I found differences in hdparm -i (lower case) and -I (upper case):
hdparm -i /dev/sda display unknown settings
hdparm -I /dev/sda display Advanced power management level: 96

While quietHDD set APM (255) hdparm display:

hdparm -i /dev/sda display disabled (255)
hdparm -I /dev/sda display disabled

I did some Mod 02 comparison with fzabkar's tool, and it could be Section 39 Offsets 4-5. Yours is 0xC800 (200) while I have 0x5802 (600) on my 2.5" and 0x0000 on my 3.5".

Now my problem is, I don't have any old hard disks that has Section 39 and I don't want to brick my 1TB playing with this.

If you feel adventurous, search for this in the source code :
buffer[0]=0x2a;
buffer[2]=rw;
buffer[4]=0x02;
buffer[6]=0x0d;
buffer[8]=0x36;
buffer[10]=0x02;

and change it to :
buffer[0]=0x2a;
buffer[2]=rw;
buffer[4]=0x02;
buffer[6]=0x27;
buffer[8]=0x04;
buffer[10]=0x02;

Recompile and do a "-g" and you should see 0xC8 (equivalent to 200). Try changing it to 600 (-s 600) and power off-on your hard disk to see if it stops parking. If it doesn't, try disabling it and power off-on again. To undo, use "-s 200".

I haven't really tried hdparm -i before, so I don't really notice that. I'll check out later.

Ensure your idle3 is disabled or it'll interfere with your tests.

Thanks for your help but I also do not want to make bricks out of my 320GB

From what I understand we have 3 different options related to APM:

• Idle Timer - possibe set with wdidle3 (DOS)
• APM Timer - possible set with APMtimer (Linux)
• APM level - possible set with quietHDD (Windows) or hdparm -B255 (Linux or Windows)

Am I right?

Only 2, first is the timer that controls the intellipark and the other has something to do with spin-down.

Idle3 is not related to APM but if it triggers intellipark before APM, then we won't know the actual timer value for APM. The APM default should be 20 seconds, so if idle3 kicks in at <20 seconds, the head would've been parked before 20 seconds.

APM level relies on the APM settings. Those set by quiethdd and hdparm are not persistent upon power cycles.

I tried to set the Idle Timer to 255 (3810s = 63.5 min) using idle3ctl (linux) but any value does not change the behavior of the disk. Interestingly, using this tool can set the original value Idle Timer of 3 seconds:

Now wdidle3 in DOS display 3 seconds

Just did some test and it refreshed my memory.

Model=WDC WD1600BEVS-08VAT2
Advanced power management level: unknown setting (0x0080)

I suspect it's unknown because there are 2 bytes returned by the disk. It should be 1 byte, I think.

So if you want to restore your hard disk to factory, run apmtimer again and set it to 20 seconds.

If I could source a newly made hard disk, I will try patch Section 39 and see if helps. My year 2012 320gb does not have a Section 39, while 1tb made in 2014 is too costly to be risked.

This is what hdparm's wdidle3.c says :
* Older drives (from WDIDLE3.EXE documentation):
* 0x00 timer disabled.
* 0x50-0xff timeout in tenths of a second, from 8.0 to 25.5 seconds.
*
*
* Some drives (found by observation):
* 0x01-0x50 timeout in seconds, from 1 to 30 seconds (?).
*
* Newer drives (from WDIDLE3.EXE documentation):
* 0x00 timer disabled.
* 0x01-0x49 not permitted by WDIDLE3.EXE; reports as 0.1 to 7.9 seconds.
* 0x50-0x7f timeout in tenths of a second, from 8.0 to 12.7 seconds.
* 0x80 not permitted by WDIDLE3.EXE; reports as 12.8 seconds.
* 0x81-0x8a timeout in 30s of seconds, from 30 to 300 seconds.
* 0x90-0xff not permitted; reports as 330 to 3810 seconds.

I managed to borrow a hard disk to test with, smaller capacity which is less costly if I damage it

It's made some time at the end of 2013, few months newer than yours. So my program is able to disable the APM timer, hence no autopark at 20 seconds. APM goes from 96 to 128.

I then undo the change above and went ahead to patch Section 39 (0x27) Offsets 4-5 from 0x5802 to 0xC800, power cycled the hard disk, yet APM stays at 128 and there was no autopark. Your Section 39 Offsets 4-5 has a value of 0xC800 instead of 0x5802 in the newly made ones. There is a fair chance that the APM timer is stored there, so I'll leave it to you to decide if you want to test further. Changing these values did not brick my hard disk.