Interest check: An updated MiniITX-MiniMig ( Pics!!! )

[Illuwatar]

The problem with the back cover plate is manufacturing - everything I do that you see here is made in a home environment with tools you could find in the nearby hardware store (except for the PCB manufacturing that is done by www.pcbcart.com). Mass producing sheet metal parts like these plates are out of my workshop's capabilities.

Fully populated boards are currently build by hand, so it is impossible for me to compete with companies like ACube in both production speed and cost. The first five boards will be built and tested (if I don't find more problems that affects the functionality), then I may put them out as auctions to see what people may pay. The corrected version will be available as bare PCB's for those who like to put it together by themselves.

[Illuwatar]

skurk wrote:

freqmax wrote:
Maybe you could utilise already used backplates on existing PCs ..?

Agreed. I thought that was part of the point with the ITX Minimig.

That was a long time ago this was possible. Every PC motherboard manufacturer do their own back panel layout today. Even swapping between different models of VIA Mini-ITX boards requires black plate change. So, making my own connector layout does not break any rules (if the Big Ones can do it, so can I).

TheDaddy - Thank You for the offer. I will keep it in my mind. But first, I must see if this board works...

[Illuwatar]

I assume You would like that connector to bring out more FPGA pins? I had actually an idea to add this, but I'm unsure how it could affect the performance of the board. Currently, I'm using a rather tight routing to keep the track length to a minimum between CPU, FPGA and memory. Going through a set of headers will make the traces longer and add pads to them. I left that out to be on the safe side.

But a small part of that has actually been added - all pins used for the VGA generation and the four spares are available at a 40-pin header (half of the pins are GND and a few carries power).

[Illuwatar]

Now I got what you was looking for. I don't think that is so simple by just bringing out more pins from the FPGA. It will require additional electronics to make it compatible (the MiniMig runs at 3.3 V internally for example) and to make it robust (you don't want to blow the FPGA by bad hardware or stupid usage). Also, it will not make the MiniMig mini anymore, taking away the original intention of the design.

My idea with pin headers was more to make it possible to create add-on modules for custom applications, not only for Amiga usage. But, as I said before - in this design, I have put reliability and performance as top priority without bringing the costs too high.

[Illuwatar]

Put the power section of the board together today and did the first power-up test. No "bang" or smoke did appear and all voltages measured correct. So, next step is to put that FPGA in place...

[Illuwatar]

The first prototype board is (almost) completed. All parts needed to run it is in place and it is ready for testing. Power-up test has been done and all voltages looks fine (and nothing went kaboom). A full run test with all firmware in place will be done within a few days - stay tuned...

Some teaser pic's:

[Illuwatar]

The first sign of life has been seen from the MiniMig - I have managed to boot it up with the Jakub modified core and PIC firmware. I got fooled by the firmware first, thinking that the board or the crystal parts for the PIC was busted in some way as it did not start up at all. But it seems like the firmware was compiled to operate from a bootloader - loaded the TinyBoot into the PIC and everything just started to work.

There is a strange issue with the RGB->S-Video converter that must be solved - when this is done, a new series of boards could be made (with all corrections implemented).

[Illuwatar]

The video converter issue is solved: it was an software issue. The hardware runs just fine (would be strange as it is a straight copy of the 1.0 MiniMig). It seems like the Jakub core version outputs a slightly different PAL signal compared with the original Dennis core (27-04-2008 core/firmware used) that causes the AD724 (using 4.433619 MHz crystal) chip to disable the S-Video output (and give a slow scrolling composite picture). The VGA/RGB output is fine though and my TV does not have any problems on SCART. And I really like the Jakub implementation with the on-sceen diagnostics, joystick control of menu and the added setting. So, I'm little bit disappointed by this. Hope Jakub reads this and takes a look at his code...

The NTSC core by Jakub have not been tested (yet) as I have no NTSC crystals for the AD724.

[Illuwatar]

I did the test You wanted me to do. By hardwire vsync to VCC (5V) of the AD724, I got a stable picture. But instead, all colors went wrong (black background with hardly noticeable contours of the floppy hand) and the pictures are skewed. Restoring vsync to be controlled by the FPGA, I got back the smooth vertical scroll, otherwise a clear and sharp (as sharp it can be at composite) picture with correct colors.

yaqube wrote:
Ok, I had a look at my sources and found that in not scan doubled mode Dennis' core outputs a static one on vsync output and my core outputs vsync signal. Both cores outputs csync on hsync output. The AD724's vsync and hsync inputs are XNOR'ed so in this case it may lead to missing vsync (is the screen slowly scrolling verticaly?)

If you have connected the AD724's vsync input to the FPGA's vsync output with a resistor please remove it and apply  high logic level to the vsync input of the AD724. Please tell me if it works so I will make a modified version.

To all that have mailed and PM'ed me regarding buying boards: You have not been forgotten. I just don't want to give out any unclear information and I must be sure that the board works to 100%. I have to put together a complete BOM with Digikey numbers and prices to be able to get a cost figure for the parts. The board layout must be corrected and finally, the total numbers of boards have to be counted for order. For those who wants complete, working boards, be aware that I have to put them together by hand, one by one.

[Illuwatar]

I re-ran the test with the vsync connected to logic high, and this time I got a stable picture with correct colors. The previous try was failed due to a bad connection (the vsync pin of the AD724 was floating - these tiny connections are hard to see). I haven't tested overscan, only with the boot picture, but You certainly will fix this. I'm waiting for an updated core to test - when it runs fine at my board, I will declare it completed regarding electronic functions.

I did the Mini-ITX verification too today - the hole pattern of the PCB matches perfect a Epia-board and the case it was installed in. The verification left to do is to test the headers for the serial port and the joystick ports. When that passes, I will finalize the design in Protel (some cosmetic changes) and start counting the mails/PM's. A BOM must be assembled and a new project to be created at my homepage.

[Illuwatar]

Yes - I'm working at Telefonplan. Nice that not every Amiga freak is on the other side of the Sea... :-)

[Illuwatar]

My home page is updated now. The new MiniMig got its own project page (not to mix it up with the old MiniMig). Schematics and some pictures are there and more will come as work is progressing.

[Illuwatar]

Hi again - You are not forgotten!

I have been working on the production part of this thing and can present BOM and Protel drawings at my homepage. I have managed to find Digikey part numbers for almost all parts needed. And I have an estimated cost for the thing in materials: €160 for the components and €18 for the bare PCB. These prices does _not_ include any profit, shipment, VAT or assembly and they are estimated.

But I have also ended up in a rather big problem that could (would) delay the project: The SD-card holder aren't made anymore. It seems like Alps has phased it out of their product line. And there aren't any pin-compatible replacement part. This means that I have to do a partly re-design of the PCB for both the Mini-ITX and Mini versions.

Edit: The SD-holder model number is ALPS SCDA1A0900

[Illuwatar]

The PCB is now updated with a newer ALPS SD/MMC card holder. The SCDA3A0102 type became the replacement. Still, the Mini MiniITX has to be done.

I have also been reading through all PM's and mails and assembled an "interest" list that looks like this:

skipp604
Firedawg
trip6
chiark
amigadave
IOWASURFER
pyrre
jimbo100
rja4096

Some wants two boards, some wants assembled and tested boards and some just want one. Totally estimated is about 15 PCB's wanted. The prices for the PCB shown in the entry above (€18) is calculated on a 20 board run. The part price is for a single board, single order (€160). Larger buys will make it cheaper. If someone has contacts with China for direct buy, it will be much cheaper...

[Illuwatar]

There is a 10-pin header just at the left side of the card socket that replicates all the pins for the card. It is not mounted at the pictures. The idea is to make a small break-out board that you could place anywhere in your chassis. Currently, I don't have any PCB layout for that yet - the MiniMig itself have taken a lot of time to design and test. The same idea have been applied to the LED's and buttons - they are all available at a pin header.

[Illuwatar]

Current plan: assemble the remaining three boards from this prototype run (they are fully usable even with the small glitches in the layout). One of these boards will be fitted with 10ns SRAM chips, just to test if they work in the MiniMig application. If they are OK, it could be an option. These 10ns memory was needed for these arcade game emulations someone did here. There will be some ordering to get this going...

For the production run of the final board (there are now enough with interest for at least a 25-board run), I will need payment in advance for the PCB. The cost for 25 boards are little to high for me to pay. To get a proper figure, I would like everyone that really wants a board (assembled or bare PCB) to send me a email at with your forum id, number of boards and if you want PCB only or a fully working unit (or maybe just a kit with all parts). This allows me to calculate costs for both parts and boards.