after some researches i found that motherboard of F2-220 is the same.
i'm thinking to buy ( if i find one) secondhand F2-220 and use it in my box.
@TMnorah
did you manage to recover your data ?
F4-220 - Start problem
Re: F4-220 - Start problem
Hi @kavdiev,
This is most likely the haunted Atom C2000 series Bug that is also applicable to intel baytrail SOC based products. It is a known bug of all Intel atom series CPUs which has been repeated with later apollo lake devices. So all Terramaster 220, 221, 421 etc. are affected. On the 423 and 223 we don't know yet if they are affected as the Intel jasper lake CPUs are relatively new.
Intel only admitted the problem for embedded SKUs, Servers and later also IoT markets, which all got a fixed stepping. There is no fixed stepping for NAS and consumer CPUs by intel because "these are not supposed to last that long".
The symptom is that the system hangs in post (no boot, no image) state because the clock on the LPC bus cannot be detected. LPC is used for example for TPM modules and is not in use by Nas systems as far as i know.
Luckily there is a known fix. Even Cisco did apply this dirty patch for their embedded network devices using the C2000 series CPUs.
- The fix applied by Cisco is by putting a 110 ohm pull-up resistor from either LPC_CLKOUT0 or LPC_CLKOUT1 to +3.3V
Synology users report a 100ohm resistor between 3,3V and the LPCclockOUT works fine to revive all those dead DS 414+ motherboards.
Now we only need the Pinout from @TMsupport as sadly the 12pin single row LPC_COM used by Terramaster is not a standard header.
If you are in the process of looking up the pinout, the VGA one would also be highly appreciated! The 12Pin VGA Dongles seem to be sould out at Alibaba and i don't know how to guess the Pinout.
We can measure the 3v3 easily, but to find out the LPC-CLock Pin probably requires an oscilloscope.
best regards, Maks
This is most likely the haunted Atom C2000 series Bug that is also applicable to intel baytrail SOC based products. It is a known bug of all Intel atom series CPUs which has been repeated with later apollo lake devices. So all Terramaster 220, 221, 421 etc. are affected. On the 423 and 223 we don't know yet if they are affected as the Intel jasper lake CPUs are relatively new.
Intel only admitted the problem for embedded SKUs, Servers and later also IoT markets, which all got a fixed stepping. There is no fixed stepping for NAS and consumer CPUs by intel because "these are not supposed to last that long".
The symptom is that the system hangs in post (no boot, no image) state because the clock on the LPC bus cannot be detected. LPC is used for example for TPM modules and is not in use by Nas systems as far as i know.
Luckily there is a known fix. Even Cisco did apply this dirty patch for their embedded network devices using the C2000 series CPUs.
- The fix applied by Cisco is by putting a 110 ohm pull-up resistor from either LPC_CLKOUT0 or LPC_CLKOUT1 to +3.3V
Synology users report a 100ohm resistor between 3,3V and the LPCclockOUT works fine to revive all those dead DS 414+ motherboards.
Now we only need the Pinout from @TMsupport as sadly the 12pin single row LPC_COM used by Terramaster is not a standard header.
If you are in the process of looking up the pinout, the VGA one would also be highly appreciated! The 12Pin VGA Dongles seem to be sould out at Alibaba and i don't know how to guess the Pinout.
We can measure the 3v3 easily, but to find out the LPC-CLock Pin probably requires an oscilloscope.
best regards, Maks
Re: F4-220 - Start problem
Thanks for all your tips/info.
I'll try to search for pinouts.
Regarding VGA it's quite easy.
On back of mainbord you have name of each pin
Than for VGA cable you have Wikipedia. If you cut existing cable VGA cable you need to play with multimeter to find what wire is corresponding to what pin. Thick ones are for RGB. Others are for syncv and h. Lot of grounds... And 3 are unused ( 15pin to 12)
https://images.app.goo.gl/kH5fXZRBKLmzFMrV6
I'll try to search for pinouts.
Regarding VGA it's quite easy.
On back of mainbord you have name of each pin
Than for VGA cable you have Wikipedia. If you cut existing cable VGA cable you need to play with multimeter to find what wire is corresponding to what pin. Thick ones are for RGB. Others are for syncv and h. Lot of grounds... And 3 are unused ( 15pin to 12)
https://images.app.goo.gl/kH5fXZRBKLmzFMrV6
- Attachments
-
- VGA pinout
Re: F4-220 - Start problem
Regarding resistor, it should be connected permanently?
and for QNAP TS-251 i saw this -> Solder 100 ohm resister between CLK and GND on connector LPC-CN1.
why here between clk and +3.3V ? and not ground?
Regards,
and for QNAP TS-251 i saw this -> Solder 100 ohm resister between CLK and GND on connector LPC-CN1.
why here between clk and +3.3V ? and not ground?
Regards,
Re: F4-220 - Start problem
As you can (hopefully) see on this picture,
only two pins are connected. i need to check with power to be sure.
and find a resistor.
only two pins are connected. i need to check with power to be sure.
and find a resistor.
- Attachments
-
- IMG_20221010_234850166.jpg (21.58 KiB) Viewed 2259 times
-
Re: F4-220 - Start problem
Thank you for the VGA Pinout!
Posting pics and detailed description is a great way to leave breadcrumps also for those who are googling this some time later.
Most machines will probably fail in a few years and people will be glad if they can find this here.
I don't think that only two pins are connected because this is definitely a 4 layer board. All other connections are in the other layers.
the 100ohms to 3,3V is a pullup resistor because the clock signal level gets too low to be detected. There should be an internal pullup like an any µC but this is probably what is faulty over time in the atom cpus.
Posting pics and detailed description is a great way to leave breadcrumps also for those who are googling this some time later.
Most machines will probably fail in a few years and people will be glad if they can find this here.
I don't think that only two pins are connected because this is definitely a 4 layer board. All other connections are in the other layers.
the 100ohms to 3,3V is a pullup resistor because the clock signal level gets too low to be detected. There should be an internal pullup like an any µC but this is probably what is faulty over time in the atom cpus.
Re: F4-220 - Start problem
yes the resistor should be there permanently. so you can solder it to the pins from the back side.
It would probably be sufficient to place it only during the POST but that would be more complicated than to fix it permanently. your device would not reboot after a power outage if you don't put the resistor back in. So just keep it permanent like everybody else does.
Btw, it will be hard to find an oscilloscope that goes to 33mhz which is the speed of the LPC bus clock
(some sources say 24mhz, but i think its 33)
It would probably be sufficient to place it only during the POST but that would be more complicated than to fix it permanently. your device would not reboot after a power outage if you don't put the resistor back in. So just keep it permanent like everybody else does.
Btw, it will be hard to find an oscilloscope that goes to 33mhz which is the speed of the LPC bus clock
(some sources say 24mhz, but i think its 33)Re: F4-220 - Start problem
My oscilloscope has only 25mhz bandwidth, and those with 200mhz are mostly 400-500$+
BTW: Is it possible in this forum to edit your own posts?
BTW: Is it possible in this forum to edit your own posts?
Re: F4-220 - Start problem
I was thinking about multilayer.
t this point,
I have nothing to loose. So I'll try to add a resistor randomly
Will keep you posted.
t this point,
I have nothing to loose. So I'll try to add a resistor randomly
Will keep you posted.
Re: F4-220 - Start problem
Be careful! I personally would not risk it but rather ask somebody with a 200mhz Oscilloscope to measure it quickly for me to be sure.
Have you measured which pin the 3.3v is? There is a 3V standby and a 3.3V.
You probably want the 3.3V as i don't know if the 3V standby is sent to sleep once the system is awake and we don't know the allowed standby current too.
On most dual row headers on other Motherboards the Clock is on Pin1. This could be the case here or not. If pin1 is neither gnd nor 3v standby nor 3.3v you have a certain chance that it is the clk.
Have you measured which pin the 3.3v is? There is a 3V standby and a 3.3V.
You probably want the 3.3V as i don't know if the 3V standby is sent to sleep once the system is awake and we don't know the allowed standby current too.
On most dual row headers on other Motherboards the Clock is on Pin1. This could be the case here or not. If pin1 is neither gnd nor 3v standby nor 3.3v you have a certain chance that it is the clk.