Illuwatar's Page

Finally, things are coming together - the board is on the way and I'm eager to start working on it...

But first, a short introduction:

Illuwatar, who is he? I'm just a regular guy closing in to the early 40. I'm living in Sweden but I'm partly Hungarian. Normally, I'm working within the telecom market with testing and developing software for business telecom, but at the free time, other kind of design work i done. One of the bigger projects coming from my tiny workshop is the Mini-ITX MiniMig that are currently developing into version 2.0. When I saw this amazing project, I was instantly hooked. As my first home computer was an Amiga 500, I couldn't resist ordering one board (even if I'm not having an old A1000 to put it in).

Up to now, all have been about waiting (and more waiting) and finding the parts needed to complete the build. Most of the tricky Amiga unique parts have been collected but there are still some strange memory IC's that must be found...

I will also do a parallel project page at my own homepage (where You can find my other projects too): www.illuwatar.se


2009-06-02: Got all the parts for the PSU board, completed it and did a dry run (GBA1000 board not connected). Success on first try. Nothing when up in flames and all voltages looks fine, just as designed. Time to figure out the design for the case of the complete thing. Also, the programming of the CPLD must be solved.


2009-05-16: Finally, I'm back on this project. It took a while due to the things that happened in my family. Also, the MiniMig took some time too as I actually managed to sell off almost all boards I had.

Back to the GBA1000 - the PSU is completed as design, PCB's ordered and they got delivered.


The PCB for the PSU (DC/DC-converter). Matching blue of course...


Assembly started but as usual, I ran out of parts.


2008-12-15: This build and the design of the PSU has been suspended for a while as a tragic event occurred within my family. My mother passed away on the 10'th of December after a 2.5 year struggle against cancer. She was fighting to the end, but sadly, the disease won.

At least, today, I managed to put that Zorro-slot in place. There are suitable card sockets at Digikey.


2008-11-26: Not a big step, but an important one - the 29F800 arrived from somewhere "Over There". I found it at eBay (as usual...). The tiny IC got soldered to the board of course (it is rather useless otherwise):



Otherwise, this project progress is rather slow now. The board is almost complete but there are some key components still missing. These will be included in a larger order to Digikey as I also needs parts for my MiniMigs. An other factor slowing things down are cash - building this board isn't cheap and sometimes I have to wait for the salary to arrive. I also hope that the MiniMig will start to pay back a little bit by selling some of the boards I'm building.

The design of the power supply is ongoing - the schematics work is almost done. For the 12V section I will use the LT1950 in a SEPIC configuration.


2008-11-18: Received the scandoubler memory parts that Marc managed to source. Thank You for the effort - one big problem in this project solved. The last odd IC to find id the real time clock chip. Anyone that have one spare?


2008-11-17: Got a package from Elpro today. All resistor nets are in place and so are most of the internal connectors too. The progress in designing the PSU is slow but it moves forward. The +5V/10A and the -5V/250mA have their solutions, the 12V still to be done. a SEPIC design will be used as it allows you to power the thing from a +12V external power source. A 29F800 was also found at eBay and it got ordered. So, now there are only a few odd parts left: the real time clock IC, the two trimmer caps, the ZORRO-slot connector and some of the rear connectors.


2008-11-13: Not much done - the power supply design is ongoing and some parts for the board has arrived. All IC-sockets are now in place and the clock battery issue is solved. Instead for using the recommended solder-on cell, I will use a AA-standard cell instead. A holder for AA-batteries has been added to the board, making the battery replaceable without the need of a soldering iron. The parts for the scandoubler are on the way - Marc managed to source these rare memory parts needed.


Forgot to clean that board... ;-)


3.6V lithium cell in standard AA size.


2008-11-07: Things are getting together - only a few parts remains to be obtained and mounted to the board. An order for some parts (the 5-pin SIL resistor networks) has been places at Elpro that hopefully should appear soon. Other parts has arrived from Electrokit, but still there are issues with the 5.6 Mohm resistors (first, they missed it in the delivery and now, they sent 8.2 Mohm instead). One single resistor is left, preventing me to put the final IC socket in place. Some connectors are also missing, hopefully they will be available at Digikey so I can order them together with some MiniMig stuff.

The PSU design for this machine is forming - the main voltage (+5V @ 8A) will be delivered by a design around the LT1339 high power switch controller from Linear Technologies. This controller is able to handle several hundred amps without problems using external power-MOSFETs. I will go for a design capable of delivering 10A, giving me some margins. The +12V is easy obtainable by a Simple Switcher IC and the -5V is still open (no suitable IC found yet).

So, to the pictures of an almost complete PCB:












2008-11-03: Got the numbers for the original Amiga 1000 power supply, so now I can start designing one. My plans are to make a custom case for this machine (I don't have an Amiga 1000 to use) based on earlier experiences of custom case design. A design based on CNC-cut aluminum and plexi parts is in the plans. For, that I need a compact PSU for the machine.

So, there are three different voltages: +5V (8.0A), +12V (1.0A) and -5V (0.25A). The Amiga 1000 PSU is rated 80W. My idea is to use a laptop "brick" as the primary source. There are models at up to 120W today. Switching regulators will be used to regulate down the three voltages needed (one has to inverted). By this design, I will remove all high voltages (in Sweden, 230 VAC) from the machine, making the computer both children safe and 100% legal. It will also remove a heat source from the case, making the machine possible to cool entirely passive (= totally noise free). Time to start up that Protel DXP...


2008-10-28: Finally, a small package from US arrived! In there was the two SMD 74F245 (together with some other stuff) i bought via eBay. The time from unpacking to soldering to the PCB was rather short... :-). Now, time for some vacation at the African plains...




2008-10-25: Some progress - got the connectors removed from the broken A500 board I got for a few days ago. It was no problems to get them off with the help of a heat gun and some pliers (things really goes hot during the process). The smell of -87 vintage PCB cooked well done aren't nice though...


A revision three Amiga 500 board from 1987 - broken in some unknown way.


That poor board did not have any chance against the heat gun...


Part of my workshop, the bench where I assemble all my design. There is an other bench for more heavy work too with some power tools, making it possible to build chassis, cases and guitars...


Current board progress. One shipment from Electrokit has not arrived, preventing me from completing the board. When I got that, it will be time for the first power-on test without the IC's mounted (will it blow up or not... :-). I have to construct a PSU first though...


2008-10-23: Got the delivery from Digikey with all the EMI-filters (ferrites), inductors, resistors, capacitors and the 24-pin IC-sockets. Still, I'm still waiting for the shipment from Electrokit. With the ferrites in hand, I started to put them on the boards. There are a few to solder in place, and the 48 ohm and 220 ohm looks almost the same. You have to be careful to avoid mixup. Yes - I selected to go for the 220 ohm option instead of trying to find the 150 ohm originally stated by Georg. I think the DC-characteristics are more important than the inductance at 100 MHz. And with the 220 ohm part, the reduction of EMI should be improved. All these EMI suppressor parts makes me think a little bit - Georg put a good piece of work in reducing the emission of HF from the design. But he forgot the ESD protection at the ports. A set of Ceradiode ESD-protections at each port should have reduced the risk of killing the chips by inserting or removing joysticks while the machine is running.


2008-10-20: Finished off the 8 MB SRAM section of the board. It was rather tedious to solder all these IC's, so I did it in two passes (8 each day). But compared to the FPGA on the MiniMig, these SOJ parts are rather easy to solder. Just be careful at the positioning so they are straight on the pads. The last big orders for parts have also been done. One to Digikey and one to Electrokit (a Swedish parts supplier). When these arrives, I have almost everything needed for the board to be completed. Left is the scandoubler parts and the real time clock IC. As a bonus, I got the Mini-ITX MiniMig V2.0 running for the first time - still there is an issue with the video converter to be solved. But the VGA/RGB output was fine and all controls was working.






2008-10-17: Got the tantalum capacitors from Digikey, so now are all "normal" value caps in place. There will be final order for the passive parts to do to get all the missing ferrites, inductors, resistors and capacitors. Also, a old and broken Amiga 500 board has been found that will be the donor of the connectors and the crystal. Finally, the 74F245 SMD´s have been found and ordered at eBay.


2008-10-13: Board is a few steps closer to completion. I ran out of SMD resistors, so the next task is to order all these odd value resistors needed. Also some ceramic capacitors and all the EMI-filters are missing and must be found. Here are a few pictures of current progress:


The Chip-Mem section with the 4 x 512k (1M x 4) DRAM chips.


Part of the (incomplete) audio filter section.




2008-10-11: Picture update - assembly progress up to what was done yesterday. Also the Kessler Electronic delivery. After the pic's was taken, I added even more parts to the board...






2008-10-10: Got the memory chips for the Chip-Mem and the main 8 MB SRAM from Kessler Electronic during the week. Also some other IC's and the 48 MHz crystal oscillator. A scrap A500 board is on the way that I will save the connectors and the other crystal oscillator from. The board itself is ongoing - a large amount of the IC sockets are in place (several of them can be mounted without blocking the small SMD parts). I will take some pictures of the current progress during the weekend.

What is remaining in the department of hard to find parts are the special memory circuits used by the scandoubler. Also, I need to find a way to put the firmware into the Lattice ispLSI chip as the GBA1000 board is missing an ISP connector.


2008-10-06: Finally, got the pictures of the board (today, the weather was nice, sun and clear sky...), so enjoy:


Component side.


Solder side, with most of the parts in place.


Detail - just to show the quality of the fabrication.


2008-10-05: All 100nF 1206 capacitors are in place at the solder side of the board. More stock parts ordered. Otherwise, no nice photos of the board today. The weather was really bad with heavy rain, lot of wind and rather dark. Not really the best environment for taking pictures. Did some work on the Mini-ITX MiniMig too - got the FPGA and the other core parts in place without destroying something.


2008-10-03: I'm really eager to start putting this board together, but I found that the original documentation lacks some quality in the pictures to be able to use them as a build guide. So, to solve this rather important issue, I loaded up the original Eagle files from Georg and used them to create new, high quality prints of the assembly (top and bottom component print) and all schematics. I have put them up at my homepage and also right here below. They are in standard pdf, so it should not be a problem to view or/and print them.

Assembly Drawings 
Board Schematics 


2008-10-02: The board has arrived! After packing it up (Kristian did a really serious project with the packing - as a bonus, I have several meters of bubble wrap to use if I need to ship something), I just sat down to admire the design. The combination of blue solder mask and the gold plating makes the board looks like a piece of modern art. But, just admiring it will never make an Amiga out of it, so this weekend I will start putting parts on the thing. As a bonus, I will be able to evaluate the gold plating in comparison with lead free HASL. If it is easier to solder at the gold board, I will do my next run of mini-itx MiniMig's the same way.

I will start with all the small 1206 size passive components as they are easiest to do before the bigger and taller parts are put in place. As the board is a tested design, I don't need to use my normal approach to this build: build the power part first and run it before building the rest. Only a short circuit check at the power lines should be enough to do after the 1206-run.

Pictures of the board will appear during the weekend when I have the ability to use daylight for the photo shoot (hope that it will not rain). The best light for this kind of photo is a bright but overcast day.


2008-09-30: Boards are starting to reach it's buyers in US, so soon (I hope) it will find its way to Sweden too...

Until then, here are a few pictures of the parts needed to build the machine:




Here are some custom chips and all the needed programmable logic (with some spare). Also, a 3.1 Kickstart and a 50 MHz CPU/FPU-set.


Here are more custom chips (with spares), 74-logic and some misc. ICs. The extra CPU is a 40 MHz part that I bought before the 50 MHz ones shown above. It will be redundant for this project so it will be used with my VXL-30 accelerator in my old trusty Amiga 500.


Sockets - makes the build simpler and repairs and fault locating easier. Not at the pictures are the needed passive parts like capacitors and resistors but they are in the workshop, waiting for a board to be populated...

Parts Found:

Value	Package	Source
27C400	DIP40-6	Amiga-Digital
Kickstart 3.1 (V40.63)
AGNUS	PLCC84	OldSoftware.com
Fat Agnus 8375B (318069-03)
AMBER	PLCC68	OldSoftware.com
Amber (390538-03)
BAS32L	SOD-80	Electrokit
75V 200mA switch diode in SOD80C package.
CIA_8520	DIP40-6	eBay.de
Got a pair from eBay (actually, I have two complete sets, one marked with C= and one with MOS). Price below is for a single chip.
DENISE	DIP48-6	eBay.de
LSI1024	PLCC68	Private (Elektronikforumet.com)
Got this one (together with the GALs) via a friend at Elektronikforumet.com. He was in China... (price unknown as they were in a bundle of other components)
MC68030RC	PGA-128	Software Hut
XC68030RC50B
MC68881RC	PGA-68	Software Hut
MC68882RC50A
Socket PGA 128		My workshop...
Found one in my storage. Sometimes it pays saving old stuff... ;-)
Socket PGA 68		My workshop...
Also an workshop find...
VIDEO_H	SIP22	Amiga-Digital
Brand new - never used (not a pull from old board)