Can anyone offer some clarification regarding PM2 (power management 2 mode)

Hi everyone,

Quite a few people or looking for an answer to this puzzle. I’ve asked around and made a pretty short and clear post on the question in a few places. Now that I’ve found a WD forum I’m trying my luck here ^_^. Here goes:

The main question is actually, what does an OS, system BIOS or RAID-card BIOS need to support staggered spin up?

The hard drive must obviously support some form of power management. The SATA standard seems to revolve around setting pin 11 on the power connector either high or floating. A drive that supports PIUS should then not spin up the motor until asked by the BIOS, OS, … Please refer to the attached PDF called " staggerd-spin-detection-pin11" . Then a PHY command should be issued and voila, the drive will spin up. Rinse and repeat for every drive.

Anyways, that does not seem to work with WD20EADS drives (we all know them). They do support some form of PIUS but call it power management 2 mode (PM2).
Please have a look here http://en.wikipedia.org/wiki/Power-up_in_standby .

  • Is this the same as the pin 11 technique?
  • Is the pin 11 technique a feature of APM or ACPI?

I do know Areca RAID-controllers don’t play well with these drives and cannot get them out of stand by(I know I can flash them and then they will work but that’s not the scope of my post). One thought from someone was that Western Digital uses ACPI and most vendors use APM to send the wake up command. Areca might not support ACPI thus is unable to wake up the drive.

If someone is wondering why I want staggered spin up: I don’t want to buy a big PSU just to allow the drives to start up. As you know the start up current of hard drives is a lot more than their IDLE/operating power.

Cheers,
Jeroen

I suspect that pin 11 provides spinup control via hardware, whereas PUIS provides the same control via software.

See sections 4.20, 4.19.4, 7.52.7, 7.52.8, 7.18 (bits 5 & 6 of words 83 & 86) of the following document.

Working Draft ATA/ATAPI Command Set - 2 (ACS-2):
http://www.t13.org/documents/UploadedDocuments/docs2010/d2015r3-ATAATAPI_Command_Set_-_2_ACS-2.pdf

AIUI, the PM2 jumper, if installed, overrides any software control over PUIS. Therefore it seems to be associated with the ATA PUIS feature rather than the pin 11 staggered spinup control.

When a drive powers up in standby, the BIOS or RAID controller can still detect it using the ATA Identify Device command, but the drive may not report all its features and settings at this time. Bit 5 of words 83 and 86 of the 512-byte Identify Device info block indicates whether PUIS is supported and whether it is enabled.

If PUIS is enabled, then there are two ways to wake the drive. If bit 6 = 0, then any command that requires access to the platters will spin up the drive. That is, any read or write command will suffice. OTOH, if bit 6 = 1, then the controller needs to issue an ATA Set Features subcommand (0x07) in order to spin up the drive.

The above document states that any changes made to the PUIS settings are nonvolatile, ie they survive a power cycle. (PUIS wouldn’t make any sense if the changes were volatile).

BTW, you can sometimes find expired web pages and document URLs in the Wayback Machine archives.

Staggered Spin-up Detection Using Pin 11:
http://web.archive.org/web/20051014111524/http://maxtor.com/_files/maxtor/en_us/documentation/white_papers_technical/staggerd-spin-detection-pin11.pdf

1 Like

I wonder why some SAS/RAID controllers do not work then. It seems to be an industry standard command?

Or are there different PUIS revisions (APM vs ACPI)? That is still a puzzling part:)

Kudo’s for the informative reply +

AIUI, PUIS is an industry standard ATA feature set, but it is up to the manufacturer to decide whether or not to implement it.

If I were investigating whether your setup supports staggered spinup via pin 11, I would test for continuity between pin 11 and ground at the controller/PSU end. If this pin is hardwired to ground, then obviously there is no spinup control. If this pin is open, then I would check whether all such pins are wired together, in which case the drives would spin up in unison. If all pins are independent, then this would suggest that hardware control is present, in which case support will then be dependent on the firmware.

I can’t answer your questions regarding APM and ACPI.