ASUS E35M1-I Deluxe

Hi all,
can somebody tell me good settings for the BIOS of the
ASUS E35M1-I Deluxe Board.?

I found something about the Pro Motherboard,
but nothing about the ASUS E35M1-I Deluxe.

I played a litle bit with the settings, but all time with bad jitter.
At the moment I am at [Servo] 24004ns and [Base] 42730ns
(starting at [Servo] 9000ns and [Base] 17300ns).

I believe, that it must be possible to get results better than 9000ns, but what to do?

On the web-page wiki.linuxcnc.org/cgi-bin/wiki.pl?Latency-Test this chip-set had
[Servo] 12009ns and [Base] 9198.
When I perform the stress-test: stress -c 16 -i 16 -m 16 -t 14400 it starts at the moment at
[Servo] 3421ns and [Base] 3374, but after a few minutes it increase to more than 50000ns for both.

I looked for the SMI-Problem, but the chip-set isn't from Intel, so I don't know, if this is helpful for AMD-chipsets.

Additionally I include the isolcpu=1 into the /etc/default/grub, but it doesn't help.

On all different settings the LinuxCNC Program started with the Message:
RTAPI: ERROR: Unexpected realtime delay ... details in dmesg.
(but I have no idea, where to find the right screw to turn)

Maybe somebody out there use this Board and has experience.
Please share your research.

thanks in advance,
bye,
Plasturo

can somebody tell me good settings for the BIOS of the
ASUS E35M1-I Deluxe Board.?
....
I played a litle bit with the settings, but all time with bad jitter.

What exactly did you do? You need to disable all power management, fan management (set it on full all the time so it does not need to check temperature), disable hyperthreading if applicable etc etc
Every BIOS is different, you will have to ask specifically if unsure about a particular option

Additionally I include the isolcpu=1 into the /etc/default/grub, but it doesn't help.

It should be isolcpus=1 (plural)

I looked for the SMI-Problem, but the chip-set isn't from Intel, so I don't know, if this is helpful for AMD-chipsets.

No, the SMI problem is not relevant to AMD, only affects specific intel chipsets

regards

Hi ArcEye,
yes, I need to be more specifically in asking question,
but I hoped there is somebody in the forum who knows this board.

However, your comments help me in knowing, that the SMI problem
is only linked to Intel Chipsets.

The isolcpu was a typing error during writing in the forum, in my GRUP setting it is/was okay.

All the time I was only looking in the "Advanced" part of the BIOS,
the setting for the fan management I wasn't thinking about.
Thanks again, now I had a look to the "Monitor" part of the BIOS.

I started to write down what I have done, to give you more exact what I tested so far.
But due to the fact, that I found the problem, I copy this now into the forum to hopefully
help somebody who runs into the same challenge.

BIOS setting at the moment:
- Advanced
CPU Configuration

• Limit CPUID Maximum Disabled
• Cool 'n' Quiet Disabled
• NX Mode Disabled
• SVM Mode Disabled
• C6 Mode Disabled

USB Configuration

• Legacy USB Support Enabled
• Legacy USB3.0 Support Disabled
• EHCI Hand-off Disabled

NB Configuration

• IGFX Multi-Monitor Disabled
• Primary Video Device PCIE Video
• Integrated Graphics Auto

Onboard Device Configuration

• HD Audio Device Disabled
• Realtek LAN Controller Enabled
• Realtek PXE OPROM Disabled
• Renesas USB3.0 Controller Disabled

APM

• Restore AC Power Loss Power Off
• Power On By PS/2 Keyboard Disabled
• Power On By PS/2 Mouse Disabled
• Power On By PME Disabled
• Power On By RTC Disabled

- Monitor
Monitor

• CPU Q-Fan Control Disabled <= today, now the Fan runs at full speed all the time
• Anti Surge Support Disabled <= today, even if I don't know what that means

Start Test
[Servo] 7861
[Base] 16416
Start stress
after some minutes during starting 'sleep 600' in a new terminal the computer goes up to
[Servo] 8586
[Base] 16416
moving some windows, starting Firefox and SoftwareUpdateManager
[Servo] 16220
[Base] 16477
After 10 minutes
starting SoftwareUpdate and abort it
[Servo] 17845
[Base] 19852 <= more or less a good value

Switching off the 'isolcpus=1' in /etc/default/grub // reboot
After a view seconds
[Servo] 46382
[Base] 49610 <= a worse value
okay, not a good idea, so
switching back to the 'isolcpus=1' on in /etc/default/grub // reboot

Then under real stress, the board gave me the following values:
[Servo] 19064
[Base] 16885 <= more or less a good value

stopping Latency test
The maximum I have now, is [BASE] 16885 and [SERVO] 19064.

Switching on my parallel Port Hardware (myOwnMotionController)
Starting LinuxCNC
and .... after a few seconds the Message comes up:
RTAPI: ERROR: Unexpected realtime delay on :task 1
the dmesg tells me:
In recent history there are
1674498, 1673880, 1674432, 1674609, and 1674156
elapsed clocks between calls to the the motion controller.
This time, there were 2105031 which is so anomalously
lage that it probably signifies a problem with your
realtime configuration. For the rest of this run of
EMC, this message will be supressed.

I read, that someone switch off the LAN before starting linuxCNC,
now I will try the same:

Stopping Ethernet LAN
Starting LinuxCNC
and .... after a few seconds the Message comes up:
RTAPI: ERROR: Unexpected realtime delay on :task 1
the dmesg tells me:
In recent history there are
1674039, 1673273, 1674510, 16742255, and 1674465
elapsed clocks between calls to the the motion controller.
This time, there were 2179062 which is so anomalously
lage that it probably signifies a problem with your
realtime configuration. For the rest of this run of
EMC, this message will be supressed.

Switching off the Ethernet in BIOS

- Advanced
Onboard Device Configuration

• HD Audio Device Disabled
• Realtek LAN Controller Disabled <= this is the additional change
• Renesas USB3.0 Controller Disabled

Save BIOS-Settings and reboot
Start Test
[Servo] 6318
[Base] 8226
stoping latency test after a few secondes

Starting LinuxCNC
during pressing on eStop on the linuxCNC window,
again the RTAPI: ERROR: Failure comes up.

After restart of LinuxCNC I wait some minutes without touching the userinterface of linuxCNC,
but again the RTAPI: ERROR: Failure comes up.

I checked how to disable the Wifi, but in the manual was nothing written.
In the web (www.forum-3dcenter.org/vbulletin/archive...ex.php/t-511188.html)
I found the information, that this board has no Jumper and no setting in the BIOS,
but the WiFi-Board can be removed. Two screws, that's it.

I open the case, remove the WiFi-Board an start again the computer plus LinuxCNC.
The RTAPI: ERROR: doesn't appear.!!! yes.!!! Perfect.!!!


Switching on the Ethernet in BIOS

- Advanced
Onboard Device Configuration

• HD Audio Device Disabled
• Realtek LAN Controller Enabled <= this is the change
• Realtek PXE OPROM Disabled
• Renesas USB3.0 Controller Disabled

Save BIOS-Settings and reboot
Start Test
[Servo] 7109
[Base] 14941
Starting LinuxCNC, the Problem doesn't appear any more.

Starting Latency-Test
[Servo] 4571
[Base] 8051
starting stress for 10 minutes
moving the mouse
[Servo] 10936
[Base] 15680
at the end of the stress test
[Servo] 10936
[Base] 16541

Conclusion:
<<< The WiFi Board was the problem, it has no jumper nor bios setting to switch it off. >>>
<<< It is a need to removed it by unbold two screws, that's it >>>

Now I can go on programming my hardware ...
bye,
Plasturo

Well done. Got there eventually.

WiFi causes all sorts of problems sometimes, my new desktop computer had a PCI-e wifi card in it, that is obviously the way things are going.
It was completely unnecessary and I never got it working properly, so I just removed it and put it on ebay.
(still not sold!)

regards