Way to make bangle js waterproof (e.g. epoxy)

Are you using the KX023 hardware doubletap detection? There is also some sensitivity tuning

Yes, I am using the detection in the poll handler in jswrap_bangle.c to pass button io events to the queue.

For regular use this was working great but for the sports mode I will try changing the sensitivity. However the real button might be a problem, too. So let's see how my final solution will look like and if it is hopefully usefull for others, too.

48 hours until I can check if it is watertight.

1 Attachment

The first kite boarding sessions with the glued watch were a success. It survived many hours, massive impacts, bodydragging, etc. with the watch under water and heavy water pressure. Here a photo with the watch and my wetsuit.

Handling the watch in everyday life without the button using tabs works well for me.

So now I am thinking if this can be useful for anyone else? I could give the 3D printed watchcase a try, because the glue looks bad and won't last forever. The result could be a tutorial explaining how to make the 3D printed backcover waterproof, reassemble the watch. Supplying the print model and the code for the buttonless mode.
I could even imagine ready to buy kits in your shop with the unglued parts, but I think the watch is no longer produced? So maybe for the bangle 3, where I still believe that buttonless is the only way to create an affordable 5 or 10 atm watch. But bangle 3 is not coming soon? Also some people like their buttons and barometer. So maybe think about offering it in two flavors or offer a watch and a buttonless smartband?

It would be some work to make my prototype code ready for a PR and I guess we would have to discuss how to handle some details. So I would only do that if there is a real demand otherwise I am happy that I have a waterproof watch now and learned a lot about programming for an embedded device.

1 Attachment

That's great! What does everyone think about whether there's interest?

I guess a frame that went over the top of the watch that was epoxied on could make quite a tidy fix, and also would help to protect the screen from scratches. I'm not sure we have to really worry about water ingress around the screen itself, but I guess it doesn't hurt!

I think the watch is no longer produced?

Bangle.js 2 is definitely still being produced and sold, I've got thousands of them in stock here :)

The issue with something like a Bangle.js 3 and having two flavours is it's a lot of extra expense (and stock to keep). Unfortunately I don't think I'm selling high enough volumes that it would ever really pay off having two SKUs (unless maybe Bangle.js 3 could be opened, and then maybe I could offer a separate case to swap the innards into).

To answer the question of Gordon whether there is more demand for watertight watches: there surely is! Everytime I go swimming I almost forget to remove my bjs2 from my wrist.

But I don't know if I could go buttonless. I need to test it to find out.

@Chriz is there a variant of your firmware that could be installed as app? And if not, can you point me to your source so that I can strip out the essentials and make it an app myself?

I would not want to switch from the official firmware for my daily driver right now, but I would like to find out if I could live without the button before going further with this...!

Hello @devsnd, my sources can be found at https://github.com/Chriz76/Espruino/tree/masterOld .

I used github for building the firmware so a lot of little commits where I tried different things But the changes itself affect only a few files.

I modified the firmware and did not create an app. But with my long term experiences I don't think that this is necesarry:

Even after several weeks the watch is still watertight. The baramoter hole seems to be the most important. The glue on the button already partly vanished. And even if glued the button could be used to do a hard reset.

Double tapping twice (4 taps with a short delay after the first two taps) to simulate a button press worked well in regular use. It should be possbile to implement this as an app. But using double tapping on the other side of the watch to simulate a long press often lead to false detected presses and returns to the clock. So I changed it to 5 double taps while my watersports app is running.

I also used the charging state for a hard reset.

For you it might just be ok to start with a global hook as app for 2x double and 5x double tap. Then glue the barometer hole and the button. If you really need a hard reset then push the glued button hard and it will work.