Page 1 of 9

Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Fri Feb 08, 2019 5:03 pm
by intric8
First, a Brief Re-cap of the Rejuvenator Project Prior to this Update
Update 3.141592 (Ad Infinitum, Feels Like)

One of the volunteers on my team named Joe is based in Idaho. You may know him around these parts as Joethezombie. He is a talented circuit board analyzer and designer. Rather than send my one-of-a-kind bare board to a company to be sanded down and recreated, I sent Joe the board for him to try and duplicate instead.
IMG_6998.png
I also sent him hi-res photographs of my working original Rejuvenator.
IMG_4785-1-b.jpg
Over time and a lot of work, Joe recreated the schematics for the Rejuvenator by hand and has since printed 6 prototype home-brew boards.
IMG_2762.png
He discovered along the way that the bare board appeared to be an early prototype, as there are differences between it and my working board.
differences-gif.gif
Joe worked those changes into his schematics, too.

He has actually taken one of the prints and begun the exciting process of populating it with chip sockets and other necessary components. Folks - look at this thing!
3IMG_2822.png
Meanwhile, I happened to be so lucky to find and source another Rejuvenator thanks to one of the members on the Rejuvenator Team, Adrian who is based in Springfield, MO. Thanks to his alert, I was able to acquire this new Rejuvenator from a gentleman in Wisconsin. I had no idea if this board worked - and neither did he. But I paid for the board in the hopes that it would, and worse case scenario I figured it could be plundered by Joe if necessary or at least provide deeper hands-on analysis.
IMG_9249.jpg
But what I really wanted was for it to be a working board so we could pull the PAL chips off of it and place them into Joe’s brand new board. Looking at the age of this Rejuve2, Adrian actually thinks there's a decent chance its PALs aren't read protected.

The PALs (which stands for Programmable Array Logic - not the video PAL/NTSC stuff) are the final piece of the puzzle as we’ll need to decode the calculations that are stored on them and eventually write them to our own chips. But before we do that, we need to prove our new home-brew board even works.

So, I installed the newly acquired board (what I have been calling internally the Rejuve2) to a spare stock A1000. I didn’t want to have to pull chips off the one known working board except as a last resort.

I flipped on the switch - black screen. I got a power light, yes, but only a black screen. That was a massive bummer. So much so I had to take a few days off of hardware projects just to clear my head.

I then replaced Paula, Denise, and all 8 RAM chips. I had working RAM from the original when we upgraded Rejuve1 to 2MB.

Still had a black screen.

I then took detailed photos of the installed board and sent them to the team. Christian looked at them very closely and compared them to images he had found of working Rejuvenators online.

And then he saw it.

There is no “notch” on the silkscreen for the Kickstart ROM, and it appeared to him that the KS was installed upside-down. This was the way I had received it from the previous owner (the board had been fully populated) and I assumed it was correct.

I put on my jeweler’s glasses (47 year old eyes, guys, gimme a break) and Christian was right! Of course, that ROM was completely fried.

But I had a spare. I quickly realized I need more KS 1.3 ROMS, as I only had a 3.1 in my stores. But regardless I popped in 3.1 and voila! The Rejuve2 sprang to life! I couldn’t believe it. We were back in business!
IMG_9586.jpg
Next steps

I have shipped the working Rejuve2 to Joe. He’s going to pop the socketed chips off the working board and put them on his new home-brew. Then he’s going to install it to his own A1000 and flip the switch. And before that, he's going to try and read this board's PALs. There's a chance they won't be protected. And as Joe told me today, "It would be a moon-landing stoke of luck if they’re readable." Because if they are readable, we're virtually home free.

But assuming Joe's board works (oh my god you guys, we’re so close!) and the PALs are protected, we will officially be down to the final step: recreating the calculations in the PALs ourselves. And we’ve already gotten a head start on that as well.

More soon, and thanks for reading.

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Fri Feb 08, 2019 6:58 pm
by rpiguy9907
Super exciting! Talk about a labor of love!

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Fri Feb 08, 2019 8:06 pm
by nonarkitten
Amazing work!

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Sat Feb 09, 2019 9:36 pm
by BloodyCactus
I suggest running ReGAL on them, protected or not.
Reverse engineering locked PAL/GAL
Another aspect of this project consists in facilitate the reproduction of locked PAL devices. When the fuse map is not readable anymore using the standard programming method, one alternative consists in dumping the truth tables by probing the outputs of the device when applying all the possibles input combinations.

Obliviously, this method is limited to devices implementing combinational logic (PAL working in registered mode are out-of-scope for this proof-of-concept).

The current implementation uses an arduino mega to drive the device under test. The setup required to perform the IO probing on the example defined above can be described as follow:

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Mon Feb 11, 2019 5:24 pm
by Jsdcsdja
Exciting work!!!

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Tue Feb 12, 2019 7:40 am
by TommyE
voodo, wizzardy and blackmagic happening :)


Great job

Tommy

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Tue Feb 12, 2019 9:49 pm
by Mr.Toast
Did you figure out a source/solution for the pin extensions? Those seem pretty exotic from this altitude. I would also imagine solder alignment of them would be a nightmare.

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Sat Feb 16, 2019 4:43 pm
by intric8
Mini-Update:

My Rejuve2 is in Idaho and was successfully installed by Joe on his Amiga 1000. He had never installed one before and needed to go through the process.

He then used Mattsoft's 3D-printed spacers that I supplied and mounted his home-brew board to his 1000. It was at this point he realized he had a slight issue with his board.

There's an area where a long rectangular slice is cut out of the Rejuvenator to let some capacitors stick through. His hole was 1mm too narrow. So he had to file it down to get the caps to pop through. Not a massive deal, and something easily fixed with future revisions.

Next:
He is going to put some dummy PALs with some dummy data on the home-brew and perform a "smoke test". We don't expect the machine to boot, but we do want to ensure the PALs aren't altered in any way. After he boots up and back down, he'll re-read the dummy PALs to make sure nothing changed.

Assuming all goes well, he'll then perform the PAL chip transplant from the working Rejuve2 and the Home-Brew. And then flick. Da. Switch!

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Sat Feb 16, 2019 7:19 pm
by oldbull
Transplant... throw the switch...

We know how that movie ends. :shock:

Re: Rejuvenator Update 3.141592 - You Can Almost Taste It

Posted: Sun Feb 17, 2019 5:52 am
by TommyE
intric8 wrote:Mini-Update:

My Rejuve2 is in Idaho and was successfully installed by Joe on his Amiga 1000. He had never installed one before and needed to go through the process.

He then used Mattsoft's 3D-printed spacers that I supplied and mounted his home-brew board to his 1000. It was at this point he realized he had a slight issue with his board.

There's an area where a long rectangular slice is cut out of the Rejuvenator to let some capacitors stick through. His hole was 1mm too narrow. So he had to file it down to get the caps to pop through. Not a massive deal, and something easily fixed with future revisions.

Next:
He is going to put some dummy PALs with some dummy data on the home-brew and perform a "smoke test". We don't expect the machine to boot, but we do want to ensure the PALs aren't altered in any way. After he boots up and back down, he'll re-read the dummy PALs to make sure nothing changed.

Assuming all goes well, he'll then perform the PAL chip transplant from the working Rejuve2 and the Home-Brew. And then flick. Da. Switch!
Ahh this is so exciting, thanks for the update. Im so eager getting a rejuvenator...

Regards
Tommy