Jump to content

VaZso

Members
  • Content Count

    1,382
  • Joined

  • Last visited

  • Days Won

    75

Everything posted by VaZso

  1. Basically Gmail is very picky about these settings and I have registered this forum using my Gmail address (although SPAM filtering is practically disabled) - so it went through. I haven't checked if SPF record was set correctly. Anyway, SPF is a setting of DNS record for allowed IP addresses (as sender), another option is a DKIM record which is basically a signature all e-mails should have and there is a DMARC record (for reports) and basically Gmail like these settings... although it may be possible to send an e-mail without them, it may be harder. 🙂 Additional problem could be i
  2. I don't know who is responsible... Anyway, you may also look at spam folder or at advertisement / promotions on Gmail. I have just checked it and recovery e-mail has arrived for me.
  3. If we had a pinout then we may have some guess but everything may happen without knowing it... I may try to replace a screen anyway and hope it will do something. I think what I have ordered should be good but I did not have time to experiment with it since it has arrived last week (my original connector is still soldered on the beard). Here it is: Anyway, the first thing I would try is still the screen replacement. Maybe looking at the keyboard's behaviour could be still checked... Edit: I mean connector #7. Edit2: I may check if closing / opening keybo
  4. That was what we were starting with. 🙂 An I2C chip has an address which used for communication and usually devices have pre-programmed address or addresses. For example, address may be selected using one or more pins at the IC or it may also be a setting. In a few cases (specific circuits) also a custom address can be given. This case we have an address of 0x14 and also found a 0x5D address is an option (above in this thread). I have looked in other model numbers like gt1151 which had a different set of I2C addresses which may have a reason or at least it is not the same IC
  5. Okay, so Pro1's touch controller seems to be model GT911, PDF attached. ...or at least it mentions our 0x5D and 0x14 addresses which is a calming fact... ...and yes, it also has 0x805B and 0x805C registers for setting blank area. These are R/W values, so it is recommended to read before writing and also reading stock values of Pro1 would be interesting. Also touch/release levels may be set differently anyway. 🙂 GT911 Programming Guide_v0.1.pdf
  6. Another option would be to find I2C line of touch controller (physically), attach a logical analyzer, then let stock OS boot and analyze its initial communication during boot process. 🙂
  7. You are right and it seems the manufacturer of touch screen controller is correct but the actual model is basically seems to be different. This document states "GT928 has 2 sets of slave address 0xBA/0xBB & 0x28/29" but we have address 0x14 with an alternate address of 0x5D which is different than what is written in this document. However, it may be a very similar command what we need or (considering it is the same manufacturer) it may also happen to be the same, however, I would not try it. I have only found it described as "gt1x" but a better model name may help and a pdf whi
  8. Another official update about manufacturing has been begun at here:
  9. You can only purchase it together with screen, so search for a replacement screen (like screen for Elephone U Pro). (Touch is integral part of screen itself.)
  10. The cable is basically visible, look at the photo below: The ribbon cable circled with "1." is the potential cause of the problem. The green cable is the cable of display and the black on the right is the touch cable. These cables may break by moving but most probably their connection to the display "glass" can release caused by movement.
  11. My theory is that part (a wide ribbon cable at the bottom of display in portrait orientation) is start to move when glue of the display starts to release - only a slight moving of the display in its frame may sooner or later break that ribbon cable or glass-to-ribbon connection may releases which can cause touch or display-related issues. So a theoretically non-moving cable can start moving if display start to detach from its frame.
  12. It sounds strange... like there was already a contact problem somewhere... That case I would not really think a short-circuit caused during disassembly... That would be good, maybe you will notice something.
  13. When has your keyboard stopped working anyway? Did it work before replacing the screen? If I²C was not working, other devices on the bus were also missing. Any devices can stop the whole I²C bus easily by pulling down data or clock wires but that would also stop other devices on the bus to work. So the bus itself (that part of the SoC) is working. Anyway, have you used a metal tool to disconnect connectors? Naturally, replacing the motherboard would help but even F(x)tec does not have replacement motherboards for Pro1 as far as I know. My suspect is somehow you mad
  14. Definitively not. It should be a multi-layer board and a discontinuity simply can happen at an inner layer. However, connector may be accessible (not easily but may be soldered) and an appropriate power wire could be picked but without documentation it is not easy, also BGA packages do not help at all. Maybe one usable 3rd-party info would be a pinout of display's wiring itself - it may happen it may be reachable somehow and may be checked if appropriate power wires are present, also it may happen power wire of display part may be used to powering touch, but these are only tips from f
  15. Interesting... it sounds really strange you have more parts which are not working... do you also have other devices on I2C-6 what I wrote earlier? Naturally I don't know Pro1's schematics but another idea may be a missing power line somewhere - so at least for touch panel, I²C line seems to be working. It uses two wires, an SDA (data) and SCL (clock) signal... device may have other pins like interrupt but basically two other pins are needed which are VCC and GND. So if these parts are otherwise working and SCL/SDA lines are connected, then either power does not reach these ICs or GND
  16. Then I would think of a discontinuity somewhere which may be a not properly plugged cable (between main board and display) or really a touch controller (practically display) fault. You may also check visually if there are any problems with display cable or its connectors (like a deformed pin). So connections, connectors, display cables and also display can be the potential cause of your problem.
  17. Do you see other devices on the bus? So in my case, also 0x08 and 0x09 exist, how about you? Right, it may happen a simple disconnection or an unfortunate event your display has a faulty touch controller, but I²C may also stall and that is why it is interesting if you see other devices on the bus.
  18. It may happen - @agent008, maybe it worth a try to check if i2c-7 has a device address 0x14 using i2cdetect, so if touch controller replies to its address or it is simply inaccessible. However, it needs root privileges. It should look something like this (if querying all addresses): ~ $ sudo i2cdetect 7 Probe chips 0x03-0x77 on bus 7? (Y/n):y 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- UU 09 -- -- -- -- -- -- 10: -- -- -- -- UU -- -- -- -- -- -- -- -- -- -- -- 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- --
  19. Okay, so it seems 0x14 is the default address for gt1x touch panel instead of 0x5D, see here: So basically gt1x codes which use address 0x5d seem to speak about the same thing.
  20. I think Windows makes it harder. 🙂 Your problem may be potentially related to the USB port you use. Try to find a port which does not have other devices attached as EDL is really picky for timings. Edit: I mean in your computer there are USB HUBs which have more or less devices attached and usually not all ports on your PC are connected to the same root HUB. Devices connected to a root HUB (like internal web camera, WLAN or other components) may generate interrupts which delays communication with the phone and may cause update process running into an error.
  21. Anyway, device 0x58 is in use by aw9523b which basically is the keyboard, 0x28 is used by nq-nci which is the NFC controller, 0x08 is used by i2c-pmic, there is a 0x38 which is used by fts, 0x4a which is used by mxt-ts and 0x14 which is used by gt1x module. So it seems touch panel is connected to bus i2c-7 at address 0x14.
  22. Anyway, Pro1's touch driver seems to be called "gt1x" and it is an I2C device. There is a related code at GitHUB which also has an update code. Also, i2cdetect shows two buses on Pro1, i2c-6 and i2c-7. In former, there are devices at address 0x28 and 0x58 and in latter, at address 0x08, 0x09 and 0x14 where all devices are seem to be used by a driver except 0x09. However, based on gt1x_generic.h, I2C address of touch panel should be 0x5D (if that is the same device). There may be other I2C buses which are not accessible through dev filesystem... Also, i2c-7 has an owner of
  23. I have checked and "CONFIG_GTP_AUTO_UPDATE" is not set in config (LOS 18.1)... ...however, as far as I know, SailfishOS also uses stock Android kernel just like how LOS does, so I don't really understand why this option would be enabled under SailfishOS. However, at least the proper firmware is there and somewhere there should be a solution to access display's firmware...
  24. I would say very hard... ...I am still waiting for the promised replacement USB board and I had to disassemble my spare Pro1 to get a working board which was a Pro1 what arrived with cracked display frame and excessive rattling what prevented me to use that device for testing or developing.... So practically it was a step I had to make in order to continue using my Pro1 which otherwise worked and which is more than two years old now...
×
×
  • Create New...

Important Information

Terms