I have a customer that brought in a Lenovo Legion S7 15ACH6. The problem is that the laptop is not posting.
My first thought was to take out the DRAM and put it back in.
No POST.
Then I swapped the dedicated RAM with good known working RAM.
No change.
I then measured to see if 1.2V was being generated for the memory to turn on, so I measured PL600 which is where the output of the 1.2V rail is.
Resistance was good; measuring ~40Ohms with the slotted RAM in.
I took the slotted memory out and measured again.
Resistance with DRAM out was ~100 Ohms and the voltage is still missing with the DRAM out.
-----------------------------------------------------------------------------------------------------------------------------------------------------
This is where I should have stopped but I had an idea that might have cost me the repair because I was curious.
-----------------------------------------------------------------------------------------------------------------------------------------------------
I took a chance and removed the onboard RAM and measured PL600 again.
Resistance jumped up to around ~400Ohms and the 1.2V is now present. I put the DRAM back in and measured the coil without the onboard memory present. 1.2V is present and resistance is 40Ohms (Hynix DRAM).
Then I took a gander around the forums and SMDFlea pointed out that there was a script to disable the onboard memory. So I attached my programmer, dumped the BIOS, made a copy for backup, and ran the script.
It changed 32 items, so I saved the BIOS and flashed it.
No change.
I then tried the script with a stock BIOS.
Still not posting but all voltage rails are present with good resistances.
CPU and GPU get warm but do not get nearly as hot as they should for a normal system power on.
My only choice at this point is to reball and reattach the onboard memory chip by chip and check the resistance each time.
The problem is that during the cleanup process (you have to remove the hard black glue that is on each corner of the RAM), one of the pads came loose and fell off on URAM3 Pad 34. This looks like a data line, and luckily the trace is still there so I can run a pad strip.
Does anyone have any ideas other than reballing and resoldering the onboard RAM? I am thinking that the script is not working on the BIOS and the onboard memory BGA pads are not getting disabled, thus the laptop is still looking for the onboard memory and since it is off the board, it will not work.
Other than the script not working, an APU/chipset failure is what makes the most sense right now.
Boardview and schematic are available on this forum. Board number is BM5070. The board that I have is a V1.4, and V1.0 is what is available on this forum.
My first thought was to take out the DRAM and put it back in.
No POST.
Then I swapped the dedicated RAM with good known working RAM.
No change.
I then measured to see if 1.2V was being generated for the memory to turn on, so I measured PL600 which is where the output of the 1.2V rail is.
Resistance was good; measuring ~40Ohms with the slotted RAM in.
I took the slotted memory out and measured again.
Resistance with DRAM out was ~100 Ohms and the voltage is still missing with the DRAM out.
-----------------------------------------------------------------------------------------------------------------------------------------------------
This is where I should have stopped but I had an idea that might have cost me the repair because I was curious.
-----------------------------------------------------------------------------------------------------------------------------------------------------
I took a chance and removed the onboard RAM and measured PL600 again.
Resistance jumped up to around ~400Ohms and the 1.2V is now present. I put the DRAM back in and measured the coil without the onboard memory present. 1.2V is present and resistance is 40Ohms (Hynix DRAM).
Then I took a gander around the forums and SMDFlea pointed out that there was a script to disable the onboard memory. So I attached my programmer, dumped the BIOS, made a copy for backup, and ran the script.
It changed 32 items, so I saved the BIOS and flashed it.
No change.
I then tried the script with a stock BIOS.
Still not posting but all voltage rails are present with good resistances.
CPU and GPU get warm but do not get nearly as hot as they should for a normal system power on.
My only choice at this point is to reball and reattach the onboard memory chip by chip and check the resistance each time.
The problem is that during the cleanup process (you have to remove the hard black glue that is on each corner of the RAM), one of the pads came loose and fell off on URAM3 Pad 34. This looks like a data line, and luckily the trace is still there so I can run a pad strip.
Does anyone have any ideas other than reballing and resoldering the onboard RAM? I am thinking that the script is not working on the BIOS and the onboard memory BGA pads are not getting disabled, thus the laptop is still looking for the onboard memory and since it is off the board, it will not work.
Other than the script not working, an APU/chipset failure is what makes the most sense right now.
Boardview and schematic are available on this forum. Board number is BM5070. The board that I have is a V1.4, and V1.0 is what is available on this forum.