SukkoPera's Open Source Hardware: A500/A600 Chip/Fast RAM Expansions

[AmigaHope]

Originally Posted by jbilander

I have now made the repo public, it is available on my github. Hopefully it'll benefit the Amiga community and people who build on this will respect the license and share their work back to the community.

Any hope of adding some "free sex and beer" address pins that we could wire to other parts of the board to let it use larger slow RAM expansions as chip RAM? Or even just include RAM that occupies slow RAM address space on the chip itself? (Obviously would require either patched ROM or something in the startup sequence to swap the memory type to chip)

[AJCopland]

Nice work on the project! It's really cool seeing you all developing this stuff.

This is the repo' I assume.
https://github.com/jbilander/ReAgnus

[SukkoPera]

As it turns out, after quite some time someone finally understood that Open Hardware is about trying to do your part rather than coming up with heaps of useless talking and moaning or blatant stealing and allowed this project to reach the next milestone.

So despite knowing he could be jumping into the void, fellow Italian living in Japan Edoardo Auteri put confidence, money, time and effort into assembling a board. He had some initial issues but these were tracked down to a faulty Paula chip. After replacing it the board booted right away so he went on with thorough testing of almost all system components:

• Internal Chip RAM (1 MB Configuration)
• Composite and RGB video output
• Audio output
• Trapdoor Chip RAM expansion
• Keyboard and leds
• RTC
• Serial port
• Kickstart 3.1.4
• Internal and external floppy drives
• Integrated drive switcher
• Several games
• Controller ports with Mouse/Joystick/CD32 Gamepad
• 12h burn-in test with *Mortal Kombat 2*
• TF536 basic testing
• 8h burn-in test with TF536
• Raspi-based RGBtoHDMI adapter
• Original shielding/case fit

Basically he tested everything except the Parallel Port and the Edge Expansion Connector. The board fully passed every test, proving to be perfectly stable and reliable. Edoardo only encountered one minor issue plus another one I was already aware of, as it had been reported by @LIV2, both of which were easily corrected. He also came up with a number of suggestions for improvements which I incorporated in a new revision of the board which is now under testing.

The new board will be released as soon as it has been tested. The full design files will be available but I changed the license to CC BY-NC-SA. TBH, if it were for me, I wouldn't have released any sources any more but I felt that Edoardo did such a huge job that he deserved to be the one to choose what to do (besides getting his moniker among the credits on the silkscreen). He chose to keep it open source and so be it. This means the "community" will soon get fully tested schematics and gerbers for an Amiga 500+. A lot of great things could be born from these, let's see what happens.

Someone else I would like to thank is an Englishman named Ibrahim Imran: he ordered some boards and even sent one to me for free. He intended to try and assemble one but was halted by personal circumstances. I am a bit worried because I tried to get back in touch with him several times but he never replied. I hope he's doing well. The board he sent me was first of all a pleasant gift (you know... after 2 years of work on that project, you finally get a physical board in your hands, aaaaaah!) but it also proved very useful for testing the dimensions, connector placement and choosing positions for the new features.

One more time, this is a small example of how the open development model works and how it thrives when everybody does a little bit. I really wish the "community" really took this to heart and started supporting open projects rather than the selfish and greedy speculators they are standing up for right now.

[ExtremeWays]

What a beautiful piece of work.

Great job Edoardo, Ibrahim and SukkoPera!

I think it's time to for me to learn to solder!

[kinmami]

Originally Posted by ExtremeWays

What a beautiful piece of work.

Great job Edoardo, Ibrahim and SukkoPera!

I think it's time to for me to learn to solder!

Thanks for the appreciation.

Working now on V2…

[cpiac64]

Good job my friends

[sampedenawa]

Originally Posted by kinmami

Thanks for the appreciation.

Working now on V2…

Which directions are you going to ?
Do you consider inserting modern features directly on board (e.g. kickswitcher, rgb output, Gotek, USB adapter, etc) or what ?

[SukkoPera]

KS Switcher and Drive Switcher were already integrated into V1. The things we added in V2 are some "commodities" that are nice to have but nothing major, as we still want to avoid any significant rerouting.

[katarakt]

Very great work from Sukko and all their helpers. Thanks a lot for all your hard work you are doing for the community

[Mathesar]

Wow.... Great work!
This is indeed how open source hardware should work!

@sukkopera, all of the (small) hardware stuff I've done for the Amiga i've released as "free" / public domain without using any specific license. The 14Mhz accelerator I am doing I plan to release under CERN-OHL-S V2. From what I understand of it, this still allows someone to sell complete products based on the design as long as they release all original sources and changes.

You did not choose this license, why?

PS; in the case of the 14MHz accelerator it is actually a bit tricky as it builds on some work by others. Following the rules I would have to get their consent to release this under the CERN license. Otoh, i have changed so much that the original design is barely recognizable anymore...

[Mathesar]

And, in the spirit of open hardware, I would like to make a design question/suggestion (and implement it if desired): My A500 sufferers from these annoying vertical clock inteference lines when connected to an LCD. Does this design fix this? Do we know how to fix it?

[kinmami]

Originally Posted by Mathesar

And, in the spirit of open hardware, I would like to make a design question/suggestion (and implement it if desired): My A500 sufferers from these annoying vertical clock inteference lines when connected to an LCD. Does this design fix this? Do we know how to fix it?

sorry to jump into the discussion. I am working "closely" with SukkoPera from Japan.

First of all, appreciated the kind interest and positive comment.
Concerning the LCD issue, based on my experience, is due to (1) poor grounding/shielding of the original A500 motherboard (any Rev.) and (2) not buffered digital H- and V- which are straightforwardly exposed to the DB23 output connector.

Well, V2 already offers improved grounding (1) and optionally (2, if not genlock used) the onboard buffer for digital sync signals. of course, the PCB is still a 2Layer stack-up.

4Layer would be a picture-perfect solution yet not cheap.
A cost-effective DIY solution implies a trade-off.

[SukkoPera]

Originally Posted by Mathesar

Wow.... Great work!
This is indeed how open source hardware should work!

@sukkopera, all of the (small) hardware stuff I've done for the Amiga i've released as "free" / public domain without using any specific license. The 14Mhz accelerator I am doing I plan to release under CERN-OHL-S V2. From what I understand of it, this still allows someone to sell complete products based on the design as long as they release all original sources and changes.

You did not choose this license, why?

I haven't read V2 of the CERN license yet, but V1.2 is what most of my open stuff is released under.


Now, the reason I used to release everything as an open project is because I hoped that someone else could take it and evolve it or derive something from it.


This did not happen very often, while people taking my designs and selling them at ridiculous prices happened much more often.


I believe this project will spark quite some interest and speculators will dive in pretty soon. With a NC license I can stop that. I will add a clause that if anybody would like to do a major contribution and is prevented to do so by the NC license, they can ask and they will be granted a separate more permissive license.


I'm really sorry for that, but the only alternative I had was not releasing anything at all, this thing has really upset me.

[robinsonb5]

Originally Posted by Mathesar

My A500 sufferers from these annoying vertical clock inteference lines when connected to an LCD. Does this design fix this? Do we know how to fix it?

The jailbars are caused by the /C1 signal on pin 15 of the 23-pin video port which is just left dangling and thus acts as a nice little antenna. Terminating that pin with a 47Ohm resistor to digital ground (pin 13's convenient) solves the issue.

[Daedalus]

Vertical jailbars can be caused by several different things, but when the problem is specific to an LCD screen, in my experience it's most likely down to the sampling done on the signal by the monitor. Neither buffering nor grounding will help that - it's simply a mismatch between the horizontal resolution and timing the monitor expects and that which the Amiga outputs.

Some monitors have a phase adjustment which can help this situation - if you adjust the phase and the jailbars shift significantly and independently of the graphics, then this is the cause. If the jailbars stay in sync with the graphics when you adjust the phase, then look elsewhere for the cause.

[robinsonb5]

Originally Posted by Daedalus

Neither buffering nor grounding will help that - it's simply a mismatch between the horizontal resolution and timing the monitor expects and that which the Amiga outputs.

That's absolutely right, but doesn't explain why the jailbars appear in areas of solid colour. If the Amiga's video output was clean there'd be no variation in signal level over the duration of a pixel, and thus nothing to cause interference patterns.

How can you make the Amiga's video output cleaner? The single most effective way is to tie off that open unterminated clock signal on the video port. I did this on my A500's DB23-VGA adapter a few weeks ago and the jailbars are quite simply not there any more (at least in solid areas - obviously you can't eliminate pixel clock interference patterns unless your monitor's phase/pixel-clock settings are flexible enough to achieve a perfect sync.)

[Daedalus]

Indeed, which is why I said they can be caused by several different things. we don't know the exact nature of the jailbars mentioned, and IMHO it's prudent to investigate and potentially eliminate simple causes before taking a soldering iron out.

[robinsonb5]

Originally Posted by Daedalus

Indeed, which is why I said they can be caused by several different things. we don't know the exact nature of the jailbars mentioned, and IMHO it's prudent to investigate and potentially eliminate simple causes before taking a soldering iron out.

Fair enough - and I'd always advocate modifying the DB23 to VGA adapter rather than the Amiga itself, in case that wasn't obvious.

[Mathesar]

Originally Posted by robinsonb5

How can you make the Amiga's video output cleaner? The single most effective way is to tie off that open unterminated clock signal on the video port. I did this on my A500's DB23-VGA adapter a few weeks ago and the jailbars are quite simply not there any more (at least in solid areas - obviously you can't eliminate pixel clock interference patterns unless your monitor's phase/pixel-clock settings are flexible enough to achieve a perfect sync.)

I'm going to try that. I also have an Amiga 3000 and a Minimig. The MiniMig does not display any jailbars but the A3000 does, also when using the DB15 output I think... Gonna check now!

EDIT: The A3000 only has it very slightly over the DB15 connector, both in normal and scandoubled mode. Gonna try terminating the clock this weekend!

[Mathesar]

Originally Posted by kinmami

sorry to jump into the discussion. I am working "closely" with SukkoPera from Japan.

First of all, appreciated the kind interest and positive comment.
Concerning the LCD issue, based on my experience, is due to (1) poor grounding/shielding of the original A500 motherboard (any Rev.) and (2) not buffered digital H- and V- which are straightforwardly exposed to the DB23 output connector.

Well, V2 already offers improved grounding (1) and optionally (2, if not genlock used) the onboard buffer for digital sync signals. of course, the PCB is still a 2Layer stack-up.

4Layer would be a picture-perfect solution yet not cheap.
A cost-effective DIY solution implies a trade-off.

So the design already improves on video? .
This would be a cool winter project to build a new A500 motherboard!

Again, great work!