T601 Curious problems (slow speeds) in caddy drive after cloning.

[Bionicman]

I'm hoping someone with similar experience could explain this situation:

A T601 Frankenpad uses SATA HDD scorpio black 320GB in the ultrabay. (ultrabay slim serial ATA FRU/N 45J7902 P/N 45J901).
After cloning the HDD (using Macrium reflect) to a 512GB SSD, if I put this SSD in exaclty the same caddy, and test the speeds using CrystalDiskMarkInfo 7.0, all speeds are actually worse than on the HDD (I could show example pictures, for now lets say 0.015Gb/s for alltype of tests).
What's worse there are significant timeouts doing simple operations, which makes it unusable. The same SSD has been used in another laptop with an OS on it without any problems.
At first I wanted to ask here in the forum if there is someone, who has used an SSD in the ultrabay of a Frankenpad ever. But then I tested something different.
The first logical conclusion was that these laptops with PATA-to-SATA connection can't handle the SSD for some reason.
So I got a 7mm HDD Scorpio black 500GB and tested it in another laptop with normal SATA-to-SATA caddy. Write speeds all normal.
Next step, cloning the 320GB HDD to the 7mm 500GB HDD, Macrium reported write speeds about 195 Mb/s.
However, after cloning the HDD doesn't work as expected in the Frankenpad caddy as well! It's read/write speed dropped several times compared to the HDD from which it is cloned.
Which shows the problem may not be in the fact that it has a PATA-to-SATA interface.
Does anyone have any ideas ?

[theterminator93]

I've never tried using an SSD in the T61's ultrabay. But as you know the T61 uses a PATA-SATA bridge for the Ultrabay. SSDs don't play nicely with PATA interfaces. The SSD will technically work but a lot of the optimizations that make SSDs scream just don't translate across the bus.

HDDs and SSDs also physically organize their data differently - HDDs have sectors and SSDs have cells. If you clone an HDD to an SSD, if the software isn't "SSD aware" it can misalign the filesystem clusters across cell boundaries, forcing the SSD to work drastically harder to perform normally simple, speedy operations. A lot of that extra work would normally not be as huge an issue on a native SATA bus, but with a PATA bus, it's exacerbated.

Rule out the caddy's bus interface acting up by doing a R/W test on a SATA HDD. You could also try formatting the SSD in the main SATA drive bay with a Windows 7+ boot disk, then copying your files over manually.

[Bionicman]

I've never tried using an SSD in the T61's ultrabay. But as you know the T61 uses a PATA-SATA bridge for the Ultrabay. SSDs don't play nicely with PATA interfaces. The SSD will technically work but a lot of the optimizations that make SSDs scream just don't translate across the bus.

HDDs and SSDs also physically organize their data differently - HDDs have sectors and SSDs have cells. If you clone an HDD to an SSD, if the software isn't "SSD aware" it can misalign the filesystem clusters across cell boundaries, forcing the SSD to work drastically harder to perform normally simple, speedy operations. A lot of that extra work would normally not be as huge an issue on a native SATA bus, but with a PATA bus, it's exacerbated.

Rule out the caddy's bus interface acting up by doing a R/W test on a SATA HDD. You could also try formatting the SSD in the main SATA drive bay with a Windows 7+ boot disk, then copying your files over manually.

Ок, I have some updates.
First off, let me explain, the ssd or HDD works just normally in caddy drive through the PATA interface while it was empty, problems started after the cloning one disk to another. I need to clone the previous disk, because I have a 200GB+ of programs, settings and files -> too much manual work.

Second, the problem was repeated and timeouts occurred on a HDD in the same caddy - again after cloning. Meanwhile, the caddy works without problem on everyday tasks on every hdd, if it isn't a clone.

So I tried a different approach: move the source HDD disk outside the laptop (used the expressCard usb3 to shorten the cloning time by 2x). Put the destination (larger) SSD disk in the caddy. Used Macrium Reflect. Perform full clone. No problems with timeouts during working with files this time! However that lasted a day, until...
Since the destination is a larger disk, there was ~200GB unallocated space, which I added to the adjacent partition. Immediately after that timeout problems reappeared as well as slow read times.
Cloning is ruled out as the culprit at this point.

Given that I clone to a larger disk, I will always have unallocated space left and extending would always be necessary. I can't find explanation why extending a partition is the trigger for this peculiarity, as it is a standard procedure which has worked in much more complicated scenarios.

So next step: regardless of the disk being very slow and producing read timeouts, clone the entire SSD (this time as source disk outside the laptop through usb) => to destination HDD disk in the caddy of approximately the same size using AOMEI partition. Then reverse the source and destination => clone from the HDD to the SSD.
This time no extending is necessary, and the SSD works fine. Performance increase noted.

[RealBlackStuff]

You may have had the settings wrong during the cloning.
For Macrium you can check this: https://pureinfotech.com/clone-windows- ... m-reflect/
Scroll down to Expanding cloned drive.

This should also apply to W7.

PS: I have used Aomei, Macrium and Acronis, all without any problems such as you described.

[Bionicman]

I'm not sure Macrium has a setting to clone to a larger disk during the cloning process.
But I don't think the problem was with Macrium, everything works fine with it.
It has happened just exactly after Expanding cloned drive, the steps are the same as in the manual.
In this manual they just show how to expand it using the standard windows tools, it doesn't seem to be an option in Macrium.

[kfzhu1229]

At first I wanted to ask here in the forum if there is someone, who has used an SSD in the ultrabay of a Frankenpad ever. But then I tested something different.
The first logical conclusion was that these laptops with PATA-to-SATA connection can't handle the SSD for some reason.
So I got a 7mm HDD Scorpio black 500GB and tested it in another laptop with normal SATA-to-SATA caddy. Write speeds all normal.
Next step, cloning the 320GB HDD to the 7mm 500GB HDD, Macrium reported write speeds about 195 Mb/s.
However, after cloning the HDD doesn't work as expected in the Frankenpad caddy as well! It's read/write speed dropped several times compared to the HDD from which it is cloned.
Which shows the problem may not be in the fact that it has a PATA-to-SATA interface.
Does anyone have any ideas ?

I have not tried this on a T61/R61 just yet but I have tried this with the same specced same 965GM equipped Latitude D630/D830 and Compaq 6910p, all of them have this exact same problem.
Then curiously, their previous gen with these exact same caddies (D620 and Compaq nc6400 with 945GM) do not have this problem.
So I thought why? I think the answer actually have to do with somehow enabling SATA2 speeds on the primary SATA SSD makes the IDE optical drive interface to panic and freak out if that is running too fast as well.
I remember this was cited as a reason why manufacturers held off enabling SATA2 speeds on 965GM/PM/GL960 equipped laptops until later BIOS revisions.
So this might sound very weird to do, but is there a way to disable SATA2 speed patches of the middleton BIOS and see if the issue is solved?
Or maybe, try an older WD Black hard drive in the main HDD bay with a jumper on the HDD that sets it to SATA1.5G (SATA1) and see if that works.

[Bionicman]

I have not tried this on a T61/R61 just yet but I have tried this with the same specced same 965GM equipped Latitude D630/D830 and Compaq 6910p, all of them have this exact same problem.
Then curiously, their previous gen with these exact same caddies (D620 and Compaq nc6400 with 945GM) do not have this problem.
So I thought why? I think the answer actually have to do with somehow enabling SATA2 speeds on the primary SATA SSD makes the IDE optical drive interface to panic and freak out if that is running too fast as well.
I remember this was cited as a reason why manufacturers held off enabling SATA2 speeds on 965GM/PM/GL960 equipped laptops until later BIOS revisions.
So this might sound very weird to do, but is there a way to disable SATA2 speed patches of the middleton BIOS and see if the issue is solved?
Or maybe, try an older WD Black hard drive in the main HDD bay with a jumper on the HDD that sets it to SATA1.5G (SATA1) and see if that works.

Let's see if I got that right.
You suggest that the main drive bay sata2 enabled speeds are forcing some strange signal behaviour in the secondary bay ? Isn't the secondary bay in those motherboards SATA1 anyway ?
What is more puzzling is the fact that I encountered a problem with a cloned hdd as well. WIll have to test again with it.
Meanwhile - since the previous post the SSD in the caddy which worked fine for 3 days, began to face these timeouts again.
The performance gain in the main drive bay with SATA2 is of more benefit than what kind of disk is used in the ultrabay.
I'm not sure if I could or if I want to try to disable that
At this stage I hope the problem is only with ssd's.
But what I can test is using the same caddy in another machine which doesn't have middleton BIOS

[axur-delmeria]

The CD/DVD drive bay is PATA/IDE, not SATA. The HDD caddy for the T60 series has an IDE-to-SATA bridge chip. I have one (Fenvi brand) for my X61 ultrabase, and it runs frickin' hot. :O

[Bionicman]

The CD/DVD drive bay is PATA/IDE, not SATA. The HDD caddy for the T60 series has an IDE-to-SATA bridge chip. I have one (Fenvi brand) for my X61 ultrabase, and it runs frickin' hot. :O

Does that mean that such a caddy is different from that one : ultrabay slim serial ATA FRU/N 45J7902 P/N 45J901
could you tell the FRU number of your caddy ?

[kfzhu1229]

Let's see if I got that right.
You suggest that the main drive bay sata2 enabled speeds are forcing some strange signal behaviour in the secondary bay ? Isn't the secondary bay in those motherboards SATA1 anyway ?
What is more puzzling is the fact that I encountered a problem with a cloned hdd as well. WIll have to test again with it.
Meanwhile - since the previous post the SSD in the caddy which worked fine for 3 days, began to face these timeouts again.
The performance gain in the main drive bay with SATA2 is of more benefit than what kind of disk is used in the ultrabay.
I'm not sure if I could or if I want to try to disable that
At this stage I hope the problem is only with ssd's.
But what I can test is using the same caddy in another machine which doesn't have middleton BIOS

From my experience that's what I saw. Try to test it in a T61/R61 which for sure has no SATA2 speeds and see if the problem persists.
Either way though, I highly recommend you buy some tiny raspberry pi style aluminium heatsinks, open these Ultrabay adapters up (which in fact has the IDE to SATA converter and as axur-delmeria said it runs boiling hot), and slap those heatsinks on with some thermal paste/epoxy. I already have one of these things die on me and they are quite expensive nowadays due to the rarity (even for generic optical drive IDE to SATA) and the extra costs due to the converter. And I have since done this to the rest of my converters.
Or if you are fine with a slower hard drive (for data storage), just get the IDE version of this adapter which has no converters and it is always 100% reliable.

Does that mean that such a caddy is different from that one : ultrabay slim serial ATA FRU/N 45J7902 P/N 45J901
could you tell the FRU number of your caddy ?

All T6x/R6x (as well as T4x/R5x) use IDE for the ultrabay connector (and is unfortunately propietrary from the standard IDE connector found on G4x series). My observation would be invalid towards T4x0/T5x0 series which all use SATA and has no such conversion chip inside the caddies. Those can have 3 dollar ebay optical drive to HDD adapters, swap the mounting hardware and faceplate of your optical drive over, and would work rock reliable.