|
Post by riton39 on Oct 7, 2014 6:11:49 GMT
the best resolutions is with a potentiometer, the resolution is infinite. With a potentiometer, we can use all the input resolution 'of the AMC.
the higher the resolution is important, better the result will in games.
so, why choose something other than a potentiometer, the resolution of the potentiometer is much better! more resolution is perfect, but decrease the result with a less accurate sensor is mess up / deteriorate the performance of the AMC
|
|
|
Post by tronicgr on Oct 7, 2014 17:27:24 GMT
The potentiometer is based on contact of the wiper on the surface of the resistor, it will get worn fast for such application. And there is the noise issue, these potentiometers are prone to electrical noise the more resistance the have. That is the reason that many joysticks use 1k or 2k potentiometers, to avoid noise, and the resolution then is enough for the 128 - 512 points they provide.
If you really need better resolution and no electrical noise the only solution is the quadrature optical encoders but the hardware related to them is too expensive. I recently bought a US Digital encoder 2500 CPR resolution for the steering wheel project that total cost was $150 !!
|
|
|
Post by valmar on Feb 28, 2015 5:01:15 GMT
Hi guys! I have built a simulator, there is a problem. I have a 500mm linear guides, I set the sensor through the gear. And the guide is put forward only 300 mm. How to do it right?
|
|
|
Post by hexpod on Mar 13, 2015 20:38:03 GMT
I have just thought again about sensor resolution.
Please correct me if I am wrong.
Most of the sensors are claiming 12 bit resolution, I guess it's 12 bit per revolution.
That means that a 180 deg sensor has about 2048 increments. 160 mm crank shaft is giving me approx. a course of 500mm over 180 deg, which means that the smallest increment is 0.25mm with 180 deg sensor.
With another example of 12 bit/1080 deg multi turn sensor, the same course of 500mm has to be divided by 12288 which is giving me a smallest movement of 0.04mm or even 0.01mm with 3600 sensor.
Considering the precision of our gearboxes, I am asking, if it will bring a benefit to go for increments smaller than 0.25mm?
Cheers, - Ignacy.
|
|
|
Post by riton39 on Mar 14, 2015 9:39:48 GMT
and with a potentiometer with a track 320 ° electrical and a cam displacement of 180 ° although a reduction belt is made between two what is the resolution?
|
|
|
Post by hexpod on Mar 15, 2015 22:01:37 GMT
and with a potentiometer with a track 320 ° electrical and a cam displacement of 180 ° although a reduction belt is made between two what is the resolution? I have no idea how you can measure the resolution of a potentiometer. Anyway it seems that the 12 bit sensor, attached to the gearbox with ratio 1:1, is almost twice more sensitive as a 10 arc-minute gearbox.
What is your gearbox precision (backlash) in arc-minute?
Cheers!
|
|
|
Post by riton39 on Mar 18, 2015 22:49:51 GMT
"What is your gearbox precision (backlash) in arc-minute?"
i don't know. 0.5 /1mm with 20cm crank
|
|
|
Post by hexpod on Mar 21, 2015 23:45:48 GMT
"What is your gearbox precision (backlash) in arc-minute?" i don't know. 0.5 /1mm with 20cm crank 0.5mm on 20cm crank seams to be 0.14deg. (9 arcmin.) 12 bit single tour sensor smaller step is 0.09deg (6 arcmin.) However considering constant pressure on the crank lever, I can not say if it matters to go for more precision. (over 12 bits) I think only empirical test will bring a true answer. cheers
|
|
|
Post by hexpod on Mar 21, 2015 23:59:24 GMT
I guess it's 12 bit per revolution. With another example of 12 bit/1080 deg multi turn sensor, the same course of 500mm has to be divided by 12288 which is giving me a smallest movement of 0.04mm or even 0.01mm with 3600 sensor. sorry for wrong info. 12 bit multi turn sensors are 10 times less sensitive that single tour ones. cheers.
|
|