Oh I see, so I could buy a connector for IPEX 20879-030E and solder USB female to the other side…? I do see a lot of grounds, one USB 2.0 DP, one USB 2.0 DM but there seems to be no 5V that USB needs - what powers the webcam or am I blind? Also I misused the word propertiary, I meant “not standard USB, but made for the webcam and you can´t just go buy an IPEX 20879-030E to USB Female in a shop” not “custom made to not be available” - english is not my first language, I am sorry.
3 Likes
I concur with this. Of course, wear is inevitable, and connectors do eventually fail, so the Framework’s expansion cards do significantly reduce the risk of normal wear on the mainboard’s connectors - but for a project like this, I would say it’s a necessary and acceptable risk to take. After all, it’s still replaceable down the line!
1 Like
You’re right. Unless I’m blind too I don’t see 5v on the webcam pinout. Looks like it runs on 3.3v. Pin 22: 3V3 (common way of writing 3.3v in electronic designs). Usb devices don’t have to use 5v, it’s just that is the standard on external ports. Some externally attached usb devices don’t even use the 5v for anything internally, instead they regulate it down to 3.3v, because a lot of chips these days run on 3.3v. So when a device is internally attached and has no need for 5v, it saves components and complexity to just serve it 3.3v.
Looks like the Input Cover Interface has 5v on pin 38 (for touchpad) and pin 39 (for fingerprint). Note [2] and [3] says 38 is not always on and 39 is always on. 39 is probably still on when the laptop is sleeping, maybe even on when the laptop is off. You wouldn’t want that for a wifi adapter so pin 38 is the better choice.
Input Cover Interface pinout
But if you wanted to use the webcam’s usb instead, you could tap into 5v elsewhere. The fan interface connector might be a convenient place to piggy-back off of. It’s a JST SH 1.0 mm pitch connector. Those are common enough that you can buy cables and connectors from aliexpress or ebay.
But yeah, it might not be easy for you to interface with the input cover port. Due to needing to source a cable that mates with it.
1 Like
There is USB 2.0 on the Display, Webcam, and Input Cover (for fingerprint) connectors.
5 Likes
Important! Lengths of screws in my bucket list are incorrect, they just arrived and they don´t fit, turns out that M3x5mm screw doesn´t have 5mm long thread but is 5mm long overall. I will update them after I measure how long they should be.
5 Likes
@Surja_Knap Have you seen this?
A board for sale from a forum user that gives you USB 2.0 from the fingerprint connector. It looks like it comes with a female USB port on a lenght of wire. Though might want to check that.
3 Likes
@MJ1 Thanks, that´s exactly what I need, furthermore this occupies the fingerprint/keyboard slot so this means I can still get an USB from the webcam one - meaning I could possibly get 2x USB 2.0 within the tablet! Definitely ordering one
2 Likes
Looks like he has only one left. Might want to grab it quick. Hopefully he’ll make more, but who knows.
The 2nd port you need is just for a bluetooth adapter, right? I don’t imagine that would demand a lot of bandwidth, so how about a small usb 2.0 hub?
If you have a soldering iron there are some small boards based on the FE11SX4 chip available.
$10 tindie https://www.tindie.com/products/tinkerboy/tinkerboy-4-port-usb-hub-breakout-board/
$3 aliexpress https://www.aliexpress.us/item/3256801220206638.html
$4 clearer pics aliexpress https://www.aliexpress.us/item/3256803326487370.html
Adafruit uses the same chip for their raspberry pi zero hubs Zero4U - 4 Port USB Hub for Raspberry Pi Zero v1.3 : ID 3298 : Adafruit Industries, Unique & fun DIY electronics and kits
datasheet https://cdn-shop.adafruit.com/product-files/2991/FE1.1s+Data+Sheet+(Rev.+1.0).pdf
1 Like
Small update - correct screw lengths to hold case together are M2.5x4 mm and M3x6 mm - btw the project will be updated, I just don´t have time due to personal reasons. Has anyone tried to 3D print or somehow modify the project files? Any feedback is most welcome.
5 Likes
I would very much like to make this. Would you say we are at a point now when printing for the masses is viable?
1 Like
Well the situation is that it works, it is printable and daily driveable, will certainly be upgraded in the future, but no big upgrades planned in near future. I did make some updates of the design though, because I noticed that the display is now pressing a little against some parts of the reinforced chassis (which wasn´t an issue pre-reinforcement) and some other details, so please make sure to wait for me to upload the updated files before you print, you will have a better experience using it. Also please note the updated screw lengths.
6 Likes
Oh, I never uploaded the files, sorry! There you go!
https://drive.google.com/drive/folders/1dUqw_XWMt_TX7aPqv7IFRYEdS0o3vmo-?usp=share_link
3 Likes
Unfortunate update - after about a year of use, this design has proven to be a failure. I was trying to diagnose an issue where BD Prochot error would come up and CPU performance would get limited to 400 MHz. I always thought I had faulty board, because I received it for free with some minor issues + I did some stuff that might have damaged it.
But now I have a new board that I bought second hand and the BD prochot is still present.
I am almost certain that the battery causes it - unplugging it immediately resolved the issue. So the theory is that the battery overheats because of its placement. The only other chance is some driver conflict in Windows, which I will test.
Anyway, if anyone is considering printing it, I have to discourage you for now, as this issue can be a potential safety hazard.
I am working (funny enough unrelated to this issue) on another design that would feature a smaller powerbank inside the case and battery would be external, also potentially a different display.
I am sorry that I received the board for free and yet failed to deliver a good result after such a long time, I have some personal issues that prohibit me from working on it as much as I’d love to, but hopefully they should be resolved next month after long 2 years of struggle.
After that, I am planning to make big changes to the project:
- I finally have the board for palmrest connector, so the tablet will feature built in USB type A, wifi and bluetooth.
- It would be powered by battery bank that would be inside the case, it would give up some capacity (44wh seems to fit) but also should save some space and resolve the overheating issue.
- I would love to experiment with water cooling, which would be beneficial especially for AMD boards that can have 54W TDP.
- A different display. I may ditch touchscreen though for I really want to use eDP display. Or use a digitizer. I did research and nobody so far managed to connect a touchscreen via eDP. External display such as this one needs cables that run inside the tablet and basically take up most of the space. I may keep this display and increase the tablet size though.
I don’t know at this rate if I finish it in reasonable time, I promised a lot in this thread and didn’t deliver much, but I did buy about 60% of the stuff I will need for the project update, so it’s not like I am not planning to do it.
3 Likes
So, it´s not battery overheating. I am running the board now with battery in a standard configuration, outside of the tablet and I am still getting BD Prochot and 0.39 GHz but the battery is literally room temperature. Either there´s something wrong with it or the issue is somewhere else.
So I guess the design isn´t dangerous after all, eventhough yesterday I was very sure that the battery was overheating. Now I am puzzled.
2 Likes
Is it possible it’s the vrms overheating?
I don’t know, on 2 separate boards? And why would the problem be immediately fixed after battery unplug?
It would likely be a function of the airflow design so it happening on 2 different boards would be somewhat expected.
Pretty sure you get different power profiles with and without battery so the one without battery may just be gentler on the power delivery.
Yeah but now the airflow is good, everything is like an open testbench in a pretty cold room, also when the CPU is running at 400 MHz for 10 minutes, VRM should have cooled down long ago and immediately after I unplug the battery, that very second it jumps to 3 GHz range again.
It might latch on overheat.
You are unplugging the battery while running?! Pretty sure you are not supposed to do that.
Anyway, vrm overheat is just one possible source of prochot that is somewhat non-obvious because you usually don’t get access to the sensors.
1 Like
Well I am probably not supposed to do that, but I don’t think it can damage anything, I don’t do it regularly but it seemed a like a good step for diagnostic. I changed the board, so anything on the board is very unlikely - what is the chance of having 2 boards with the same issue?
What I transferred: RAM, SSD, Windows 11 installation , battery and peripherals. Assuming it’s nothing on the board, peripherals are easy to test, it can be maybe some faulty driver, RAM overheating but I can measure that with IR thermometer (it isn’t), SSD overheating (it isn’t) and battery overheating. That was the most likely cause as the battery was placed where it could get warm. So I went for the most likely cause and unplugged the battery and it stopped. If it were VRM, they would cool down in seconds after switching to 400 MHz.