Hack@Home: 16GB mSATA SSD to Amiga IDE - for about £10

[nogginthenog]

Originally Posted by Spacemonkey

The Adapter is a KA60 (see below) and the CF reader (set as Salve) is plugged into the 40 Pin IDE port with a IDE single cable. The m-Sata was plugged in to the 44 Pin port which should be master.

That's pretty nice. I guess that's not your picture as it's an SD card.

I would still be tempted to try putting the master at the end of the cable but you can't do that with your adapter.

[Spacemonkey]

no mine is under the Amiga 600 rack Case post

[Shadowfire]

Just a few notes from back when I was doing testing of IDE drives.
1. The cheapo IDE -> SATA adapter cards use microcontrollers, not state machines, to perform bus translation. There is always a speed loss when going through one. The degree of speed loss can range from around 50% ("probably won't notice this on a slowass amiga") to 95%+ ("why did they even bother").
2. The Transcend PATA SSD drives (16GB etc) seem to be based on ancient USB memory stick technology. Writes are abominably slow, I mean near-FLOPPY drive slow. Reads aren't that much better than a 5400rpm hard disk from the Amiga era. STAY AWAY.
3. With a mSata SSD, you at least have a chance to avoid the old SSD controller abominations (such as in the previously mentioned Transcend drives). You still, unfortunately, have to deal with the SATA -> IDE translators.
4. There is a stickied thread in EAB about drives compatible with the A600. A lot of CF-IDE drives take "liberties" with the IDE specification (such as responding to both master and slave addresses).

[patrik]

Originally Posted by Shadowfire

3. With a mSata SSD, you at least have a chance to avoid the old SSD controller abominations (such as in the previously mentioned Transcend drives). You still, unfortunately, have to deal with the SATA -> IDE translators.

Have bought several of the cheapest adapters, both SATA->IDE40 and mSATA->IDE44. All of them were like Zetros based on the JM20330 chip (can be seen in his images).

I have used them for connecting mSATA SSDs to acard SCSI->IDE bridges. If connected through a such bridge (ultra wide) on a CSPPC, it posts the following numbers for the 256kB test in DiskSpeed on a PFS volume:

Testing with a 262144 byte, MEMF_FAST, LONG-aligned buffer.
Create file:     19267584 bytes/sec
Write to file:   31948800 bytes/sec
Read from file:  28573696 bytes/sec

Don't know what kind of type this controller would be categorized as, but it seems fast enough for use on the Amiga.

Also did a quick test by connecting this setup to a linux PC with a ultra wide scsi card and doing some sequential read/write tests with dd on the raw device. With 4kB blocksize it managed 90MB/sec read and write so my guess is it "links" at ATA100 with the acard bridge.

With 512B it managed 75MB/sec read and 9MB/sec write. I suppose the low write speed at 512B must be and issue with dd, linux itself or the scsi card drivers as the PFS volume on the amiga also uses 512B blocks and manages >30MB/sec writes.

[Predseda]

05 October 2016, 10:07 Predseda #85

Is there a solution for the LED problem?

Originally Posted by Zetr0

Yes, I will do a "simple guide" when I get mine back from Havie =)

Eeeerm, bump?

[Zetr0]

@Predseda

LOL, sadly my friend I need to get another one to re-do this as mine has been lost unto the ether of the postal service =(

[jagLally]

hi,

I'm new to the scene and I saw this thread to try to use an SSD instead of a CF card. I followed this thread and the one on using large disk > 4Gb. http://eab.abime.net/showthread.php?t=61666 and http://eab.abime.net/showthread.php?t=32256

I installed ClassicWB FuLL 3.1 using FS-UAE. Patched the SCSI.device to 43.45 and managed to add the PFS3-aio to RDB block. Everything went well and I can boot it in FS-UAE.

However, when I plug it into real A1200, it can't see the partitions under Boot Options. I get the line:
CC0: ..... Enabled CC0 priority 3 ADOS carddisk-0

But no partitions are visible on the SSD

My A1200 has ACA 1221EC Accelerator. kickstart 3.0

sudo rdbtool /dev/rdisk2 info
Password:
PhysicalDisk:               0  1465788    46905248   22Gi  heads=1 sectors=32
LogicalDisk:                5  1465788    46905088   22Gi  rdb_blks=[0:159,#160] used=[hi=145,#146] cyl_blks=32
Partition: #0 'RDH1'        5    70000     2239872  1.1Gi    4.78%  PFS3/0x50465303 max_transfer=0xffffff mask=0x7ffffffe num_buffer=30 bootable pri=0
Partition: #1 'RDH2'    70001   140000     2240000  1.1Gi    4.78%  PFS3/0x50465303 max_transfer=0xffffff mask=0x7ffffffe num_buffer=30
FileSystem #0 PFS3/0x50465303 version=18.0 size=69388 seg_list_blk=0x4 global_vec=0xffffffff

It's taken a whole day to get to this point and when it worked in FS-UAE, I thought I was home and dry. I'm stumped as what to do next. My current partitions are all below 4GB!

Please help

[daxb]

Can't help but saw that buffer values are to low.

[Predseda]

I encountered a problem. Bought everything from those exact ebay dealer and links that Zetr0 posted and it doesnt work. Finally discovered the problem is with the mSATA adaptor, that works under WinUAE, but not in a real Amiga. Strange. I need to order another one and hope it is working this time :-/

[FOL]

Originally Posted by Predseda

I encountered a problem. Bought everything from those exact ebay dealer and links that Zetr0 posted and it doesnt work. Finally discovered the problem is with the mSATA adaptor, that works under WinUAE, but not in a real Amiga. Strange. I need to order another one and hope it is working this time :-/

Ive tried 3 so far, none of which work.

[Predseda]

We tried one, from other ebay seller. Looked the same, but there were more prints on PCB, and it worked.

[indigolemon]

Originally Posted by Predseda

We tried one, from other ebay seller. Looked the same, but there were more prints on PCB, and it worked.

Do you have a link?

[Predseda]

We have photos, comparison of a working and non-working adaptor. Problem is all the eBay sellers have stock photos.

[jagLally]

To try to isolate the issue I went back to basics and created a disk image and installed a standard Workbench 3.1 from Amiga Forever disks using FS-UAE.

I then copied the same image to an SD card and the SSD disk.

The SD card booted up immediately and the SSD refused.

The other observation is that when I choose 'Boot without executing Startup-Sequence' from the Boot menu because frankly I couldn't think of anything else; occasionally it boots into DOS window and if I run the 'info' command it shows the SSD partitions correctly. There is not pattern and it only happens randomly. Most of the time it's just black screen.

I'm using the components linked in Post 1 except the SDD is of higher 22G capacity (same make).

So I agree it's most likely the SSD adaptor compatibility. The fact Zetro has it working in A600 makes me think it's the adaptor + A1200 combination perhaps.

Commands used to prepare the disk image before going to FS-UAE.

rdbtool  ~/FS-UAE/Hard\ Drives/Workbench31.hdf  create chs=16383,16,63 
rdbtool  ~/FS-UAE/Hard\ Drives/Workbench31.hdf init rdb_cyls=2 + info
rdbtool  ~/FS-UAE/Hard\ Drives/Workbench31.hdf add size=50Mib name=DH0 dostype=ffs + info
rdbtool  ~/FS-UAE/Hard\ Drives/Workbench31.hdf change DH0 bootable pri=0 + info
xdftool  ~/FS-UAE/Hard\ Drives/Workbench31.hdf open part=DH0 + format disk1 ffs + list
# use FS-UAE to install
sudo dd if= ~/FS-UAE/Hard\ Drives/Workbench31.hdf of=/dev/disk2 bs=10m

Has anyone managed to make it work in A1200?

[Predseda]

Here is the picture. That one which has no printing doesnt work, the other works

https://photos.google.com/share/AF1Q...JoYm9UMU9ZWjhB

Beside those texts they look identical.

[jagLally]

Originally Posted by Predseda

Here is the picture. That one which has no printing doesnt work, the other works

https://photos.google.com/share/AF1Q...JoYm9UMU9ZWjhB

Beside those texts they look identical.

That’s interesting. Which SSD worked with the adaptor? Mine is exactly the same as the printed one bought Dec 2017 and I’m using Sandisk 22gb SSD. Can see the drive but not partitions in Boot menu. If I boot from floppy, HDTools can’t detect the drive.

[Predseda]

Originally Posted by jagLally

That’s interesting. Which SSD worked with the adaptor? Mine is exactly the same as the printed one bought Dec 2017 and I’m using Sandisk 22gb SSD. Can see the drive but not partitions in Boot menu. If I boot from floppy, HDTools can’t detect the drive.

This one. It is also from Zetr0's link:

https://photos.google.com/share/AF1Q...JoYm9UMU9ZWjhB

[DaMi]

Hi, I own a working mSATA IDE adaptor, it's the one Predseda mentioned. On its main chip there is a label JM20330 1630 TGAZ0 C0 B761F0141 and on the smaller one can be found AMS 1084CD 3.3 1529.
I use this adaptor with 32 GB mSATA SDD SanDisk memory, it works flawlessly on both WinUAE and real Amiga 1200. Actually, according CF or SD adaptors, this one is much warmer.
Predseda's not working adaptor is visually almost the same, but it has different labels. The label on the main chip is JM20330 1618 TGAZ0 C0 B75Z20341 and on the smaller one AMS 1117CD 3.3 1548.
We have suspicion that the reason why Predseda's adaptor does not work is that his adaptor contains visually the same components but technically they are probably different (because of different labels).
Can someone compare your labels?
Detail photo of my working adaptor is here. And here is Predseda's not working.

[jagLally]

Mine has the labels as in the working pictures and visually looks identical BUT does not work. The main label was under the QC sticker and is very faint; best I can read is JM20330 1708 T6AZ0 C0 B766Z0141. The smaller chip is clear AMS 1084CD 3.3 HS1714.

Is there more than one set of numbers that work and I suppose the vendor could be asked before purchase.

[AwoLStill]

Originally Posted by jagLally

I suppose the vendor could be asked before purchase.

Spot the differences:

CAUTION:
The following URLs are direct to .pdf files that appear ok, but the links on the homepage of that domain appear to redirect to an unrelated site, suggesting either that the site has been hacked, or my laptop has malware on it. I suspect the former. I will check this later and confirm, in the interim here are some details:

This appears to be the datasheet for AMS 1084.
http://www.advanced-monolithic.com/pdf/ds1084t1.pdf

"GENERAL DESCRIPTION
The AMS1084 series of adjustable and fixed voltage regulators are designed to provide 5A output current and to operate down
to 1V input-to-output differential. The dropout voltage of the device is guaranteed maximum 1.5V at maximum output current,
decreasing at lower load currents.
On-chip trimming adjusts the reference voltage to 1%. Current limit is also trimmed, minimizing the stress under overload
conditions on both the regulator and power source circuitry.
The AMS1084 devices are pin compatible with older three-terminal regulators and are offered in 3 lead TO-220 package, 3 and
2 lead TO-263 (Plastic DD) and TO-252 (D PAK) package."

This appears to be the datasheet for AMS 1117.
http://www.advanced-monolithic.com/pdf/ds1117.pdf

"The AMS1117 series of adjustable and fixed voltage regulators are designed to provide up to1A output current and to operate
down to 1V input-to-output differential. The dropout voltage of the device is guaranteed maximum 1.3V, decreasing at lower
load currents.
On-chip trimming adjusts the reference voltage to 1.5%. Current limit is set to minimize the stress under overload conditions
on both the regulator and power source circuitry.
The AMS1117 devices are pin compatible with other three-terminal SCSI regulators and are offered in the low profile surface
mount SOT-223 package, in the 8L SOIC package and in the TO-252 (DPAK) plastic package."

I have emailed them to report their webpage and asked if they have a U.K. based dealer, or a quote for some AMS 1084CD 3.3 1529 shipped direct.

Website was last updated in 2007.
I have no idea if they are still in business.
LinkedIn and Bloomberg seem to think so.