ATTO UL3D / UL4D SCSI Pains

[MDBenson]

Hi,

I have a number if New World G3/4 PowerMacs and I am trying to get hold of a SCSI card to use some late 80-pin U320 drives with tgat works in OS 9 and OS X.

I have had success with OS X 10.4 and 10.5 on these drives with a ATTO UL4D so I am pretty sure my cabling, termination and sacraficial rites are all correct. The drives are lashed up using 80-pun to 68-pin+Molex adapters. Again, they worked flawlessly in a G4 PowerMac QS booting and using Mac OS X.

I have a UL3D card but I can't get a lot of sense out of it. I initially tried it in a Rev. 2 Blue & White G3 with a (yes, I know) 1GHz G4 upgrade in it but couldn't get a lick of sense out of the drives. No boot and seemed to sit scanning busses for boot drives for a very long time with no success. The UL4D didn't seem to behave a lot better so I dismissed it as a test candidate.

I put the same 2 drives and the UL3D it in a G4 Gigabit Ethernet model and it was still unhappy, flashing "where am I supposed to boot from Folder/? and it seems to take an age to decide there's no drive. I eventually booted the system from the MacOS9Lives 9.2.2 Install image on a FireWire SSD and the card and drives were not recognised. As the UL3D card was from an unknown source, I decided to flash the firmware with the OS 9 flash utility. Although it already had 1.66 firmware I forced an overwrite usin the 1.66U updater.

This produced some level of success in that 9.2.2 now tried to mount the drives. Only one of themmounts, but the other drive with 10.5 Leopard on won't, but I suspect it has no OS 9 drivers because Leopard doesn't care about that nasty old Classic OS.

The drive with 10.4.11 and 9.2.2 on is recognised but still won't boot.

As I can boot from MacOS9Lives via FW, I decided to attach a different drive (same model, different cspacity) and start fresh. A restore of MacOS9Lives 9.2.2 went just dandy but that isn't bootable either. More concerning is the FW 9.2.2 boot freezes and the new drive just headbangs constantly. Could be a bad drive I guess but I ran out of energy.

I guess the takeaway TLDR; question is are UL3Ds usually this painful to work with? Of so what alternative do I have to work with OS X and OS 9 at the same time?

I also found a stray post somewhere tgat said Version 1.01 of the UL4D firmware worked with OS 9 (or both?) but I've never been able to find it, does anyone have it?

 

[SnakeCoils]

Hello,
I have an archive with the original CD content of UL3S/3D and UL4S/4D drivers and firmware updaters including the one you are looking for but I doubt it could be useful because at the time I made the mistake to simply copy the content without compress it with some OS9 archiver so the fork/data information could be missing and the updaters not executable anymore... :-( If you know how to make them executable under OS9 it could work again, there is also an OSX section with various "flashbundle" so I hope they will be somewhat useful to you.
I attach the archive here, good luck!

 

[MDBenson]

Thanks, I really appreciate that. It's a shame it's not directly usable, but it does provide a very good starting point.

 

[ojfd]

1. How were the drives formatted? On what OS? Which utility?
2. Take a trip to web.archive.org and check what ATTO said back then about OS/SCSI card/driver compatibility.

 

[MDBenson]

1. How were the drives formatted? On what OS? Which utility?
2. Take a trip to web.archive.org and check what ATTO said back then about OS/SCSI card/driver compatibility.

The original pair of drives were formatted in their respective OS X installers (10.4 and 10.5) while they were on a UL4D.

The third drive I tried was formatted using Drive Setup (OS 9.2.2) and I also tried using ATTO ProTools but neither seemed like it worked. Notably no matter what I used ProTools wouldn’t write the ATTO driver to the disk, it whined about not having any free drive space.

I have already dredged archive.org and found the same information, and no copies of the UL4D firmware, sadly. It’s one if only a few files missing from the archives of ATTO’s FTP server.

 

[ojfd]

It seems that UL4D is the wrong card to be used with OS9.

Spoiler

 

[MDBenson]

As I’ve stated, there is a Mac OS 9 firmware for the UL4D. The chart is apparently a lie. 😂

The UL4D was really designed for G5s but I suspect they backtracked on the OS X only features after launch because of people still wanting to dual boot G4s in which the UL4D works just fine.

 

[ojfd]

So what? That's just a firmware updater for OS9, but, have you ever seen any statement or official document by ATTO that confirms that UL4D is OS9 bootable?
Attached are two ATTO documents containing "lies". Dated 2004 and 2005.

I doubt UL4D were designed with G5 in mind. Rather something Windows or so. Check the reports here:

Apple PowerMac G5 tech docs, G5 software updates

Powermac G5 tech docs, G5 software updates and install guides

www.xlr8yourmac.com

P.S. I've uploaded a whole bunch of my SCSI stuff that I had - apps, firmwares etc. from NuBus times up to PCI G4 DA and everything inbetween, to the preterhuman Hotline server some time ago. You might try to look there using KDX or Nostalgia. If I remember correctly, there was some ATTO stuff too.
Also, as far as I can remember, ATTO ProTools was somewhat finicky. I used it for benchmarking and to mount hotswap SCSI drives. Sometimes one version didn't work as expected and I had to switch back to some older one. (Let me check, I might have some of them on one of my SL machines).
For formatting SCSI drives I only used FWB Hard Disk Toolkit. 4.5.2 was the last version, btw. Garden has it, give it a try.

 

[MDBenson]

Nitpicking aside, thanks for you info, and for the tip on preterhuman’s archive - I’ll be sure to check that out. I also didn’t try FWB HD Tools so I’ll give that a whirl.

 

[MDBenson]

Myself and another community member have positively established that we need an ISO or Toast image of ATTO ExpressPCI Tools CD-ROM from somewhere around 2000-2003.

If anyone has that disc or an image of I’d be incredibly grateful if you could provide a copy or better yet upload it to Archive.org

 

[Ivan Shpak]

So what? That's just a firmware updater for OS9, but, have you ever seen any statement or official document by ATTO that confirms that UL4D is OS9 bootable?
Attached are two ATTO documents containing "lies". Dated 2004 and 2005.

I doubt UL4D were designed with G5 in mind. Rather something Windows or so. Check the reports here:

Apple PowerMac G5 tech docs, G5 software updates

Powermac G5 tech docs, G5 software updates and install guides

www.xlr8yourmac.com

P.S. I've uploaded a whole bunch of my SCSI stuff that I had - apps, firmwares etc. from NuBus times up to PCI G4 DA and everything inbetween, to the preterhuman Hotline server some time ago. You might try to look there using KDX or Nostalgia. If I remember correctly, there was some ATTO stuff too.
Also, as far as I can remember, ATTO ProTools was somewhat finicky. I used it for benchmarking and to mount hotswap SCSI drives. Sometimes one version didn't work as expected and I had to switch back to some older one. (Let me check, I might have some of them on one of my SL machines).
For formatting SCSI drives I only used FWB Hard Disk Toolkit. 4.5.2 was the last version, btw. Garden has it, give it a try.

Hey! Please tell me in which directory and folder you uploaded your staff in preterhuman, I just have been sitting for 3 hours and I can’t find what you are talking about

 

[SnakeCoils]

Hello

Myself and another community member have positively established that we need an ISO or Toast image of ATTO ExpressPCI Tools CD-ROM from somewhere around 2000-2003.

If anyone has that disc or an image of I’d be incredibly grateful if you could provide a copy or better yet upload it to Archive.org

Hello,
I don't know if in the meantime you have found what you were looking for but recently I was lucky enough to find at a very good price an ATTO UL4D along with an original ExpressRAID CD-ROM (version 2.8.1) and this time I did not made the same previous mistake in archiving the disc so I have compressed it on my MDD FW400 booted in OS9.
Inside there are various firmware updaters both for OS9 and OSX and of course the almost impossible to find ExpressRAID software.
The previous owner gave me also the activation code and license key but he doesn't allowed me to share it, only the program that otherwise is limited to a trial od 30 days with full features, sorry...
About the UL4D card the OS9 updater is not the latest 1.01B you are looking for (it is in fact the 1.0) but maybe it could be useful anyway

Below you will find the Stuffit Archive with the entire content of the ATTO CD-ROM, it is re-compressed in Zip otherways I could not be able to upload it. Enjoy!

 

[MDBenson]

Ahhh man, you're a Legend, and I'm really thankful for you posting this.

Sadly trying to roll back to the 1.00f Firmware bricked the UL4D I tried it on un-recoverably (I can't even get the OS X utility to flash 1.50 back on it now) :\

All my cards have 1.52 firmware on and apparently aren't compatible with the older firmware version. So dang close :\

Meanwhile I'm still going through a few things with the UL3D I have. It works okay in 3 of my systems (B&W, G4 AGP 500 and a G4 DA) but it won't boot Mac OS 9 and hangs the boot-select up on all three (I used

multi-boot

on the G3).


I'm also looking at Adaptec cards, 29160 and 39160s are easy to find and Mac flash utility and ROM files are still avalable from MICROCHIP, the now-owners of Adaptec's IP. The problem is support for the cards ceased at OS X 10.3 officially and I've not had a lot of luck getting a driver KEXT to work on OS X 10.4 thus far with a flashed 39160 (that works flawlessly in OS 9 and even boots). Someone I have talked with said their **29160** worked in Sorbet Leopard (10.5 with patches and fixes) uysing the 10.2 driver so I have that to try. Even if it means busting my New-in-Box 29160 Mac card out of the cellophane!

The most reliable and widely usable card I've used this far has been the good old AHA-2906 cards that Apple fitted as OEM BTO options. They work in anything PPC flavoured I've tried thus-far because Apple kept supporting them. Someone also suggested maybe a 2940 might work well across versions too for the same reason. If I can find a ROM and flash utility for a 2940 I think I have one in my PC spares box.

 

[SnakeCoils]

Really sorry about your UL4D, I also have just tried the OS9 updater on mine and yes, now it is bricked too... 😭 both my MDDs (FW800 and FW400) refuses to boot if the card is installed, I don't know if this matters but when I launched the firmware updater the system clock was not set correctly (exhausted lithium battery...) and strange things happens with a wrong date/time setup, for example if you format a disk drive you will not be able to write on it until a more recent date and time are set, also the drive information are not reported and size or capacity are blanked.
About the other SCSI cards I fully agree about the Adaptec 2906, it was one of the most compatible with my scanners and also removable drives. If I remember well there was some KEXTs to remove for working at its best: the only caveat was that I had to leave a SCSI terminator plugged on the rear 25 pin port, otherwise the Mac refuses to boot (this happened only after the KEXT removal). The UL3D is a strange beast: on some cards the OS9 boot is possible only on external ports, on other cards (same 1.66f firmware) also on internal ports, never understood why.
I had many SCSI cards on my Macs and the best for ratio speed/compatibility I would reccomend if both OS9 and OSX support is needed is the ACARD AEC-67162M but nowadays they are rare as unicorns, I am really disappointed to have sold mine long time ago... In the meantime have a spare UL3D I can use until I will find a way (if exist) for resurrecting the UL4D, maybe on a G5 or a PC, perhaps the firmware code can be extracted from the flasher using ResEdit, I will try in the next weeks.

 

[Tratkazir_the_1st]

I thought about purchasing UL4D from local second-hand market. But after such troubles - no, thanks . At least I have several adaptecs (2930CU, 2940U2W, 29160). I suppose - even non-flashed, they will work in OSX? (My Canister, G4 MDD, still refuses to install masos 9 ).

 

[DearthnVader]

I will find a way (if exist) for resurrecting the UL4D, maybe on a G5 or a PC

If it uses a standard EEPROM the normal way is to lift the ground pin and make a switch circuit. Assuming the issue is an FCODE ROM on the chip that makes your Mac unbootable or the card no longer shows up on the PCI bus.

If you have access to a PC and the Utilities to flash the card, that can be an easier route. Tho sometimes PCI ROM's will not work in a PC or a Mac and you are stuck with lifting the ground pin anyway.

If you lift the ground pin and boot the Mac/PC the EEPROM can not be read, so the card will show up as whatever the PCI Confg Registers are set too. This can be an issue depending on how they wrote the utility to look for the card.


However, if everything goes as planned after boot you just make the switch sending the EEPROM to ground and the flashing utility will see the card and be able to read and write the EEPROM.

If that fails there are things you can do with Open Firmware to present the PCI card as the flashing utility expects so that it can R/W the EEPROM. We don't have this ability with BIOS/PC's.

So try and identify the EEPROM on the card and find the ground pin>create the switch circuit and see if the system will boot with the switch off.

If all that works, make the switch to on, and launch the flashing utility and see if you can recover the card by flashing a known good ROM to it.

If that fails at any point let us know and we'll dig deeper into the issue....

 

[ojfd]

Now, how do you safely lift a pin on PLCC-32 chip?

 

[SnakeCoils]

If it uses a standard EEPROM the normal way is to lift the ground pin and make a switch circuit. Assuming the issue is an FCODE ROM on the chip that makes your Mac unbootable or the card no longer shows up on the PCI bus.

If you have access to a PC and the Utilities to flash the card, that can be an easier route. Tho sometimes PCI ROM's will not work in a PC or a Mac and you are stuck with lifting the ground pin anyway.

If you lift the ground pin and boot the Mac/PC the EEPROM can not be read, so the card will show up as whatever the PCI Confg Registers are set too. This can be an issue depending on how they wrote the utility to look for the card.


However, if everything goes as planned after boot you just make the switch sending the EEPROM to ground and the flashing utility will see the card and be able to read and write the EEPROM.

If that fails there are things you can do with Open Firmware to present the PCI card as the flashing utility expects so that it can R/W the EEPROM. We don't have this ability with BIOS/PC's.

So try and identify the EEPROM on the card and find the ground pin>create the switch circuit and see if the system will boot with the switch off.

If all that works, make the switch to on, and launch the flashing utility and see if you can recover the card by flashing a known good ROM to it.

If that fails at any point let us know and we'll dig deeper into the issue....

Well, the flash chip on my UL4D is an ST brand M29W040B in PLCC package (or I think is it, I was unable to identify other candidate ICs on the board). According to the datasheet the ground (Vss) is on pin 16, the third from bottom left corner with notch on top left corner. I think I could lift the pin without damage anything using soldering iron, fresh stain and much care but for sure I will not be able to re-solder it in the same position, on the IC hidden side. About the switch for re-flash the EEPROM, may I use any other ground signal from the nearby and more accessible ICs? I know it as noob question but I does not handle PCB maintenance every day

 

[SnakeCoils]

Success! I was able not only to bring my UL4D back to life but also to flash the OS9 firmware 1.00U that previously bricked the board. I can confirm that ExpressPro Tools 2.8.2 now see and scan both SCSI channels but I have not yet tested the OS9 bootability.
It was a very tricky and annoying process, a PC with PCI slots and XP installed was also needed along the ATTO Configuration Tools, Drivers and flash files for Windows, plus others that were on the PC session of the ATTO ExpressRAID install CD... tomorrow I will detail all the steps, now I really need to go to sleep ;-)

 

[SnakeCoils]

Okay, here is the procedure I followed to reanimate my almost bricked ATTO UL4S.
Tools needed for reflashing to a working state:

- An old Windows XP real PC with PCI slots and a working 1.44 Mb floppy drive installed
- ATTO SCSI driver for XP (winx86_drv_epciu320_314)
- ATTO Config Tool Utility for XP (win_app_configtool_325)
- ATTO firmware for UL4D (win_fsh_epciu320_080218)

Install the SCSI drivers, the Utility and unzip somewhere the folder with the provided firmwares
The Config Tool Utility is almost a carbon-copy GUI of the OSX version so the next steps are the same either on XP or on a Mac
Open the ATTO Utility and go to the Flash sub-page, browse for the correct firmware (it is the one and only with *.ul4d extension) and then start the flashing process with the appropriate button. That's it.
On the next reboot the card will be fully functional again, also the Mac Firmware has been restored so it can be used again on OSX too.
At this point we are where all this story started and, guess what? If you perform another firmware flash on OS9 with the 1.00U updater the card will brick again.
I think that from a certain version on, the microcode structure changed so much that only subsequent versions will flash correctly and going back to some really old ones would not be possible without bricking the card... or at least this is what we tought ;-)

The UL4S (like many similar controller cards) has a configuration utility embedded in the low level boot code that after the POST screen perform a scan of the two SCSI chains but also permit to configure some internal parameters AND also performing a flash of a different firmware that must be present on a physical floppy drive with a precise structure of directories. To launch this tool it is sufficient to press CTRL-Z on the keyboard once, after a couple of second the menu will pop up and after a little wait the menu items wil appear on the center of screen.

Of course we need a firmware to flash, and every ATTO drivers CD has an hybrid ISO that presents for OSX or Windows only the files that belong to a certain operating system so in the PC partition I found some DOS files that not only performs an update on the PC code but ALSO on the Mac code!

A script in the archive has assembled the correct files and directories on a 1.44" floppy disk and then from the bootstrap Config Memu (CTRL-Z) I was able to re-flash the UL4D with both the 1.00 firmware version for Mac and PC, the result can be seen in the attached picture and, of course, it was fully bootable under OS9... ;-)

No further flashing is needed on OS9, the Mac firmware is already there! I must say that with this firmware the card is almost unresponsive under OSX, the device drivers are too much newer for this but on OS9 is pure joy!

 

[SnakeCoils]

Good news!
I was able to update the UL4D firmware for OS9 to the latest 1.01f0 available using the same method above. The link to the self-extracting archive to looking for is epci0101.exe from old ATTO WayBack Machine snapshot.

Copy the contents of the archive on a 1.44" floppy FAT formatted, launch the ATTO Utility (it must be the latest 2.25, the version is on the top line on screen) after the POST screen on a PC (CTRL-Z), go to the Flash page, insert the floppy. Start. After the flash you will be good to go also on OS9

Also my troubles under OSX are gone, also with latest 4.42 drivers, it was not a problem with the UL4D but with the HDD drive itself.

I have two Seagate CHEETAH ST318404LW in two different LaCie enclosures, both of them worked well with the UL3D adapter (OS9 and OSX) but on the UL4D one of them was a nightmare to use under OSX, it also prevented the system to shutdown!
So I have tear them apart to see the differences but nothing visible, also the only internal jumper (Delay Motor Start) was in the same place on both HDDs, then I noticed two different firmware revision:

Rev. 002 - Troubles on OSX, perfect in OS9
Rev. 006 - Perfect either on OSX or OS9

Al the rest was the same: same cables, same terminators, same troubles swapping the UL4D ports, it was really a Seagate firmware matters...

I searched a bit on the net about firmware updaters for the ST318404LW but I found none, if someone is able to find one please tell me where I must looking for

 

[ojfd]

* Check if drive's installation manual has any links in it.
* I don't see anything here at the first glimpse:

HDD Firmware Downloads. PC-3000 Support Downloads. Data recovery and HDD repair tools

Hard disk recovery software, HDD firmware updates, various tools downloads. Powered by HDD Guru Community.

files.hddguru.com

* Try to ask on their forums. SCSI is not that old. Recovery guys might have it. Or some Seagate tools.
* What about this? Might be a starting point.. They talk aout some 'sgtfw1.iso'
dell.com/support/home/en-uk/drivers/driversdetails?driverid=r59709
* Contact Seagate tech support.

Also, I don't remember ever using 'Delayed Motor Start' on Seagates while using OS 7.6.1-9.2.2.

 

[SnakeCoils]

Thank you very much Some of the links you provided I already stumbled upon in the past hours, anyway I will search again.

In the meantime I went in touch with the guy I sold the UL4S some time ago and he was so kind to send me a Stuffit archive with ATTO ProTools 2.8.2 that include the very latest UL4D/UL4S Updater "version 1.0.1 Rev B"

Since I had the UL4D flashed to 1.0.1 version using the PC flasher procedure the updater tells me that version was the same but I decided to go ahead anyway (one of the option allows a re-flashing even the adapters with the same firmware) and the process this time went succesful, no bricks.
Apparently nothing was changed, the OS9 System Profiler reported 1.0.1 as before but when rebooted in OSX I've seen with my big surprise the ATTO Control Panel Info page this time reported the correct firmware version!

With the previous 1.0.1 firmware (from PC flash) that info was something like "10 november 65536" but now it is perfect!
In short this is the procedure I reccomend for a safe OS9 flashing

- In a PC with XP installed and working floppy drive insert the UL4D/UL4S and install the latest ATTO drivers/config tool then flash the board with appropriate latest PC firmware (see my previous post)
- Prepare a FAT formatted floppy disk with unzipped files from "epci0101.exe"
- Reboot the PC without the floppy inserted and when you see the ATTO Config Utility screen press CTRL-Z
- Insert the floppy and select "1. Adapter Menu" and then "4. Update Flash ROM"
- Go ahead, if everything went smooth in a couple of minutes the red window disappear going back to menu
- Turn off the PC, extract the UL4D/UL4S and put the card in a OS9 PowerMac (in my case, an MDD)
- Boot in OS9, look in System Profiler if the SCSI card is version 1.0.1, if not repeat the procedure from beginning
- Extract the Stuffit Archive attached to this post
- Go to: MacOS 9 / ProTools / Scsi / ATTO Firmware Utls (SCSI) / ExpressPCI UL4D/UL4S Updater
- Click on the only executable you will find there "ExpressPCI UL4D Upd 1.0.1U Rev B"
- Choose the option that force the adapters to be updated disregarding the current firmware version
- Wait for a couple of minutes and the updater should tell you everything is fine
- Reboot the Mac in OS9 and check the Sytem Profiler still report version 1.0.1
- Reboot in OSX (if present) and check in Config Tool the flash version is reported as in the picture above

Done!

I don't know if this "Revision B" of the flasher allows the ATTO card to be programmed straight in OS9 without corrupting the flash Eprom as the "1.0.0" did but in doubt I would follow the procedure above.
Keep in mind you are assuming every risk, for me it worked, I cannot guarantee the same on other systems.

Enjoy!

 

[DearthnVader]

If there is a later version of the Mac firmware for the SCSI card that works better with OS X but lacks OS 9 support we can likely extract the OS 9 'NDRV' from the older firmware and add it to the later.

Assuming there is enough room on the EEPROM.

 

[ojfd]

This sound like a good idea, if there will be actual compiled binaries published and if it doesn't just turn into programmers festival, where they're show-off their skills in front of each other and send others to GitHub.