…or the chaos is overwhelming
This time I present to you the fourth part of the epic Tesseract project. Links to the previous three posts are below:
Now, let’s get back to work :>
Buddha flash card
This is the third try to install a Buddha flash card. This time it is the Buddha Flash 20-year anniversary edition and this one turned out to be perfect for this project. However, this time I didn’t even try to solder it directly. Instead, I’ve figured out a way to install it parallel to the mobo and still make it fit under ZZ9000. It required a bit of tinkering with a zorro socket.
This is how it was done.
With the first row sorted, I had to find a solution to connect a second row. I installed a curved, single row of gold pins and soldered it to socket pins one by one.
However, I had to remove two caps that collided with the card and find a new place for them. I did that by designing and 3D printing a small holder.
And it fits! Yay!
And it works!!! Wooohoo!!
From this point on, it is total chaos but I think it is necessary to show you how this all goes. I do mistakes that are corrected later but mistakes have to be documented too 😀
I’ll try to comment on some pics a bit more.
It all happened all of a sudden. The Amiga started to behave weirdly – random hangs or GURUs. Obviously, I am working on a totally custom project with lots of mods so it is very hard to control such a situation. It was eventually fixed and I will show how but just have a look at how it all developed 😀
I’ve started working on a keyboard/mouse part. The goal was to somehow connect Bluetooth-driven KB and a wireless mouse but through a mouse/joy auto switcher.
I’ve tested it first without a mouse/joy switcher to see if it actually works
The BT keyboard was connected through its dongle plugged into a SUM 234. That USB wire is only for charging.
The mouSTer was plugged into a temporarily soldered DB9 socket.
It all worked! 🙂
GURU meditation … again
The Amiga started to throw GURUs and hanged up at random moments. At that point in time, I totally didn’t know what was wrong, moreover, I didn’t realize how far from problem discovery I was 😀
However, I’ve started investigating a ribbon wire connecting ACA500+ and a mobo.
I’ve quickly discovered that there is a problem with one of the wires being slightly off.
I’ve fixed it by reclamping this IDC connector.
Then I decided to design clamping brackets for this connection to make it more durable. After a few 3D printing sessions, it was done. It was to be reprinted in black plastic later. Obviously, at that point, I didn’t know that I will remove it completely when nearing the end of this issue.
Above work didn’t help so I was like … maybe it is the heat?
This came to my mind because I started to work with this Amiga a bit longer than quick boot and test if “something” works. I’ve decided that I need to design a cooling solution for a 030 CPU on an ACA1233n.
My initial idea was to use a heat pipe stolen from a laptop, but after a few measurements, it turned out to be a dead end.
This is when I created a 3D-printed heatsink holder. Obviously (AGAIN!) I didn’t know that I will have to remove it later because of the complexity of this project 😀
At this point, all came back to normal so I moved to another stage – The clock port adventures!
The clock port adventures
The short story is that while writing this post (and I still have material for another two posts) I am still trying to solve clock port issues. Clock ports are vastly undocumented and the major issue is that devices can be connected in two different ways but … if connected in the wrong way, it results in either a broken clock port or a device …
Let me paint you an idea of what was planned. We wanted to connect several add-ons through the famous Amiga clock port expansion port.
As Wiki describes it “The port is a remnant of an abandoned design feature for the addition of internal RAM and a clock for timekeeping. However, it was later widely used as a general-purpose expansion port by third-party developers for devices, such as I/O cards, sound cards, and even a USB controller. Although a real-time clock can be connected to the port, the clock was typically added by other means (usually integrated on CPU or RAM expansions) which leaves the clock port free.“
Obviously, Amiga 2000 doesn’t have a clock port but we have expansions that have it! Through these expansions, we can and want to connect four devices:
I’ve started working with RapidRoad USB and the best way (for this project) was to connect it to a clock port located on an ACA1233n. However, it didn’t work at all. The expansion was simply not detected. However, it was detected through an ACA500+ clock port. Did I mention that I’ve killed that expansion (RRUSB) twice lol? Fortunately, Jens of iComp, designed this one nicely and it was just a matter of replacing fried resistors that were placed there for morons like me 😉
Anyway, MrTrinsic wanted me to test a simple Zorro to clock port adapter and he’d sent it along. It was a bit different solution that we’ve tried to use in an Amiga 1000 Phoenix project but still, it was not designed for a A2000 😀
I was a bit against this test because I knew that gold-plated (ENIG) pads on a MoBo will hold these gold pins strongly.
However, I did it! I just had to slightly trim the original angled gold pins to optimize room usage.
In the meantime, I’ve 3D printed fresh and BLACK ACA port covers.
Did I mention CHAOS already?! 😀
This project is so long and complex that I don’t even remember why we’ve abandoned this expansion, but that is not important since the Amiga started to randomly crash again …
Obviously, this happened after ANOTHER re-assembly (after soldering and desoldering the clock port card) and this time after a few days of fiddling, I’ve narrowed it down to an ACE expansion – FatAgnus replacement.
However, I didn’t figure out what exactly causes these random crashes, even though it was already visible on the pic below … but more on this later.
After re-seating the ACE, the problem kinda stopped so I could move on to clock port tests again.
The next stop is Buddha Flash which has a clock port. After some tests, it turned out that that Rapid Road USB works nicely through it.
Anyway, I’ve left these problems for later because time flies and had a lot to do.
Next step – add-ons location, which means another measuring round.
Various options were considered.
Chaotic front panel job
I’ve started working on a front panel. I’ve divided it into a few parts – The power button(already described in previous posts), LCD, Mool logo, and some more stuff 😀
Let’s start with …. everything lol
The front panel part hidden under a flap has many proper holes (no pun intended lol). The plan was to re-use almost all of them but I had to figure out how to approach this first.
I’ve come up with an idea to use the original PCB that was supplied with this PC/Amiga case but I had to desolder some components, as I don’t need a firewire nor audio jacks, and cut it to a proper size. This is how I did this.
I designed a few studs/brackets to hold it in place
Next, I’ve 3D printed a simple holder for two SD card slots with flex-wire extenders to connect to ACA500+ and HxC.
The idea was to put both PCBs in a CF card slot hole.
Back to a drawing board …
I’ve decided that I’ll rework this panel and 3D print it from UV resin. Below, is the first prototype.
Then, the second, third, and so on 😀
Until I came up with a proper solution.
From left to right – USB slot 1, USB slot 2, EMPTY, DiSmo button (AVA500+), Bootselector, three HxC buttons, and two microSD card slots.
Hxc and the Mool logo
Next in line was a panel on the right. I wanted to put HxC LCD there and decided to work on the Mool logo while it was disassembled.
The problem was that the original LCD is a bit too big. Fortunately, I was able to trim it to a proper size. Then it was only a matter of designing and 3D printing a bracket for it.
The original backlight had to be changed to blue. I’ve made it from scratch using some parts that I had on the shelf.
The Mool logo will be powered via a HxC 5V rail to save space. I’ve also placed HxC in its target space using some 3D printed studs, soldered front panel wires of control buttons, and routed a flex microSD ribbon wire to it.
The final outcome.
End of PART 4
You’ve reached the end of part 4 of the Amiga Tesseract project. Stay tuned for more episodes of this build as there are at least two posts coming 😀
If you want to get the retro gear I am manufacturing or hardware modules, please visit shop -> https://retrohax.net/shop/