Update readme

This commit is contained in:
Scott Bezek 2022-02-14 10:29:06 -08:00
parent 493e783355
commit ac6ef1557a
8 changed files with 48 additions and 7 deletions

View File

@ -14,29 +14,38 @@ Premium SmartKnob experience. Under active development.
### SmartKnob Handheld
An exploration of a small handheld form-factor. Under active development.
<a href="doc/img/PXL_20220212_013249074.jpg"><img src="doc/img/PXL_20220212_013249074.jpg" width="200" /></a>
### SmartKnob Mini
Planned for the future.
## Component Info
### Magnetic encoders
#### TLV493 (Infineon)
A decent choice, and easy to prototype with using [Adafruit's QWIIC breakout board](https://www.adafruit.com/product/4366).
#### TLV493D (Infineon)
A mediocre choice. Easy to prototype with using [Adafruit's QWIIC breakout board](https://www.adafruit.com/product/4366).
In my testing, it is a little noisy, requiring filtering/smoothing that can slow responsiveness, hurting control loop stability. Or, with less filtering, the noise
can easily be "amplified" by the derivative component in the PID motor torque controller, causing audible (and tactile) humming/buzzing.
There is also apparently a somewhat common issue with the internal ADC locking up, requiring a device reset (see datasheet section 5.6). From a random sample of 4 Adafruit breakout boards, 2 of them (50%) appear to be extremely prone to this - locking up within a minute or so of use! Further testing is planned, but this may be a deal-breaker.
There is also apparently a known silicon issue that causes the internal ADC to sometimes completely lock up, requiring a full reset and re-configuration. See section
5.6 in the [User Manual](https://www.infineon.com/dgdl/Infineon-TLV493D-A1B6_3DMagnetic-UM-v01_03-EN.pdf?fileId=5546d46261d5e6820161e75721903ddd)
In the Master Controlled Mode (MCM) or the Fast Mode (FM) the ADC conversion may hang up. A hang up can
be detected by:
- Frame Counter (FRM) counter stucks and does not increment anymore.
In my experience testing 4 different Adafruit breakout boards, 2 of them (50%) regularly exhibit this lockup behavior within a minute or two of use.
As of 2022-02-08, there is limited availability of this IC.
[Datasheet](https://www.mouser.com/datasheet/2/196/Infineon_TLV493D_A1B6_DataSheet_v01_10_EN-1227967.pdf)
#### MT6701 (MagnTek)
Very promising based on the datasheet and initial prototyping.
Very promising based on the datasheet and initial tests. Seems to be less noisy than TLV493D, and more responsive (control loop is more stable) using SSI.
- Haven't testing in a full control loop yet, but data looks good and relatively noise-free.
- Lots of IO options - SPI, I2C, and ABZ - should offer good response latency.
- Lots of IO options - SSI, I2C, and ABZ - should offer good response latency
- SSI include CRC to validate data
- No power-down or low-power options - may not be ideal for battery-powered devices
- Not available from US distributors (Mouser, Digi-Key)
@ -58,11 +67,43 @@ Highlights:
[Product page](https://www.trinamic.com/products/integrated-circuits/details/tmc6300-la/)
### Motors
TODO
#### 32mm Rotor, Hollow Shaft, Diametric magnet
<a href="doc/img/motors/PXL_20220121_221746595.jpg"><img src="doc/img/motors/PXL_20220121_221746595.jpg" width="200" /></a>
<a href="doc/img/motors/PXL_20220121_221738745.jpg"><img src="doc/img/motors/PXL_20220121_221738745.jpg" width="200" /></a>
- 32mm rotor
- 15mm overall height (including magnet), 12.75mm height w/o magnet, 9mm rotor height
- low/zero cogging - excellent for completely smooth input
- 5.9mm hollow shaft
- built-in diametric magnet for encoder
- Proven option
This is overall the easiest motor to get started with. Low cogging and a built-in diametric magnet are great!
[AliExpress](https://www.aliexpress.com/item/1005001384349991.html)
#### ~17mm Rotor
<a href="doc/img/motors/PXL_20220121_220530808.jpg"><img src="doc/img/motors/PXL_20220121_220530808.jpg" width="200" /></a>
<a href="doc/img/motors/PXL_20220121_220546761.jpg"><img src="doc/img/motors/PXL_20220121_220546761.jpg" width="200" /></a>
- Not tested yet
- Stepped 17.3mm and 16.3mm diameter rotor
- 11.2mm total height, 8.6mm rotor height
- moderate cogging - not ideal for completely smooth input
- Rotor has threaded hole on the side that should be usable for keying a cap
- Very little shaft accessible on bottom - may be tricky to mount a diametric magnet to it. Maybe could replace screw with something else?
[AliExpress](https://www.aliexpress.com/item/1005003240145615.html)
# Firmware
TODO
# Acknowledgements
This project was greatly inspired by Jesse Schoch's video "[haptic textures and virtual detents](https://www.youtube.com/watch?v=1gPQfDkX3BU)" and the
corresponding [discussion in the SimpleFOC community](https://community.simplefoc.com/t/haptic-textures/301).
# License
This project is licensed under Apache v2 (see [LICENSE.txt](LICENSE.txt)).

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.9 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.2 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.2 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.2 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.2 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.1 MiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 1.0 MiB