User avatar
retropanic
Switzerland

Posted Sun Jun 24, 2018 5:44 am

Perfect Project Eric!

I am in for at least one or two of these.

Of you start the campaign i am in!

Cheers from Switzerland

User avatar
bonoedgey
Canada

Posted Mon Jun 25, 2018 10:32 am

this is an amazing, thoughtful adventure you are on Eric. it may be just what my A1000 project in progress needs..
Thanks again for all the hard work
Look forward to the road ahead with you.
:D

User avatar
AmigaDave

Posted Thu Jun 28, 2018 3:16 pm

Just signed up here to comment on this thread and project. I have 2 or 3 A1000 computers, all NTSC models, and at least one of them has a daughterboard inside, maybe they all do. The A1000 has a very fond place in my memories, as it was my first ever computer, and I learned so much during the years that I used and expanded it, before upgrading to an A2000, and the beginning of my Amiga addiction/collection.

You can add me to the list of users who is interested in purchasing at least one Rejuvenator board, either in kit form, or pre-assembled. I also have a bare pcb for the GB1000 motherboard replacement for the A1000 computer, but have not yet even begun to source the parts required to complete that project, and have heard from some users, that it is a difficult project to complete, with translation of the a1k forum threads and messages about that project, from german to english, being just one of the challenges to overcome.

My original A1000 was completely stock when I bought it used some time in 1986, but by the time I purchased my A2000 to use as my primary computer replacing my A1000, I had upgraded it with the 1.5mb Spirit Insider RAM expansion, an Add-IDE hd controller, another device which allowed the first external floppy drive to be recognized as DF0:, instead of DF1:, so I could replace the internal floppy drive with a 3.5" hdd (can't remember the exact name of it), and a Kickstart ROM switcher, with 2 different ROM sockets, allowing the installation of any ROM chips, from 1.2 to 3.1, to be booted from, simply by soft-booting for 5 or more seconds to switch from one ROM chip to the other, plus a jumper, which could be used to allow going backward and booting from a floppy disk. I installed a 105mb 3.5" IDE hdd, which at the time was massive, and one that I thought I would never completely run out of storage room while using (Amiga programs were so small in those days). Although I have not booted that original A1000 for many years, and I have removed some of the components/parts from it, I still have it stored in its original box, protected from damaging sunlight, and waiting to be used once again.

I would love to obtain a brand new Rejuvenator board to install into my A1000, as I am unsure if/when I will be able to complete the assembly of my GB1000 motherboard replacement, and I have at least 1 or 2 other A1000's I can install it into, if/when my GB1000 is finished. ;)

User avatar
intric8
Seattle, WA, USA
Website

Posted Thu Jun 28, 2018 11:08 pm

AmigaDave - wow, just wow. Reading your post made my heart flutter, I swear. What an awesome journey you've taken with those machines! To this day I have one A1000 sitting next to my A2000. That is a knockout combo. I wish, I really do, I had enough room to stick my C64 next to those two beauties at the same time - I'd be in total and utter heaven. I'll find a way eventually (I have to have more electrical sockets installed by a pro in my C= room).

The BOM has been completed by my friend Mattsoft THIS WEEK, and I instantly delivered it to the local CB shop we hope will take us the promised land.The ball is in their court now to return some sort of price guestimate, which I should be able to use for creating the GoFundMe campaign. That's really the next milestone (or hurdle). If we get past the GoFundMe successfully, it'll be up to this shop to redraw the board (they have everything they need) and get them printed. Then the final shoe shall drop, as we'll see then if our sourced parts with the new boards actually work and pass QA.

And if they do? Holy crap, Dave! Ho-lee-crappage.

Boards shall be built. And you'll get one. I doubt we'll ultimately make more than 30-40... hard to say. But for the few of us that want one - a new one - hot damn! Wouldn't it be cool? Fingers crossed.

And thanks for the backstory. Love that.

User avatar
TommyE

Posted Sat Jul 28, 2018 10:50 pm

Hello

Just registered when I found this thread.

I’m sitting here smiling (alot) when reading about your passion and persistence.

You have Done an incredible job, Thank you. :)

Owning a PAL A1000 with daughterboard I’m in.

Thanks
Tommy

User avatar
intric8
Seattle, WA, USA
Website

Posted Sun Jul 29, 2018 6:50 pm

Mini-update 1.5

This week with the help of the forever-awesome Mattsoft, who designed and 3D printed some custom spacers and guides for the original Rejuvenator, I've been able to successfully mount the board in one of my A1000s! First try, too.

THANK YOU, MATT.

So, in the next few days I'll reassemble the machine (if not tonight!) and make sure our board actually works. Kind of hard to reverse engineer a thing that isn't functional, right?

Also, a bit of a road bump.

A couple of weeks ago the local company we were working with to build the board for us pulled out of the project. We did everything they asked, and the owner of the company felt very bad about doing it. But he was worried we'd crowd-fund the project, burn through all of the money and still be left with a non-functional product.

It's certainly a possible outcome. And he didn't want to place his company in that position. So...

Also, the owner of the bare Rejuvenator board has been having second thoughts about us using his board (if we destroy it). If we took it to a shop, they could sand it down and recreate the GERBERS perfectly. But it would destroy the board entirely in the process. So we may have to try and scan it instead (more expensive - we'll see).

Regardless, Mattsoft connected me to another firm that has experience in reverse engineering retro tech. We're in the early discussion stages, but frankly this is feeling more promising. In the end, this "hitch" may have been a blessing in disguise.

I need to prove the Rejuvenator actually works. And if that works out we should be in pretty good shape.

More soon as it becomes available.

User avatar
joethezombie

Posted Mon Jul 30, 2018 10:03 pm

Thanks for the mini update! This is so super exciting! I have a couple of thoughts, I hope you don't mind me sharing.

I do have some experience in reverse engineering old hardware. In my case, it was an old piece of hardware known as the DayStar PowerCache adapter for the Macintosh SE/30. This adapter board was required to use DayStar 50MHz 030 accelerators in the SE/30. The problem being, they are simply not found anywhere. I somehow, with very much luck, found one at a reasonable price in Europe (strangely enough, on AmiBay), and had it shipped out here to Idaho.

Creating new schematics for the board was pretty elementary, if not time consuming. On the PowerCache there are 3 120-pin connectors, a smattering of resistors and capacitors, and single chip. Using a slow process, I was able to trace every single path and recreate the schematics in KiCad. I verified the connectivity 4 times to make sure I got it right. And finally, it was complete. Since you have a blank board, the same time consuming process can be used to make new schematics quite easily without damage to the board.

Now, my PowerCache adapter board is much simpler in design than the Rejuvenator. As I said, it only has one IC on it-- the chip that contains the magic smoke to allow the accelerator and cache memory to function. That chip, known as a PAL, contains programable logic. The problem is, almost always, when the PALs are "burned" in the programmer, the protection bit is set by default. This makes it impossible to simply read the logic program back out of the chip with a programmer. To make things more confusing, the PAL can be burned in several different modes: Registered, Complex, or Simple. If it was programmed in Simple mode, then you can use an adapter to read the logic functions back out without much trouble. But if was programmed in Registered or Complex mode, then there is practically no chance to retrieve the logic program inside.

On my PowerCache, the PAL was programmed in Registered mode. The only way I was able to decipher the logic was to use a 16-channel logic analyzer and capture the function while the board was running. This took me months to accomplish, and I finally found a break when I discovered the cache functions in an old Macintosh Hardware manual. I'm not sure if I would have accomplished my goal without that stroke of luck.

I bring this up because looking at the pictures of the Rejuvenator you've posted, I see no less than 4 PALs, perhaps more. These are identified by having PAL16L8 or PAL20L8 written on the chip. These will be the greatest challenge of the entire project if they are not programmed in Simple mode. I fear, most likely, they are Registered, and have the protection bit fused, and will be impossible to read back.

If possible, you can check with the amazing developer you were able to find, Greg Tibbs, and see if he still remembers the logic equations in those chips. Otherwise, it's going to take someone with a lot of time and skill with a logic analyzer to decipher. It can help to look at other expansion boards that we know the logic. Address decoders and bus latches are pretty common and most likely are the same from different boards.

The road doesn't stop there with these troublesome little guys. Even if you are able to retrieve or re-assemble the logic from them, creating new ones requires knowledge of an obscure logic programming language known as CUPL. I was lucky I had some friends in Arcade repair that are proficient, and wrote the new program for me.

In the end, we were successful, and created several runs of the board. It was dream come true. I sincerely hope your dream comes true as well, because it is shared by me, and the others that have posted here, to have one of these Rejuvenators for our favorite Amiga 1000.

User avatar
SEADRIVE

Posted Mon Jul 30, 2018 10:21 pm

Here it is the end of July, but I would still love to get in on this game...I have two A1000s that would love such an upgrade path. Please count me in as well...both of my machines are NTSC machines, since I live in Canada...very cool and it's nice to see other fans of the first Amiga...cheers.

User avatar
dalek

Posted Tue Jul 31, 2018 12:32 am

joethezombie wrote:...
I bring this up because looking at the pictures of the Rejuvenator you've posted, I see no less than 4 PALs, perhaps more. These are identified by having PAL16L8 or PAL20L8 written on the chip. These will be the greatest challenge of the entire project if they are not programmed in Simple mode. I fear, most likely, they are Registered, and have the protection bit fused, and will be impossible to read back.
...
Firstly awesome post - sounds like you did a great job reversing that device.

For this one though PAL "L"'s are only ever simple right? It's the PAL "R"s that could be either simple or registered. So hopefully they are all reversible!

User avatar
joethezombie

Posted Tue Jul 31, 2018 7:47 am

Unfortunately, that is not the case. Even the “L” series chips have a registered mode, albeit much less in functionality as an “R” series. Alas, even the PAL on the PowerCache was an L series. What the registered mode accomplishes is the ability to take the output of a set of logic equations and feed them to the input of a another pin internally. These equations can’t be read, because the output and input pins don’t show the results of the equation, it’s handled completely internally. It also provides a method for storing data in the gate array for use in a later equation.

Now, the PALs on The Rejuvinator may not depend on complex equations like the PowerCache did. Like I said before, if they are just address decoders and bus latches, likely the equations can be analysed out or guessed as they are somewhat universal for the application at hand. You can get a good idea of the purpose by identifying what signals go to each pin.

Please be aware, i am making a lot of worst-case assumptions here. It is entirely possible the PALs are not protected, or are simple in nature. My advice would be to decode the PALs firstly, burn new copies, and try them in The Rejuvinator to test function before spending a lot of effort elsewhere.





Return to “Hardware”