idmil / t-stick Goto Github PK

TinyPicos can charge the Lipo batteries, but not the 18650. In a conversation with the folks from Unexpected Maker, it may be something related to chemistry (the 18650 shouldn't be able to charge all the way up anyway).

It can also be that the batteries purchases at IDMIL are not great. We need to run some tests and see how to fix it. In the meantime, we need to charge them outside the T-Stick.

Create T-Stick Design specification and document for easier maintenance and support

• Establish core functionality and features of T-Sticks
• Design tests to measure t-stick performance against spec
• Make decision log to keep track of major design changes and justifications for said changes

What we need to do is put Joe's code and take Travis' library out of the firmware.
It also needs some configuring and testing to see if it works.

Notes from testing signals with webmapper:

• battery: needs ranges, values not correct
• brush: range is too large, should be -1-1
• count: why is this a signal?
• double_tap: no signal values aside from 0
• jab: should be instanced, hard to detect
• multibrush: range too big
• multirub: should only be positive
• tap, triple tap: should be instanced
• accel: should make a normalized version, unknown ranges
• capsense: another normalized centroid signal
• fsr: needs normalized signal
• mag: values stay around zero

Hello

r and y are inverted

The button namespace outputs both 0 and 1 at the same time when I release the button. It is supposed to output only 1.

The connection guide doesn't necessarily reflect the details of newer wireless T-Sticks. The differences are not especially problematic or significant, but it does create an unsettling feeling as I set up the T-Stick on my desk.

• My T-Stick has no power switch. How am I meant to power it on?
• If my battery is dead, how will I know?
• My T-Stick's setup network SSID is very different from the one shown (TStick_312 vs T-Stick_SN_black2_194).
• The setup webpage did not open automatically on my (unusually configured Arch Linux) device.consistent/simpler method that could be suggested.
• The design of the configuration page has changed and doesn't reflect the screen captures
• The T-Stick appears not to automatically switch over to the configured network and needs to be rebooted now?

I managed to figure it all out, and I've made some edits to the connection guide to help anyone after, but a more formal update may be in order, by someone who knows what's going on...

We want to adjust the shake sensitivity by setting an offset value.

Lower the difficulty and time required to assemble a T-Stick

• Simplify Touch sensor Assembly
• PCB Design for fabrication
• Switch away from Shrink tubing
• Easier access to components for maintenance

There is no reference to where the configuration button should be connected in the documentation or on the schematic. (Latest Firmware says Pin 15) but this should be shown in the schematic diagram and/or referenced in documentation.

Libmapper crashes when creating signals during boot time. This is not a libmapper issue (the libmapper-Arduino example works fine).

It appears the T-Stick firmware is marked as MIT licensed, whereas libmapper-arduino uses the stronger copyleft, and therefore incompatible, GPL 3+. I've left an issue on the latter in case this was not intended to be so (mathiasbredholt/libmapper-arduino#11), but in case it is the T-Stick firmware probably needs to be changed to also use the GPL 3+ license. @edumeneses I'm guessing the MIT license here was inherited from the Puara template?

The problem is that it returns a negative value once you set the offset value. Clip the value.

• Better feedback for when a T-Stick is on/off
• better battery state feedback (current signal is unreliable and requires manual tuning when battery is replaced)
• replaceable batteries
• Update build guide to state the batteries need to have an on cell protection circuit as the tinyPico's BMS doesn't have one

When one of the I2C sensors is having connection difficulties, the entire device becomes inoperable, without any indication of what's wrong. This makes it difficult to tell when something is amiss, since the behavior is almost the same as a dead battery.

Ideally, when one of the sensors in inaccessible, the firmware should detect this, report it in some fashion (e.g. set all the sensor values to NaN? Or perhaps there should be status signals that can be set appropriately if a problem is detected?), and move on to trying to do the rest of the work that the T-Stick usually does, like reading all the other sensors.

Improve on the current capacitive sensor solution of the T-Stick with the aim to improve the spatial resolution, speed, reliability of the capsense

FW211124 code for the TinyPico LED isn't working. Probably some typo or wrong implementation.