VRM modify ICC_MAX to run high TDC OEM cpu
- Thread starter RolloZ170
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Okay, let me make another correction. The VRM pin for H3C C35 platform is J266
Among them, 1 scl, 2 gnd, and 3 sda
Because its PMBUS output is hung on the bridge, the multimeter cannot detect it.
Among them, 1 scl, 2 gnd, and 3 sda
Because its PMBUS output is hung on the bridge, the multimeter cannot detect it.
Thank you, I found the reason. The workstation must be powered on, and then run the command.
scan don't show icc_max. try the mod. and you see.
"MCP2221a_iccmax_FF.exe -TPS53679 60 61"
ok, it working
tested 8171M work
8173M work
8259CL not work (maybe bios not support 82xx cpu) it boot but hang when memory init ...
Attachments
Last edited:
Just got a Cisco C240 M5, and I found the vrm chip (PXE1610CDN) which is close to a 4 pin header. I cannot read found the trace of the board on what’s leading to the pin. Is there a way to identify them without using fancy equipment? Any help is much appreciated!
Attachments
header must not have connection even if close to the VRM controller.
look at the circuit diagram from manufacturer, if you don't have it no.
what helps is: multimeter, magnifying glass or head magnifier.
note: there can be a 20-33ohm series resistor in SDA/SCL signal.
edit:
check if GND is connected to Ground (e.g. any shield)
update:
pinout is correct for Cisco C240 M5
The addresses are 70 and 72
Last edited:
@RolloZ170 Thanks for the good work. I have a X11DPi-NT Rev: 1.21. Is it impossible to perform this mod?
I saw the following 3 commands and I am not sure which one applies to me:
- MCP2221a_iccmax_FF.exe -MP2955A 20 REM X11SPM/X11DPi first
- MCP2221a_iccmax_FF.exe -MP2955A 21 REM X11DPi second
- MCP2221a_iccmax_FF.exe -MP2955A 20 21 REM X11DPi first second
Thanks!
1.21: this is the rev. with the VRM design bug, if blowed up, SM will replace it even after warranty period.
rev. 1.21 uses still TPS53679 (TI) controller not MP2955A (MPS)
do a
MCP2221a_iccmax_FF.exe -scan 10 7FBUT I DON'T RECOMMEND TO USE HIGH TDC PROCESSORS ON THIS REVISION.
( bec. it can blow up even with the std. SKUs )
Last edited:
Thanks for the info. What does the -scan option do? Does it read the addresses? If I really do perform the mod, do I need to run:
MCP2221a_iccmax_FF.exe -TPS53679When you say even standard SKUs can blow up the VRM for the 1.21 rev board, does it mean even if I stay within 205W TDP (but very close to the limit) as per specs, I risk blowing it too? Asking this as I was previously considering 8259CL for cheap because it is close enough to the 205W limit, but knowing this I might consider sticking with much lower TDP CPUs.
Can I mount an additional heat sink on the VRM for CPU1 on my board to minimize that issue?
For BIOS, am I correct in assuming that it does not matter what version I have when running this mod as long as it is one that supports my CPU?
Thank you for your response!
Last edited:
-scan 10 7F
scans from address 10 to 7F for vrm controllers. you need the address to programm the right ones.
e.g. if two TPS53679 are found note the addresses. if for example the 2 addresse are 20 and 21 you use this:
MCP2221a_iccmax_FF.exe -TPS53679 20 21sure. this rev. has a VRM design BUG
not a bad idea.
from rev. 2.0 and up the VRM controller was changed and Heatsink holes for CPU1 VRM powerstages added,
required if NOT used in a server chassis with jet engine FANs.
but even in an server chassis the VRM BUG results i a blown VRM power stage.
correct. there is no TDP block in the BIOS.