[SOLVED] SSD replacement for 1.8" PATA Toshiba MK8007GAH (unusual connector)

Feb 14, 2020
10
1
15
I need to find a SSD substitute for a 1.8" 3.3V PATA SSD, Toshiba part number MK8007GAH. This was used in Fujitsu P1610 subnotebook and at least one similar-era Sony PC.

The problem is that the connector is not a "ZIF" (zero insertion force) type; rather it is a female 2-row receptacle (probably 50 pins total), and the P1610 cannot support a ZIF type PATA SSD. The Super Talent DuraDrive ZT4 was the direction I was going until it arrived and I discovered it was ZIF type.

Are there any part/type numbers I should look for? Even if I'm constrained to the used/recycled marketplace, that would be OK. I am happy to go with any available size above 8GB, but something in the 64-256GB range would be ideal. An alternate would be (probably even better) an adapter housing (same size/form factor as the MK8007GAH) to accommodate a MSATA module, but I've not seen that either with this type of connector. I'd like to avoid CF adapter (same problem with connector, but further complicated by the ability to serve as OS/ boot device.)

URL for user guide for the Toshiba MK8007GAH that shows an end sketch of the drive is at:

Toshiba MK8007GAH PDF

Thanks for any suggestions.
 
Solution
I would measure the voltage at the laptop's power pins, just to be sure that the docs are not in error. You can easily verify the ground pins by testing for continuity with a known ground point on the chassis or motherboard.

One other possible problem could be the CF-TF adapter. CF cards can come preconfigured as either ATA devices (HDDs, SSDs, etc) or ATAPI devices. It may be that your laptop may not be able to boot and run from the latter.
Thanks, that provides the needed clue. The Kingspec devices are not available (at least on my initial search) - if anyone reading this knows where I can find one of those drives, please message! If all else fails I can go the CF adapter route as 50 pin IDE to CF adapters exist. Now I know what a 50 pin IDE connector looks like!
 
OK - thanks for the replies, this is very helpful.

I'm going down this path - 1.8" IDE (same connector) to CF adapter, followed by CF to TF adapter (type 1 module that accepts one or two TF cards.) The CF to TF adapter claims UDMA support, and supports 3.3V operation (power from the laptop to the Toshiba drive.) Pin 1 is marked in all of the involved interfaces. While the IDE/CF adapter is an exposed PC board, it can be protected from inadvertent contact inside the laptop. I used a similar arrangement years ago with a couple of earlier-generation Lifebook machines (IDE-CF) with good results, however, both of those used 2.5" IDE form factor drives. In any event, this isn't a costly experiment - if it doesn't work, proves mechanically incompatible, or results in some damage, I'll scrap the project. In any event, I'll report back results.

Thanks again for the very helpful replies, suggestions and comments!
 
I would measure the voltage at the laptop's power pins, just to be sure that the docs are not in error. You can easily verify the ground pins by testing for continuity with a known ground point on the chassis or motherboard.

One other possible problem could be the CF-TF adapter. CF cards can come preconfigured as either ATA devices (HDDs, SSDs, etc) or ATAPI devices. It may be that your laptop may not be able to boot and run from the latter.
 
Last edited:
Solution
Thanks much!

I'll check voltage on applicable pin on the IDE/CF adapter before attaching the CF/TF adapter.

As for behavior of the CF/TF adapter, that is what I ran into in the earlier era (at the time I required a specific CF device that would appear bootable to the OS.) In this case, there is no documentation to that level for the CF/TF adapter, so if it is not bootable, then my next choice is to give up or seek a native CF device which can be bootable (they exist, but may be more expensive.)

Thank you again for the helpful advice!

Dave
 
Some success! The 1.8" IDE to CF adapter arrived yesterday (the CF to TF adapter module is still in transit.) I installed the IDE/CF adapter with a Transcend TS4GCF200I (4GB, Ultra DMA Mode 0 to 5 (UDMA4 as default)). On first try, the locator pins at the edges of the P1610's connector interfered with small guide extensions at the edges of the adapter connector preventing full mating, but these were trimmed from the adapter connector and it then fit fine. Pin 1 is not labeled on the P1610 IDE cable but this labeling was easily added. Installed XP onto the 4GB CF card and it runs much faster in the P1610 than the HDD ever did.

Next update once the CF/TF adapter arrives.
 
Some more success - the CF/TF card adapter arrived (it takes two TF cards and combines the capacity and splits the traffic, so in theory it's faster then a single TF card of the same type).

The P1610 recognized the setup as FC-1307 SD CF adapter (with about 30 GB storage). The FC-1307 turns out to be a chip specifically for this sort of purpose (bridge from SD/TF to CF/IDE to support obsolete computing devices), which was released from a Korean outfit in late 2012 (datasheet overview is available online.) The P1610 BIOS shows support for UDMA 5 (the chip supports UDMA 6).

After a failed (and extremely slow) attempt installing Windows 10 (32 bit) - timeouts in the initial setup dialog sequence - I replaced the two 16 GB TF cards with a pair of 32 GB Samsung EVO cards (intended for my wife's bicycle camera); these run much better (faster, particularly for writing), however, Windows 10 still cannot get through it's initial configuration sequence (a series of dialog screens which run too slowly and appear to time out.) There's a procedure on www.configserverfirewall.com (search for OOBEREGION to find the relevant post) to help get past the dialog screens, but as yet Windows 10 is not running. I'll post final outcome as an update.

In a nutshell, the combination of 1.8" IDE to CF then CF to TF adapter based on the FC-1307 (you won't find that part number, just look for UDMA on the CF/TF adapter labeling) solves the conversion of the P1610 to SSD without relying on antique SSD or costly native CF devices. Whether I can get Windows 10 installed remains to be seen (fallback will be Windows 7, and failing that, XP), but the original problem I posed here - the need for 1.8" IDE SSD - has had a successful outcome.

Thanks again to all that replied to my original post!

Dave
 
OK - windows 10 ver 1909 is running on SSD on Fujitsu P1610 (minus touch panel and bezel drivers, which were unique to Fujitsu P1610 on XP).

This confirms the hardware solution, and the balance of the issues were/are related to the machine being under-spec for Windows 10 (1G RAM, 1.2 GHz single-core CPU).

Taming windows 10 for this environment Igetting as many unneeded processes and functions disabled as possible) is ongoing (so far so good), and 2G RAM module is on order.

Thanks again for all of the help and pointers! Much appreciated.

Dave
 
EPILOGUE (2/28/2020)

The P1610 is now running Windows 10 (32-bit) in 2 GB RAM (upgrade modules available on ebay $50), using a pair of Sandisk Extreme Plus 32GB microSD cards (type A1) for mass storage.

Learned that the SD Card association has created two standards (called A1, succeeded by A2), which make SD (and microSD) cards suitable for hosting software applications (with numerous random read/write cycles):

https://www.sdcard.org/developers/overview/application/index.html

I went with the A1 option as those were available locally the day I worked on the project. The difference in behavior is significant (gets rid of the pronounced lags experienced when the OS did random writes to the conventional SD cards, including Sandisk Pro types I tried.) CPU speed/throughput on the P1610 is now the limiting factor (not snappy, but fully adequate for the need.) The A2 type cards are readily available for about same cost (but not the day I was working on the project), but I doubt the P1610 would take advantage of the (4x higher) random write IOPS specification of the A2 type cards.

I can't verify how the FC1307 SD/TF adapter juggles the two micro SD cards (does it indeed split the load among the two cards, presumably improving apparent throughput?) Enough time spent!

So, lessons learned -

  1. 1.8" PATA to SSD isn't a viable option these days due to lack of devices, but 1.8" PATA to CF, and CF to TF, works as a good solution.
  2. Type A1 (or A2) micro SD cards are mandatory to get adequate responsiveness. I'd go with A2 in general, but A1 was adequate for the P1610 due to it's fairly meager CPU resource.
  3. I turned off a number of Windows 10 features/functions (background tasks, cortana search, etc.) using registry changes - this helped. I also turned off windows defender - a procedure below that requires temporary boot to Linux live CD - replacing it with an AV application that was more flexibly configured (Windows Defender runs in the background at maximum priority at boot time, leaving this machine pretty unresponsive, so another AV solution is very important in this respect.)

https://www.techpowerup.com/forums/threads/how-to-disable-windows-defender-in-windows-10-1909.262699/post-4183761

https://endurtech.com/how-to-turn-off-and-disable-microsofts-cortana/

I hope this information helps someone interested in preserving the Fujitsu P1610 going forward - I'm astonished that Windows 10 runs acceptably (to me) on a nearly 15 year old machine, giving up only the touchscreen capability (getting the XP driver to run on Windows 10 has so far proved fruitless.)

Thanks again to those on tomshardware forum that got me going in the right direction!

Dave