Tip: How to identify the Read Channel test points

[fzabkar]

Tip: How to identify the Read Channel test points

The read and write differential pairs are identifiable as two thin parallel signal traces running between the HDA connector and the MCU. They are terminated in resistors at the MCU end. The read terminator connects across the differential pair, while the write pair has a resistor on each side. Each pair of traces is of equal length, and they terminate at the HDA pads in a characteristic "Y" pattern.

Here are several examples:
http://www.users.on.net/~fzabkar/HDD/RW_TestPoints/

Some newer PCBs have no visible termination resistors on either the read or write differential pair. In such cases it may be possible to identify the read channel by measuring the two signal pairs both on and off the HDA.

The write channel is driven by the MCU whereas the read channel is driven from the preamp. When the PCB is off the drive, the voltage on the read channel test points will be 0V with respect to ground whereas the write test points will be 1V or thereabouts (possibly 1.8V). When the PCB is on the drive, the read channel test points will now rise to around 1V or so, and the write channel will settle into a similar range.

I'm not certain if the above observation is applicable to all drives, but I have verified it on a Hitachi.

Original article:
http://malthus.zapto.org/viewtopic.php?t=123&p=1632

Attachments

WD800BB-53CAA1_RW_TP_comp.JPG (121.91 KiB) Viewed 15517 times

WD800BB-53CAA1_RW_TP_bot.JPG (58.28 KiB) Viewed 15517 times

ST3120022A_RW_TP_comp.JPG (81.77 KiB) Viewed 15517 times

ST3120022A_RW_TP_bot.JPG (95.79 KiB) Viewed 15517 times

SP2514N_RW_TP_comp.JPG (81.05 KiB) Viewed 15517 times

DiamondMax16_80GB_RW_TP_top.jpg (89.34 KiB) Viewed 15517 times

DiamondMax16_80GB_RW_TP_bot.jpg (53.8 KiB) Viewed 15517 times

[HaQue]

Thanks for posting that here, I was looking for it the other day while Spildit's was down.

Could you please clarify what is commonly refered to as Shorting the Read Channel?

Does that mean just put tweezers ove the 2 read test points, or you actually ground them? What is the shorting supppsed to achieve, is it a prposely built in feature to enable added functionality to Terminal, or enable it when disabled, or it tricks the MCU in some way?

[fzabkar]

The read channel is the data path between the MCU and the preamp. It is also the path that servo data takes. AIUI, the read channel is interrupted by bridging the read pair with tweezers, or by introducing a noise signal into the signal path. I believe one can also add an RC filter across the pair under certain circumstances.

AFAIK, this procedure is used to prevent access to the SA at appropriate points in the power-up process. For example, the 7200.11 BSY bug had a corrupt SA module that prevented the drive from reaching readiness.

[fzabkar]

Here is a Toshiba/Hitachi PCB:

DT01ACA300_HDA.jpg (179.42 KiB) Viewed 15424 times

There is a kink on one side of each R/W channel pair. These kinks are intended to equalise the lengths of each trace.

[fzabkar]

Here is a Seagate 7200.12 PCB:

Seagate_100535704_RW_Channel.jpg (116.08 KiB) Viewed 15384 times