Build Complete

[Dan W]

Despite some setbacks and a few full on fights with glue, I seem to have completed the phone.

Few comments:
1: With the manner of how I assembled my dialing assembly, it is possible to press the main board and the daughter board too tightly together and bind up the dialing mechanism, which I was reminded of once I got both halves of the case in place. I suggest checking the operation before affixing the e-paper as after that the screws to loosen it are hard to access.

2: As far as fights with glue are concerned, a tip for others: have some isopropyl alcohol handy. When used immediately it prevents the super glue from hardening up properly and will allow you to scrape it off more easily. It also breaks the bonds of standard hot glue.

General design/parts comments:

1: The dial cap appears a bit on the thin side, leading the clip points to be rather obvious:


2: My e-paper seems to be mounted just a bit high, leading to some wavyness in it.


3: This is a picture of what it looks like when you try to use the wrong screw to secure something.


4: The side window pops out of place quite easily even after assembly is completed. Not sure if I'm just too heavy handed or what other solution could be come up with.


Overall this has been a lot of fun I'll post again when I get my hands on a SIM to stick in it.
(Sometime this week)

Link to photo page to give the host some traffic.

[rwyeth]

My build is also complete! And, it does some stuff. It powers on, and I put a SIM card in it from US Mobile. As for the SD card, I need to hunt down an adapter to upload the files to it. There do seem to be a couple of issues in my initial testing.

1. It only dials the number 2. No matter what number I try to dial, it dials a two. The white light comes on when I hit the finger stop; but, the blue light does not pulse as the dial returns. I'm not sure where to begin with that.

2. I'm not sure if it is an issue with the ePaper display; but, as you can see in the picture it is an "otary UnSmart Phone."

3. Checking with US Mobile's chat, the phone is not compatible.

That is about all I know so far. If anyone has suggestions, please let me know.

[Dan W]

Mine does the same thing with the missing 'R.'

As far as the picking up only 2 goes, I think the "limit switch" might need to be adjusted so that it engages the cam a bit more and clicks on each rotation.

[justine]

The only 2's is VERY odd. I need to look through your post comments more carefully but were you careful about the Cam Gear alignment before assembling the daughterboard?

-As for the missing "R", it turned out I accidentally deleted the lines that write that letter from the splash screen function... it's not a problem with the screen.

-Re. the waviness in the display, I've revised the placement of the foam tape on the display, which seems to eliminate the problem.

-Re. the side window and dial cap, yes I agree. Room for improvement in the future.

I made some other updates to the firmware, and before doing anything else I'd suggest flashing it with the latest, also to see how it affects your issue with the 2s. I documented this process here: https://docs.google.com/document/d/e/2P ... orLRnM/pub

[rwyeth]

Justine,

I believe I was careful in the Cam Gear alignment; I do remember that step. But, I guess that is where I will begin if flashing with the new firmware doesn't resolve the issue. Honestly, I'm excited that I got the whole phone together and that it powers on and does something. This sort of project is a bit out of my scope of expertise. Flashing it should be another small adventure; but, it looks like you have very detailed instructions. I'll give that a shot when I have time (probably this weekend).

Thanks again!

[justine]

I should add now (amazed I didn't earlier), beautiful builds.

I've made some tweaks to the assembly process and to the way the ePaper is mounted to fix some of the problems in the newest version of the instructions.

[joelsef]

I also got mine put together...finally! But having the same issue as rwyeth. It only seems to dial 1s or 0s. Sometimes 2s.

[imyxh]

Hey, sorry for the delay on this. We at Sky's Edge have been busy with a lot of things and I'm only now getting a chance to go through the forum!

My build is also complete! And, it does some stuff. It powers on, and I put a SIM card in it from US Mobile. As for the SD card, I need to hunt down an adapter to upload the files to it. There do seem to be a couple of issues in my initial testing.

1. It only dials the number 2. No matter what number I try to dial, it dials a two. The white light comes on when I hit the finger stop; but, the blue light does not pulse as the dial returns. I'm not sure where to begin with that.

[...]

If the blue light does not pulse, it means the firmware is not seeing any clicks from the limit switch. The 2 is actually a negative 2 (you can see this in the very upper left of your screen). This is because there is a "fudge factor" of −2 (sometimes −1) when converting limit switch clicks to the dial number, due to the mechanical workings of the rotary mechanism. You can see this in the old firmware here: https://github.com/skysedge/RotaryUnSma ... rs.ino#L52. You can try flashing the newer (work in progress) firmware at https://github.com/skysedge/rusp_firmware, but most likely there's a mechanical or electrical issue. Some steps you can take to debug this:

1. Disassemble the phone (I know ...) enough so that you can reach into the cavity that normally houses the gears, and manually click the limit switch. Start clicking the limit switch, get a permanent magnet and swipe it close enough to the phone that the white light comes on, then continue clicking the limit switch with your finger a certain number of times. For example, "click click SWIPE click click click" would indicate to the phone that the dial clicked three times while returning, which would indicate that the number one was dialed (subtract two).
2. If that works fine, then the mechanical assembly is not properly engaging the limit switch while dialing.
3. If you do step #1 and don't see any blue flashing, then there's something wrong with the firmware or the motherboard. You could check solder joints, maybe run some continuity tests, etc.

[...]

3. Checking with US Mobile's chat, the phone is not compatible.

I actually wouldn't be super sure about that. If US Mobile is implementing some sort of IMEI allowlist then that's news to me—though I also wouldn't be surprised. It's likely they tell you that it's not compatible because they don't know what the RUSP is and worry it might be carrier-locked (it's not). Even without the dial working, you can check if the phone is compatible by connecting to the RUSP over serial and sending AT commands to the LARA. This is easier than it sounds! If you get a serial console up, you can try typing e.g. ATD+13034997111; (with the semicolon at the end!) to place a test call to the NIST telephone-time-of-day service. If that doesn't work it's worth checking the output of AT+CREG?, AT+UCFSCAN=7, and some other commands.

There's a cheatsheet of AT commands that I've compiled here: https://raw.githubusercontent.com/skyse ... atsheet.md

and the documentation for LARA has a list of all of them here: https://content.u-blox.com/sites/defaul ... 046719.pdf

[imyxh]

I also got mine put together...finally! But having the same issue as rwyeth. It only seems to dial 1s or 0s. Sometimes 2s.

If you can get any amount of variability in the number you dial, it's at least a good sign that the limit switch can talk to the motherboard. It's possible that one of the following is happening:

1. the cam is not engaging the limit switch often enough,
2. the dial is returning so quickly that the firmware is ignoring multiple limit switch presses for debouncing purposes, or
3. the dial is returning so slowly that the firmware has decided it's done returning when it really hasn't.

Something you can try is flashing our work-in-progress rewrite of the firmware: https://github.com/skysedge/rusp_firmware, and if that doesn't fix things, playing with some of the timing in the code. Specifically:

• at time of writing, the limit switch is debounced by 30 ms intervals; you can make that longer here: https://github.com/skysedge/rusp_firmwa ... re.ino#L51
• at time of writing, the timeout that decides to stop waiting for limit switch presses is 500 ms; you can make that longer here: https://github.com/skysedge/rusp_firmwa ... e.ino#L233

Of course, neither of these approaches are ideal solutions. Ideally, the mechanical assembly would simply allow for the dial to return at the right speed. But I'm fairly new to the team and am not sure how to address that issue yet.