after reading this thread, I decided to build a factory cable.
It is pretty stright forward indeed: I bought a micro-b usb connector, removed the connector of an old USB cable and soldered wires in correspondance to color and number. Because this is a factory (McGyver) cable I also soldered a 100kOhms resistor from VBUS to ID (i.e 1 to 4). For reference look here.
Now the interesting part (I am on linux):
Code:
$ lsusb
...
Bus 001 Device 067: ID 22b8:41d4 Motorola PCS
this is different from how my defy is reported using a normal cable:
Code:
Bus 001 Device 069: ID 22b8:41db Motorola PCS Droid/Milestone (Debug mode)
While using a normal cable pluging it into the phone triggers a "USB connected" symbol on the phone nothing happens when using the modified cable, which is strange.
Also note the fact, that using the modified cable, the phone is reported earlier during bootup.
Like expected from a factory cable I can now boot the phone with battery removed. But after booting without battery my defy reports that no SIM card is inserted, and that the battery is at 5% (wth!) and must be charged... otherwise (WLAN, apps, etc.) it is working fine.
Starting a terminal app with su shows me that an instance of adbd is running, but adb client on my host machine lists no available devices. (Starting adbd from 2ndinit works fine, but the phone locks up if I try to continue booting, probably due to failing to aquire ressources for another adbd instance)
The question now is whether there are means to circumvent this lmitation i.e. forcing detection of SIM and restarting adbd in another mode.
My defy is rooted but otherwise stock european build without any updates.
Can somebody confirm my observations?
Thanks in advance
EDIT: another source with explanations
There are reports that this damages the battery...
So better stay away from this don't put the battery...
This method mainly help when ur juice in battery is empty and u wanna flash ur phone with RSd but can't Coz u see battle low cannot program ... so using this cable eliminates this(helpful when u have a non removable back(Raze))
However from what I read this wont charge ur battery...(I don't know)
Now ur issue...
Can u see baseband in ur about phone??
Sent from my MB525 using xda premium
nogoodusername said:
There are reports that this damages the battery...
So better stay away from this don't put the battery...
Click to expand...
Click to collapse
Thanks for the warning... Does the µC drive the battery controller berzerk?
nogoodusername said:
Now ur issue...
Can u see baseband in ur about phone??
Sent from my MB525 using xda premium
Click to expand...
Click to collapse
Baseband is EPU93_U_00.60.00
When I reboot the phone it will work fine, just when booting without battery it stops recognizing the SIM card.
Another thing is strange: If I plug the "factory" cable after a clean reboot on battery then unplug it and use the standard cable the phone will still not recognioze the usb connection.
Until next reboot that is...
I also looked at sources of adb and observed, that using the factory cable on my linux box the phone does not report the adb subdevice, ie. adb.h:
Code:
372 #define ADB_CLASS 0xff
373 #define ADB_SUBCLASS 0x42
374 #define ADB_PROTOCOL 0x1
-- so this is the reason why adb fails, the phone just does not register the appropiate subdevice, no idea why...
My last try was to remove the battery after connectong "factory" cable, which caused the phone to bail out a warning about low battery and then instantly reboot. aw snap!
You should be able to short pin 2 and pin 3 (D- and D+) together, and then this cable would be considered a "charge cable". It will not be usable for data any longer, and the phone should report "AC charging" (try the Battery Monitor Widget app) when it is plugged in with this cable, and it will charge at a much higher rate (higher current draw) with this cable.
Sent from my MB525 using Tapatalk
My actual goal now is to get adb running when on factory cable, but I do not have much time at the moment.
In a second stage it would be interesting to force the phone to detect the SIM.
This would allow to use the phone without battery....
I have a Nook Simple Touch and its battery life is coming to an end. Due to the intensive use I give it.
I usually use my Nook in a fixed place. I have been searching around for a way to use the Nook, without a battery, just plugged to the wall energy. I haven't found anything.
Does anybody know how to use the Nook with no battery?
Thank you.
--
P.S. If you are looking for a battery, I found this http://www.powerbookmedic.com/Nook-Simple-Touch-Battery-p-22959.html
Dissassembling it looks pretty easy https://www.youtube.com/watch?v=Zr0dCMavbPo
I'm not sure if physically removing the battery would affect anything, but if the battery is there I know that you can plug in your usb cable to a PC and then unmount it but leave it plugged it. That should allow it to charge/run and let you read with the device.
In Windows you would go to the "Safely remove devices" and choose the nook. In Linux you could either unmount it through a file manager or go to a shell and do "umount /dev/sdc" or w/e your device id is (or equivalent steps in your distro). I am not sure how to unmount in OSX... maybe drag the drive icon to the trash?
I've tried that experiment, pulled the battery and tried to get it to boot with external power.
It won't.
Although the PMIC (Power Management IC) is designed to work with a dead, missing or shorted battery, something doesn't let it.
I don't know if it's the lack of the temperature sensor when the battery is unplugged.
In any case, if the battery is mostly dead, who cares, just plug in a charger permanently.
File Under Useless Information
I used a parts phone ordered from eBay to repair my LG G5 then repurposed leftover parts from both phones as a spare battery charger. I had been considering ordering the LG G5 "Coffin" Charger but didn't see many listed on ebay and those that were, the prices were a bit too high to justify for a convenience on an old device. I still use my LG G5 as a daily driver and keep an Essential PH-1 as a backup device, yeah I know that seems backwards but I like the LG mobetta.
In my instance, parts leftovers did not include a good LCD/Screen so to check charge level, I have the status light on top right (red = low/green=good... der). Also, if USB debug happens to be enabled by default on motherboard used in charger, you can get exact charge status by sending command in ADB shell [adb shell dumpsys batteryproperties].
I'm sure this cannot be a first. Ha
also posted on LG5 reddit
I run my hot spots 24/7 and they overcharge the batteries and puff up.
Verizon already did one recall on the Franklin "Ellipse".
I do a direct USB powering since many devices won't run without a battery.
The photo show the one for the MiFi 8800L with a diode, a 51k ID resistor and a 100k thermistor resistor.
It works fine, but I still wasn't happy with how warm the device got even without a battery.
So I threw away the back plate, broke the glass/touch panel off the bezel and removed the OLED display.
Now the tin covers on the processor get direct access to air and it runs a lot cooler.
It seems to work fine.
Still, I wanted to be able to run "top" and make sure that it's not thrashing too badly without the OLED and the I²C touch panel.
I plugged the MiFi into Windows 10 and identified a single HID interface with a 4 byte input record and a 4 byte output record.
I tried to read the input record, nothing was coming out.
So I wrote my 4 bytes of zeroes (in Windows that makes it 5 bytes since you have to add an unused record ID).
Code:
memset(report, 0, 5);
WriteFile(hid, report, 5, &n, NULL);
And lo and behold, the MiFi reset and came back with 7 interfaces, including the original HID.
One of the interfaces was a CDC serial that came up as a virtual com port (VCP).
Code:
mifi login: admin
Password:
Login incorrect
Any idea what the password could be?
I wrote a password spammer but after 30-40 login attempts it stops listening.
I added a 40mm square heatsink. Now it runs even cooler. I cut away some of the top cover so that you can still plug in the LCD if you need to.
If you're thinking this is all pointless: In marginal areas a hot unit decreases the S/NR. I've often stuck my old Jet Pack in the freezer when nothing else would get me a connection.
So I believe the password is listed in the device settings. On mine its under advanced settings on the screen.
I want to get more outta mine and am trying to put custom firmware on it.
I'm not sure if you mean the regular WiFi or administration logon password. I've tried those. I don't know what the login itself is either. I've tried root and admin. Can you give me a screenshot of that password using the regular web interface? (Ok, with the password blacked out.)
In other news, I discovered that the network manual selection allows me to select AT&T or Telekom (since I paid full price for the unit). It may not have all the right bands, but I think that I'll try test drive one of them since the Verizon tower here is horribly overloaded. That's 8 kb/sec download during prime time.
I got mine second hand and dont have a data plan currently but its the verizon one.
No, that's the regular Administrator password. It doesn't work here:
Code:
mifi login: Admin
Password:
Login incorrect
mifi login: admin
Password:
Login incorrect
mifi login: root
Password:
Login incorrect
mifi login: user
Password:
Login incorrect
mifi login:
This is on the virtual com port coming over the USB cable, not the front screen or the web access.
how are you accessing this com port when i boot and enable usb connection it doesnt appear in windows
That's the whole point of that memset/WriteFile stuff at the top. I've never had to resend that so I don't know if that's a permanent enable. It did disconnect and reconnect when I did it, so I think that it's real. If you're not showing the same as I see now, that confirms it.
Do you have UsbView.exe? It's here: https://www.ftdichip.com/Support/Utilities.htm#MicrosoftUSBView
This is the what I have (trimmed down to the essentials). Is your "Total Length 0x136" the same as me? Do you see 7 interfaces?
Code:
Device Descriptor:
idVendor: 0x1410
idProduct: 0xB023
bcdDevice: 0x0318
Configuration Descriptor:
wTotalLength: 0x0136
bNumInterfaces: 0x07
Interface Descriptor:
bInterfaceNumber: 0x00
0x0409: "RNDIS Communications Control"
Interface Descriptor:
bInterfaceNumber: 0x01
0x0409: "RNDIS Ethernet Data"
Interface Descriptor:
bInterfaceNumber: 0x02
(Something custom)
Interface Descriptor:
bInterfaceNumber: 0x03
(Something custom)
Interface Descriptor:
bInterfaceNumber: 0x0C
0x0409: "CDC Abstract Control Model (ACM)"
Interface Descriptor:
bInterfaceNumber: 0x0D
0x0409: "CDC ACM Data"
Interface Descriptor:
bInterfaceNumber: 0x0E
0x0409: "HID Interface"
(You get extra credit for noticing that the interface numbers are not contiguous! (0, 1), 2, 3, (12, 13), 14. That breaks some Linux libs.)
Edit: Oh! My 8800 was originally 1410/b010, then when I did the HID thing it turned into 1410/b023. I just looked at my code.
So this is a Qualcomm processor, so apparently this is a Qualcomm DIAG port. I still can't find a password, but I strongly suspect that "root" is the (only useful) user.
Ok, so I managed to reset it back to "Normal", a/k/a 1410/b010
I see that in this mode there is only one interface, that HID.
So I wrote a little utility:
Code:
C:\>mifilink /?
MiFi Configuration Utility
mifilink <cmd>
cmd = normal, diag
C:\>mifilink
Found MiFi 8000/8800 (normal mode)
C:\>mifilink diag
Found MiFi 8000/8800 (normal mode)
Sending Diag (0) ok
C:\>mifilink
Could not find any MiFi devices
C:\>mifilink
Found MiFi 8000/8800 (DIAG mode)
When it couldn't find anything that was because it was rebooting. I put that in there to demonstrate that the utility knows that. Without a command it just shows what it sees.
Note: This is not a zip file. Don't unzip it, just rename it mifilink.exe
I had a question on how I was powering this. Just to clarify:
I'm using a USB cable that I cut up. Normally the color code is to be trusted, but check with a voltmeter if you have it.
Black is ground, green is D+ (unused), white is D- (unused), red is Vbus i.e. +5V.
The photo in the OP shows it all.
From the bottom is the ground connection to black.
Then the 100k resistor for the thermistor to ground (to make the device think that the temperature is reasonable).
Then the 51k resistor to ground to fake the device out to thinking that a battery is connected.
Then the 1 Amp diode (like a 1N4001 to 1N4007) to drop the voltage down a bit to make life easier for the device.
The banded end (i.e. the cathode) goes to the contact, the plain end (i.e. the anode goes to the red wire.
If you don't like murdering your device you could take a block of something (wood, plastic) and put four contacts on it.
Then you could leave your device unmodified. I've made some blocks for cell phones that I have for testing.
Here's a Moto E6 that I was trying to get into EDL mode. The plug has one resistor in it for thermistor.
Renate said:
I had a question on how I was powering this. Just to clarify:
I'm using a USB cable that I cut up. Normally the color code is to be trusted, but check with a voltmeter if you have it.
Black is ground, green is D+ (unused), white is D- (unused), red is Vbus i.e. +5V.
The photo in the OP shows it all.
From the bottom is the ground connection to black.
Then the 100k resistor for the thermistor to ground (to make the device think that the temperature is reasonable).
Then the 51k resistor to ground to fake the device out to thinking that a battery is connected.
Then the 1 Amp diode (like a 1N4001 to 1N4007) to drop the voltage down a bit to make life easier for the device.
The banded end (i.e. the cathode) goes to the contact, the plain end (i.e. the anode goes to the red wire.
If you don't like murdering your device you could take a block of something (wood, plastic) and put four contacts on it.
Then you could leave your device unmodified. I've made some blocks for cell phones that I have for testing.
Here's a Moto E6 that I was trying to get into EDL mode. The plug has one resistor in it for thermistor.
Click to expand...
Click to collapse
So I would recommend something slightly different, use a tp4056 to charger the battery, they don’t get very hot and it bypasses the internal charger and usb c port. So you could use that for data still.
Renate said:
I had a question on how I was powering this. Just to clarify:
I'm using a USB cable that I cut up. Normally the color code is to be trusted, but check with a voltmeter if you have it.
Black is ground, green is D+ (unused), white is D- (unused), red is Vbus i.e. +5V.
The photo in the OP shows it all.
From the bottom is the ground connection to black.
Then the 100k resistor for the thermistor to ground (to make the device think that the temperature is reasonable).
Then the 51k resistor to ground to fake the device out to thinking that a battery is connected.
Then the 1 Amp diode (like a 1N4001 to 1N4007) to drop the voltage down a bit to make life easier for the device.
The banded end (i.e. the cathode) goes to the contact, the plain end (i.e. the anode goes to the red wire.
If you don't like murdering your device you could take a block of something (wood, plastic) and put four contacts on it.
Then you could leave your device unmodified. I've made some blocks for cell phones that I have for testing.
Here's a Moto E6 that I was trying to get into EDL mode. The plug has one resistor in it for thermistor.
Click to expand...
Click to collapse
Thanks for this. it worked like a charm. This was the best solution for me even after trying two of Netgear's LTE Modems and Modem/Router. Those devices frequently lost connection to Verizon. Neither Verizon nor Netgear could help me.
There was a question on firmware updatting the 8000/8800
I tried a bit to find EDL test points on my 8800. Since this is my main connection that was in use I didn't try too hard at the time.
I had an Orbic Speed that I got under warranty and don't use so I could attack it at my leisure.
I found the EDL test points. It uses Red Hat Linux. Since I'm more Android and don't need to do anything I dropped the matter there.
This is all related here: https://forum.xda-developers.com/t/...-firmware-flash-kajeet.4334899/#post-86616269
(My main device, an Onyx Boox Poke3 ereader is also modified with a reed switch as I do a lot of slinging partitions around.)
Can you still power via the USB port and ommit the DC at the battery terminal?
If not, would a capacitor work in the batteries place?
You need active power coming in on the battery terminal. If you just want to just use the USB port you can jumper from the USB through a diode to the battery in. Of course you still need the two resistors. There is a test point near the USB connector where you can pick off power. I don't have a photo off-hand.
Renate said:
There was a question on firmware updatting the 8000/8800
I tried a bit to find EDL test points on my 8800. Since this is my main connection that was in use I didn't try too hard at the time.
I had an Orbic Speed that I got under warranty and don't use so I could attack it at my leisure.
I found the EDL test points. It uses Red Hat Linux. Since I'm more Android and don't need to do anything I dropped the matter there.
This is all related here: https://forum.xda-developers.com/t/...-firmware-flash-kajeet.4334899/#post-86616269
(My main device, an Onyx Boox Poke3 ereader is also modified with a reed switch as I do a lot of slinging partitions around.)
Click to expand...
Click to collapse
There's three test pads on upper right of the 8800 board (the same side as the battery). Shorting the left pad to the middle pad and plugging in a USB cable results in it booting in EDL mode.
mjg59 said:
There's three test pads on upper right of the 8800 board (the same side as the battery). Shorting the left pad to the middle pad and plugging in a USB cable results in it booting in EDL mode.
Click to expand...
Click to collapse
Unfortunately there doesn't seem to be a publicly available loader that'll work here
Renate said:
I had a question on how I was powering this. Just to clarify:
I'm using a USB cable that I cut up. Normally the color code is to be trusted, but check with a voltmeter if you have it.
Black is ground, green is D+ (unused), white is D- (unused), red is Vbus i.e. +5V.
The photo in the OP shows it all.
From the bottom is the ground connection to black.
Then the 100k resistor for the thermistor to ground (to make the device think that the temperature is reasonable).
Then the 51k resistor to ground to fake the device out to thinking that a battery is connected.
Then the 1 Amp diode (like a 1N4001 to 1N4007) to drop the voltage down a bit to make life easier for the device.
The banded end (i.e. the cathode) goes to the contact, the plain end (i.e. the anode goes to the red wire.
If you don't like murdering your device you could take a block of something (wood, plastic) and put four contacts on it.
Then you could leave your device unmodified. I've made some blocks for cell phones that I have for testing.
Here's a Moto E6 that I was trying to get into EDL mode. The plug has one resistor in it for thermistor.
Click to expand...
Click to collapse
Can you power an SSD via the USB-C (for file sharing feature 8800L) still with doing the power where the battery would be going? Or do I need to send 5v down the USB-C wire since the battery input would be ~3.3v?
pcoplen said:
So I would recommend something slightly different, use a tp4056 to charger the battery, they don’t get very hot and it bypasses the internal charger and usb c port. So you could use that for data still.
Click to expand...
Click to collapse
Would you care to share how the tp4056 would hook up to the charging pin? This sounds like the way I would need to go in order to power an SSD via USB-C port.
stealthrt said:
Can you power an SSD via the USB-C (for file sharing feature 8800L) still with doing the power where the battery would be going? Or do I need to send 5v down the USB-C wire since the battery input would be ~3.3v?
Click to expand...
Click to collapse
The device has a boost converter to take battery power and boost it to 5V for the Type C when you are using a flash drive.
You don't need to do anything.
If you did route 5V to the Type C yourself it would say, "Oh, I'm connected to a charger, of course there is no flash drive."
I can plug in a flash drive, there is 5V supplied to it, it's takes 70mA or so.
It is enabled on the MiFi Share setting page.
I can't find it on my Windows machine. Maybe because I disabled so much of that net sharing.
If you really wanted to keep the battery, yes, you could throw in a TP4056 charger board and have it charge your battery.
stealthrt said:
Would you care to share how the tp4056 would hook up to the charging pin?
Click to expand...
Click to collapse
Just connect the two battery pads on the PCB to the + and - of the battery.
I tried the EDL, the test points are as stated, "Left" in the instructions means towards the middle of the device.
Code:
HWID: 0007d0e100000000
Hash: 62b4a62f72d6c323
I couldn't find a loader either.
As I've said before I think this device runs hot.
That's why I cannibalized it and added a heatsink.
Maybe this thing was tested at an ambient temperature of 20°C (68°F) but it just runs away at higher temperatures.
At 38°C (100°F) ambient temperature even with a heatsink this thing is a nuclear reactor.
I had a little 12V 40mm fan. I plugged it into 5V and that's enough to keep this thing cool when it gets hot.