What resolution do you require? Only whole degrees? Then you need an encoder with 360 counts per revolution (cpr). That's not a difficult encoder to find.
The chip you linked to is usually what you find inside an encoder. It will work if you can 3D print a holder which holds the chip perfectly in line with the shaft with the magnet on it. Personally, even though I have a 3D printer, I would use a pre-built encoder. That will ensure that the relationship of the magnet to the chip is perfect.
I really like the USDigital MA3 encoder. It's pretty much the same thing as that chip except all the hard work is done for you. The high resolution version has a resolution just a little better than 0.1 degree. It may be outside your budget. Their optical encoders look good too, but I have yet to use any of them.
But if you want to continue with that chip, then it really doesn't matter which package you choose. SOIC has larger pins so it will be easier to solder.
An analog output requires an analog input on your Arduino. It's also subject to noise and interference. I would pick the SPI option. This will use the MISO, MOSI and CLK pins in parallel for all encoders, plus one more "select" pin for each encoder.
As far as resolution goes i am open to options depending on price of the parts as i need 6 of these encoders to make 1 joystick and it can become very expensive pretty quickly if they are $20+ each.
Range of motion of the encoders is about 30° - 150° (120*) from what the guy says about the Stewart platform.
From memory similar application that i know of that uses hall sensors if a gimbal for a RC radio the Frsky M9 Hall Sensor has 1024 resolution. These can be purchased for ~$27 per gimbal and have 2 hall sensor and 4 ball bearings in a + configuration "joystick" if you will.
Thrustmaster warthog uses a hall-sensor setup as well although i don't know what exactly it is using i am trying to google to find answers with not much luck.
Most other joysticks say they are using "pots" so that's some kind of potentiometer setup?
So 1 encoder for $55 seems a little expensive for this project i was thinking something could be found a little cheaper but i am not a electronics person and i don't really how much this stuff is worth
Reference mlx90316 is listed for about $16 for 10 and the pcb brake out board for $2 mind you i would have to manufacture the rod + magnet although with not much experience in this sort of stuff might be beyond my capabilities although i am confident i can get something working.
Maybe start with analog potentiometers and see how much noise you get. With a little luck, it may be adequate for your purpose. Pots are cheap and not much bigger than a small encoder, so you can replace them with encoders later if required.
MorganS:
Maybe start with analog potentiometers and see how much noise you get. With a little luck, it may be adequate for your purpose. Pots are cheap and not much bigger than a small encoder, so you can replace them with encoders later if required.
I don't really know much about them i just know they are called pot's (LIN pots) from what i read its just a better quality potentiometer. The resistance part i don't understand how much is needed the range is massive from 50 Ohms - 1 MOhms
Pot's it whats currently used for this application because they are cheap i guess. Higher end applications use Hall Effect Sensors in the currently market so i would think that a "final version" would use these. I just have no clue what i am looking at when it comes to these or what to purchase. I think for now by the sounds of it a nice pot will be sufficient as a prof of concept. Some examples of each would be really helpful if the price is not to much hall effects would be nice but i think they might be a little to much.
Thanks for all the help so far i am learning as much as i can in my spare time.
What's the diameter of the shaft? You might be able to 3D print a flywheel with notches cut out of it all around the inside of the flywheel that mounts onto the shaft, then use an optical encoder like this. The number of notches on the flywheel will determine your resolution.
androidfanboy:
What's the diameter of the shaft? You might be able to 3D print a flywheel with notches cut out of it all around the inside of the flywheel that mounts onto the shaft, then use an optical encoder like this. The number of notches on the flywheel will determine your resolution.
This part of the project i have to re-design so currently there is are 6 arms and each have a cut out for a 9g servo that i am replacing with a angle measuring device so i will have input instead of output.
Also https://www.aliexpress.com has a lot of options for what seems like very good prices but i still don't know what i am looking at exactly for my requirements.
I think the optical option may end up a little large depending on the flywheel size as far as printing a flywheel the edges wouldn't be that smooth i would think if i were to do this destroying a old ball mouse for parts and building a housing new housing for those parts would be better for a more compact result. At that stage i might as well try build a hall effect setup from the parts in the first post as it would be more professional and more inline with the final product. Although i don't know how hard it is to get the distances correct and how much forgiveness there is in the setup.
Potentiometers type Conductive plastic
Very smooth feel and high resolution, can be constructed to perform millions of cycles. Can only handle a limited power and are expensive. Often used in high-end (audio) equipment where a high resolution and low noise are important.
This type seem to be around ~$11 each sold in packs of 5 so i would need to by 2x packs for this project
So my understanding noise will be bad as it will throw the reading out. It would be possible to code some of this variance out something like update change of the rotation if its in this reading is different longer than 0.x of a second or change is greater than X.
It looks like you are linking to expensive audio pots. (Have they sold you any [useless] gold-plated speaker wires yet?) Something with "LOG" in the name and dual pots is actually a volume control for stereo.
You want linear, or "LIN" or "B". You should be able to find a bag of 20 for $5-$10. The actual resistance is not important at all so anything in the range 1K to 100K will be fine. (10K is popular.)
The hall-effect magnet sensors aren't too sensitive. If you're willing to experiment, then go for it!
MorganS:
Yes, you can deal with the noise in software.
It looks like you are linking to expensive audio pots. (Have they sold you any [useless] gold-plated speaker wires yet?) Something with "LOG" in the name and dual pots is actually a volume control for stereo.
You want linear, or "LIN" or "B". You should be able to find a bag of 20 for $5-$10. The actual resistance is not important at all so anything in the range 1K to 100K will be fine. (10K is popular.)
The hall-effect magnet sensors aren't too sensitive. If you're willing to experiment, then go for it!
Excellent much more info on specifics what i am after as i said i really don't have much of a idea when it comes to electronics.
By the sounds of it i will save the hall-effect sensors for a more final version when i have worked out all the design bugs with the cheaper pots and what seems somewhat reliable. For now it will be a prototype version 1 basic no bells and whistles yet save the good stuff for the pro version with a "Fatal1ty" tag on it LOL.