Difference between revisions of "How to put SATA in old ThinkPads"

From ThinkWiki
Jump to: navigation, search
m (HX-811 2.5-MSATA: probably Soarland Hexin)
(added more models affected by safety feature)
 
(6 intermediate revisions by 2 users not shown)
Line 1: Line 1:
=ATA-ATA bridge=
+
__NOTOC__
[https://wikidevi.com/wiki/ATA-ATA_bridge At least fourteen (14) ATA-to-ATA parallel-serial bridge controller microchips exist (ref WikiDevi).]  The role of such a chip is: to connect two AT Attachment physical busses, bridge communication between Serial and Parallel Transports; and to translate PATA and SATA commands, protocol, and possibly alter responses.  These ICs were primarily produced for embedded applications, storage expansion solutions, and integration in complex products to reuse designs (such as many PC products, including ThinkPad).  Industry wants consumers to continually consume newer junky stuff, rather than extend life of perfectly good things, so datasheets of these controllers were not made available to public.
+
This page gives an overview of several methods that can be used to provide an old model with IDE/PATA/ATA capabilities to use newer SATA/mSATA/M.2/U.2 devices, Please be aware that this page does not offer any information to extremely old models with ancient or obscure interfaces, such as SCSI.
  
Some cheap modular translator adaptor bridge boards exist.  With a SATA to PATA adapter, you can connect a new SATA peripheral device to PATA host ThinkPad.
+
==Options==
 +
===via ATA-ATA bridge controllers===
 +
These bridge controllers translate ATA-to-SATA or SATA-to-ATA commands, depending on their application. On other note, be wary of low quality controllers as they can potentially corrupt data transfers or fail prematurely. Some of them may also lack datasheets for some odd reason, and that may just be a sufficient red flag to steer clear of them.
  
=instead of SFF-8111=
+
For reference purposes, have a look at a [http://en.techinfodepot.shoutwiki.com/wiki/ATA-ATA_bridge list of ATA-ATA bridge controllers] to judge what you need. Some may be faster or more stable than the other, but that does not necessarily mean that it is of higher quality, as these controllers could be built into very unreliable adapters (mostly due to poor PCB work).
==HX-811 2.5-MSATA==
 
[[File:HX-811defect.jpeg|thumb|alt=pair of bare pads, circled]]
 
[[File:HX-811correct.jpeg|thumb|alt=pair of pads, covered by solder bridge]]
 
{{X40}}, {{X41}}, {{X41T}} use '''SFF-8111 1.8" drive form factor (60x70mm)''' HDD. A very cheap circuit board (marked 'HX-811 2.5-MSATA'; likely manufacturer Shenzhen Soarland & Hexin Technology Co., Ltd.;) can fit within this f.f.  A seller of this shoddy device might use photos of "QC PASSED" labels to deceive you. A manufacturer of this product has minimal quality assurance, and does not inspect the pieces before they exit factory.  One specimen was non-functional: a pair of pads were not electrically connected, leaving all circuitry not- powered.  This defect is easy to fix.  Connect the twin pads by placing a blob of solder, zero ohm surface-mount resistor, or other conductive material.
 
  
==N-1801==
+
Here is one shining example of said PCB work:<br>
Another 'from China off fleaBay' choice.
+
[[Image:HX-811defect.jpeg|256px]]
 +
[[Image:HX-811correct.jpeg|276px]]<br>
 +
This is an extremely cheap, definitely of sweatshop quality, ATA-to-SATA adapter with an mSATA slot. It's possibly produced by Shenzhen Soarland and Hexin Technology Corporation Limited, or Sintech Electronic Corporation. The marking on the adapter reads 'HX-811 2.5-MSATA'.
  
=performance limited to UDMA 2=
+
Originally, this was purchased by a user desperately trying to replace his X40/X41/X41 Tablet's failing SFF-8111 1.8-inch based HDD. Unfortunately, due to the sweatshop-tier work performed by the intellectuals in said possible companies, there were two solder pads left unsoldered, rendering this adapter useless (although, it can be fixed by simply connecting them together via pouring solder over them as seen in the red circle of the image on the right side)
So you installed some sort of SATA disc in to your favourite laptop.  Next, you find its speed limited to Ultra DMA mode 2 33 Mo/s. [http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 What is wrong?]  ''That's not a bug, that's a feature!''
 
  
==over-validated input==
+
Be very careful of adapters such as this pictured here. Don't trust the 'QC PASSED' labels either, as the margins to produce this adapter was likely to be very low, causing the actual quality control to be non-existent.
IBM, or Phoenix or some group tried to be clever.  Too clever, ten years later reaping negative outcome and disappointment.  ATA Serial Transport is not ATA Parallel Transport; SATA devices (Attachments) not use PATA modes (neither DMA nor PIO signalling).  Firmware in some ThinkPad models checks whether drive reports 80-conductor cable, and configures ATA host controller likewise.  Sane operating systems limit bus to Ultra DMA 2, as a safety measure when using 40-conductor cable.  (Even at limited speed, you really should use UDMA-ready cable.)
 
  
[http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 known affected models] contain [[Intel 82801CAM|southbridge ICH3]]: {{A31p}} {{R32}} {{T23}} {{T30}} {{X22}} {{X23}} {{X24}}
+
===via removal of the Marvell 88SA8040 bridge controller===
 +
In models such as the {{T43}}, {{T43p}}, {{X41}}, {{X41 Tablet}}, and the {{R52}}, the 82801FBM I/O controller is capable of utilizing SATA. However, it was crippled in a way to use an ATA-to-SATA bridge controller for cost related reasons. This was due to the higher prices of SATA devices back in 2005, which were considerably more expensive than ATA devices. In that sense, it was a logical decision to use a bridge controller for that very reason.
  
==old firmware bug==
+
As time went on, ATA devices were slowly becoming more expensive as they dwindled in supply. SATA devices on the other hand, were turning much cheaper as they were basically oversaturating the market. This infuriated some users, as the bridge controller was holding back the potential of their models. As a result of that, some have managed to remove the bridge controller from the motherboard, and soldered a SATA connector to their models:
Tforty, Rfifty and family, contain [[Intel 82801DBM|southbridge ICH4]].  [http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 It would seem, old firmware versions exhibit behaviour just like in above ICH3 models.]
+
*[http://mikejmoffitt.com/articles/0000-t43sata.html '''ThinkPad T43: Intel 915GM SATA Hack''' T43p by Michael J Moffit http://mikejmoffitt.com/articles/0000-t43sata.html] <!-- http://mikejmoffitt.com/wp/?p=5 -->
 
 
affected models probably includes: {{T40}} {{T41}} {{T42}} {{R50}} {{R51}}
 
 
 
=modify Sonoma 2005 models=
 
Centrino Sonoma platform is comprised of Alviso chipset, which is first Intel chipset for mobile featuring SATA.  At power-on, ThinkPad firmware reports Error 2010 if device model is not in a whitelist; see [[Problem with non-ThinkPad hard disks]].  ThinkPads which house PATA drives, bridged to SATA host controller, include: {{R52}}, {{T43}}, {{T43p}}, {{X41}}, {{X41T}} (also {{X301}}, which is not Alviso generation).  Several people successfully modified their ThinkPads to use true SATA, eliminated signal conversion.
 
 
 
*[http://mikejmoffitt.com/wp/?p=5 '''Intel 915GM SATA Hack''' T43p by Michael J Moffit http://mikejmoffitt.com/wp/?p=5]
 
 
*[http://thinkpad-forum.de/threads/63974-gel%F6st-SATA-IDE-Bridge-umgehen-beim-X41 '''SATA --> IDE Bridge umgehen beim X41''' X41 by Luigi http://thinkpad-forum.de/threads/63974-gel%F6st-SATA-IDE-Bridge-umgehen-beim-X41]
 
*[http://thinkpad-forum.de/threads/63974-gel%F6st-SATA-IDE-Bridge-umgehen-beim-X41 '''SATA --> IDE Bridge umgehen beim X41''' X41 by Luigi http://thinkpad-forum.de/threads/63974-gel%F6st-SATA-IDE-Bridge-umgehen-beim-X41]
 
*[http://superyin.blogspot.com/2010/05/ibm-x41.html '''IBM X41 續命記''' by 阿倫 superyin http://superyin.blogspot.com/2010/05/ibm-x41.html]
 
*[http://superyin.blogspot.com/2010/05/ibm-x41.html '''IBM X41 續命記''' by 阿倫 superyin http://superyin.blogspot.com/2010/05/ibm-x41.html]
 
*[http://forum.thinkpads.com/viewtopic.php?f=3&t=90285 '''done: X41 with SATA SSD mod (pics)''' X41 by ZAGNUT http://forum.thinkpads.com/viewtopic.php?f=3&t=90285]
 
*[http://forum.thinkpads.com/viewtopic.php?f=3&t=90285 '''done: X41 with SATA SSD mod (pics)''' X41 by ZAGNUT http://forum.thinkpads.com/viewtopic.php?f=3&t=90285]
 +
 +
==Problems regarding SATA being used on ATA platforms==
 +
===Transfer rate limited to UDMA33 speeds===
 +
For some godforsaken reason, the BIOS in models with the 82801CAM controller (e.g. {{A31}}, {{A31p}}, {{T30}}, {{R32}}, {{X30}}) will force it to operate at UDMA33 speeds if it does not 'detect' an 80-pin cable. This clearly makes no sense in a notebook, as the amount of pins related to data transfer is only 44 (the other 4 pins are jumper pins). Unfortunately, this is considered a 'feature' as it actually does make sense in a desktop, but not really in a notebook. <br>[http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 Discussion related to this 'feature']
 +
<br>[http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 Discussion regarding models with the 82801CAM I/O controller]
 +
<br>[http://forum.thinkpads.com/viewtopic.php?f=31&t=115478&p=741729#p741729 Discussion regarding models with the 82801DBM I/O controller]
 +
 +
To bypass this issue, grounding pin 34 (PDIAG#, Pass Diagnostic) on the PATA connector is required. Unfortunately, most SATA adapters do not ground this pin, which results in the situation as described above.

Latest revision as of 18:46, 18 October 2020

This page gives an overview of several methods that can be used to provide an old model with IDE/PATA/ATA capabilities to use newer SATA/mSATA/M.2/U.2 devices, Please be aware that this page does not offer any information to extremely old models with ancient or obscure interfaces, such as SCSI.

Options

via ATA-ATA bridge controllers

These bridge controllers translate ATA-to-SATA or SATA-to-ATA commands, depending on their application. On other note, be wary of low quality controllers as they can potentially corrupt data transfers or fail prematurely. Some of them may also lack datasheets for some odd reason, and that may just be a sufficient red flag to steer clear of them.

For reference purposes, have a look at a list of ATA-ATA bridge controllers to judge what you need. Some may be faster or more stable than the other, but that does not necessarily mean that it is of higher quality, as these controllers could be built into very unreliable adapters (mostly due to poor PCB work).

Here is one shining example of said PCB work:
HX-811defect.jpeg HX-811correct.jpeg
This is an extremely cheap, definitely of sweatshop quality, ATA-to-SATA adapter with an mSATA slot. It's possibly produced by Shenzhen Soarland and Hexin Technology Corporation Limited, or Sintech Electronic Corporation. The marking on the adapter reads 'HX-811 2.5-MSATA'.

Originally, this was purchased by a user desperately trying to replace his X40/X41/X41 Tablet's failing SFF-8111 1.8-inch based HDD. Unfortunately, due to the sweatshop-tier work performed by the intellectuals in said possible companies, there were two solder pads left unsoldered, rendering this adapter useless (although, it can be fixed by simply connecting them together via pouring solder over them as seen in the red circle of the image on the right side)

Be very careful of adapters such as this pictured here. Don't trust the 'QC PASSED' labels either, as the margins to produce this adapter was likely to be very low, causing the actual quality control to be non-existent.

via removal of the Marvell 88SA8040 bridge controller

In models such as the T43, T43p, X41, X41 Tablet, and the R52, the 82801FBM I/O controller is capable of utilizing SATA. However, it was crippled in a way to use an ATA-to-SATA bridge controller for cost related reasons. This was due to the higher prices of SATA devices back in 2005, which were considerably more expensive than ATA devices. In that sense, it was a logical decision to use a bridge controller for that very reason.

As time went on, ATA devices were slowly becoming more expensive as they dwindled in supply. SATA devices on the other hand, were turning much cheaper as they were basically oversaturating the market. This infuriated some users, as the bridge controller was holding back the potential of their models. As a result of that, some have managed to remove the bridge controller from the motherboard, and soldered a SATA connector to their models:

Problems regarding SATA being used on ATA platforms

Transfer rate limited to UDMA33 speeds

For some godforsaken reason, the BIOS in models with the 82801CAM controller (e.g. A31, A31p, T30, R32, X30) will force it to operate at UDMA33 speeds if it does not 'detect' an 80-pin cable. This clearly makes no sense in a notebook, as the amount of pins related to data transfer is only 44 (the other 4 pins are jumper pins). Unfortunately, this is considered a 'feature' as it actually does make sense in a desktop, but not really in a notebook.
Discussion related to this 'feature'
Discussion regarding models with the 82801CAM I/O controller
Discussion regarding models with the 82801DBM I/O controller

To bypass this issue, grounding pin 34 (PDIAG#, Pass Diagnostic) on the PATA connector is required. Unfortunately, most SATA adapters do not ground this pin, which results in the situation as described above.