are nice because you can couple the optical disk directly to the shaft and they tolerate a bit of runout without coming loose. So +1 for USdigial....

http://www.usdigital.com/products/e4p/

Plus Andy uses them on the super shifter so they must be good right!

Re: Gear Tooth Sensors versus Encoders?
 

We've used USDigital H1-50's for the last 3 seasons. Again @ ~60 they aren't the cheapest, but they are readily available, and support has been excellent.
http://www.usdigital.com/products/h1/

We leave one drive shaft a little longer and then couple the two shafts together with surgical tubing. This helps absorb side loads, etc...

They've worked flawless and never had any issues with them.

Re: Gear Tooth Sensors versus Encoders?
 

Yes, 2 per wheel for quadrature output.

Re: Gear Tooth Sensors versus Encoders?
 

Don't forget the encoders and divider boards from Banebots. They work nice... especially with the BB trannies.

Jason

Re: Gear Tooth Sensors versus Encoders?
 

This will be our second year using US Digital's E6 line of shaft encoders, and while they're not the cheapest ever, we've had fantastic success using them with Kevin Watson's encoder code.

Re: Gear Tooth Sensors versus Encoders?
 

All this talk about encoders... so I'll put in one vote for gear tooth sensors.

Our gear tooth sensors cost about $25 each from DigiKey and one 5K ohm resistor and PWM cable later they are interfaced to the RC.

Granted they don't give direction information, but they also don't need direct mechanical connection and precise alignment. No moving parts means they will last just about forever. Plus they can be used as limit switches (detect ferrous metal their tip) and on sprockets, chain, etc.

Presently we have 0.07 inch resolution on the field with this setup using interrupt-based detection.

Re: Gear Tooth Sensors versus Encoders?
 

We also used these Grayhill encoders on our FIRST robot last year and they worked great for us. We use them on almost all of our robots on our college level team as well, except for a few times when we've bought motors with built-in encoders.

The output shaft of the grayhill is the perfect size for press fitting a K'NEX gear onto as well, which we have used successfully as the interface with an actual shaft in your drivetrain.