Tyan S2892: Missing fan speeds, Power voltage, and 1 of the 2 cpu is missing. about open-hardware-monitor

[GoogleCodeExporter]

Thx for the report. The missing super I/O chip (mainboard fan speeds, voltages,
temperatures) is most likely a W83627HF with revision 0x41. It should be easy 
to add,
as revisions 0x17 and 0x3A are supported already.

Multi CPU support is what I am working on right now, cf. Issue 41.

Original comment by moel.mich on 23 Apr 2010 at 1:08

• Changed title: Tyan S2892: Missing fan speeds, Power voltage, and 1 of the 2 cpu is missing not even showing as a second core.
• Changed state: Accepted

[GoogleCodeExporter]

Hi and Thx for the fast response!

Here is more data on the Main board Chip-set
•Nvidia nForce pro2200
•AMD 8131™ PCI-X Tunnel
•Winbond W83627HF Super I/O
•Analog Devices ADT7463
Hardware Monitoring IC

Just in case you need more specs here is the manual for it:
http://www.tyan.com/manuals/m_s2892_102.pdf

Original comment by [email protected] on 23 Apr 2010 at 10:25

[GoogleCodeExporter]

The latest revision r117 should add support for your revision of the W83627HF.

Original comment by moel.mich on 24 Apr 2010 at 8:14

[GoogleCodeExporter]

I have just uploaded the new version 0.1.29 beta, which should fix both your 
problems
(hopfully). It would be great if you can give it a try and upload an
OpenHardwareMonitor.Report.txt.

Original comment by moel.mich on 24 Apr 2010 at 9:02

• Changed title: Tyan S2892: Missing fan speeds, Power voltage, and 1 of the 2 cpu is missing.

[GoogleCodeExporter]

Hi,

Well, it sort of worked...  Things loaded and appeared to be detected ok then it
issued an error page and died.  Several attempts resulted same.

Cant get a printed report but I have an image of the one it created 
automatically...
Here is a pic of it:

Original comment by [email protected] on 26 Apr 2010 at 11:50

Attachments:

• error.png

[GoogleCodeExporter]

The problem seems to be with detection of the secondary processor.  It fails to
provide accurate data...

Original comment by [email protected] on 26 Apr 2010 at 11:54

[GoogleCodeExporter]

Thx. Yes there have been a few more bugs, which I hopefully fixed with the 
latest
version 0.1.31 beta. Please let me know how things run with that version.

Original comment by moel.mich on 26 Apr 2010 at 7:41

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Flawless!
Thanks A Bunch!

The only thing I noticed is... It is still missing fan rotational speed for the 
power
suply, and individual processors, and for the GPU.  However having the 
temperature
indicated helps a great deal!

Original comment by [email protected] on 27 Apr 2010 at 11:39

Attachments:

• [missing Fans speed.png](https://storage.googleapis.com/google-code-attachments/open-hardware-monitor/issue-44/comment-10/missing Fans speed.png)

[GoogleCodeExporter]

Fan speed of the power supply can only be read if there is an extra connection 
from
the power supply to the mainboard (and is support only by very few power 
supplies). 

The monitor chip on the mainboard does support two more fan tacho channels, but 
there
seems to be no signal on these. So unless there are other tools that can read 
more
fans I would assume they are just not connected.

The NVAPI used to read the values from the GPU supports in theory reading back 
of the
fan tacho, but I haven't seen it supported by the driver on any system yet (or 
there
is still a bug).

Could you perhaps try and load only one of your CPUs (Task Manager -> Process
Affinity), in order to see if the temperatures sensor to CPU mapping is 
correct. That
means if really the CPU you put the load on gets warm.

Original comment by moel.mich on 27 Apr 2010 at 3:14

[GoogleCodeExporter]

Mmmm, Right Power supplies are as you stated.  Only on this mainboard all 
installed
parts are premium (Power supply does have a two wire lead to main board) and 
both
processor fans are using 3 wire lead to mainboard (one which is used for fan 
speed
monitoring and for regulating rotational speed based on BIOS thermal readings).

I tested the temperature as you requested via ProcessXP and under load and both 
cpu's
rating are actual... as they reflect differences in singular loads...

Original comment by [email protected] on 27 Apr 2010 at 4:11