Most recent activity
-
Its rendered by creating an image object during boot and the data for which is stored in logo-white.js to save me time editing i just kept logo white (its used elsewhere not just in boot) the same. Then just copied the image object into the boot code. The one i posted i cant remember if i kade it 4 bit or 3 bit bit it was about 5000bytes for the full screen one
-
I managed to get the starting logo in colour again and learned why they dont do this anymore. I tried getting it full screen but it runs out of memory before i can get it to work.
Edit2: reduced number of colours in raw file to reduce size and now i get a full colour logo across full screen. With a bit more optimisation like better downscaling and colour blocking i can defo get the image smaller
Edit: using the 2x2 bayer pattern i manahed to get a full screen colour logo on startup
-
The second photo there is highlighting the button connector not the hrm connector. The green connector isn't populated on the starfield watch.
The plan now is to either recreate or make an original hrm sensor setup which shouldnt be hard. The code will be modified from the bethesda clock app since i will need to replace the pedometer with an oxygen sensor
-
i think that pretty much proves that the starfield watch comes after the G5 watch the back of the starfield watch has RT-H120D and the G5 not only is it richtechie but the model number is.....
RT-H120C
so as long as the chinese know the latin alphabet fairly well that makes the connection pretty obvious
-
Is there a place i can find the source for bangle functions? Like there is a bangle.health() function along with bangle.getcompass() and similar but i cant find out what i2c address it uses and get more of an idea on the workings of the function.
Its probaboy more obvious then i thought but ive not been able to easily find it. This way i may be abke to use the prexisting pins 9 and 10 i2c. Also on a similar note how did you know what went to the hrm header? Did you trace it out the pcb or was it provided to you?
If richtechie did a bunch of these chances are the hrm header is identical across any of the watches that has it, probably only with varying digital pin assignment but otherwise same power and gnd rails. I noticed in the bangle.js2 the hrm header looks identical with the same two shorted pins in a corner
-
@Gordon do you think there is much of a chance that the dickens watch was more original then we thought? like a little bit from A, a little bit from B and C and so on until they got to where they wanted.
It may just be a case of there is no original watch. Since you could draw a load of parallels between a lot of these watches without much of an issue, like it even has similarities with the bangle.js2. Richtechie have done another watch before the HPlus watch from 2016. Its not similar to the dickens watch but it shows they've done watches before. So it is possible that either richtechie did the schematic and PCB or that the wand company did both and Richtechie manufactured it or whatever.
I don't think the button count has any bearing on the watches origin since adding it wouldn't be at all difficult. The gw32 has the HRM connector that is the same as the one on the bangle.js2 and probably a couple other watches. With no clear original watch. My plan is to just add the HRM sensor to the dickens watch but ill love to know as much as there is to know about the origins of the dickens watch.
I should be able to have a javascript app access i2c without having to do anything fancy so I can defo do this project. It shouldnt be too difficult to add an app specific to the starfield watch so.
-
Im lookin through the GW32.py and it seems that if pins are shown under the def get_pins(): area are they unused? or simply available to be used by apps?
i2c is pins 9 and 10 on dickens (available on the supposed HRM header) but on the GW32 the i2c is on pins 3 and 47. I cant imagine the difference matters much. At some point these NRF52840 watches are starting to meld together in a bunch of ways. If we can confirm that the GW32 watch has the 3 and 47 pins on the HRM header then itll be more obvious that it was derived from the GW32.
In the Dickens.py the package is shown as QFN48 but its actually the aQFN73
I took a look at the G5 Smart Watch .js and .py file here
https://github.com/jeffmer/WatchApps/blob/master/apps/main-g5/main-g5.js
https://github.com/jeffmer/Espruino/blob/master/boards/G5.pyThis watch seems like a far more likely candidate for the dickens parent. It has the i2c on pins 9 and 10 and 8MB of external flash like on the dickens the only difference being it uses a higher resolution display. The G5 also has some pre-exisitng multiface code which could have been used for the multifaces on the starfield watch but that may be a stretch.
So along with the GW32 do you have any photos for the G5 watch PCB @jeffmer?
-
LMAO 5 minutes later update, looking through the board file for the GW32 found here
https://github.com/jeffmer/Espruino/blob/master/boards/GW32.py
The pins that are suspiciously missing from the Dickens board are present here doing exactly what I thought they'd be for pins 30 and 31 for the analog inputs of the HRM and pins 9 and 10 for the NFC coil which is a native part of the NRF52840 chip on those pins
From this i can 100% say that with 80% accuracy the GW32 is the watch that the LPV6/dickens/starfield/constellation/bethesda watch is based on
-
i might as well through my hat into this ring, for the past few weeks ive been trying to RE this watch focusing on the firmware for a while, just to find out its espruino.......
Now im trying to add a HR sensor to mine.... This should be very very possible. Im down to one of two ways to get it added, its to either find the watch this one is based on which would have the HRM sensor and the likes. or to add it in from scratch and simply use i2c to interface it. There is a spare unused connector on the pcb for the starfield watch which i believe was supposed to be used for the hrm sensor and could potentially have been used for NFC given its connection to those pins.
I would love to know more about the ancestry of the watch and i have a sneaking suspicion that the GW32 smart watch which is NRF52840 based can be a strong candidate since they are similar in some ways. My plan is to either buy one or find someone with one
ah ill definitely do that because the thing is tight for space as is, i wonder why they dont do that by default, cant imagine its slow enough to cause any issues.
Is there a way to tell how much memory/flash is left from the ide?