K1208 8mb Fastmem board for the A1200

[AwoLStill]

@Kipper2k
Crazy idea. Keep SPI, replace the IDE with an edge connector for an expansion card...

P.S.
8MB fastmem would do me. for Clock I can use NTP. http://aminet.net/package/util/time/NTPSync

[mpattonm]

Originally Posted by kipper2k

I can give a link here for them, please be aware that it is just a beta and there will be changes once i get a chance to update. I think i will dump the IDE portion from it as there appears to be no interest for it. this board will only work as 8MB fastmem and RTC. I do have changes planned

That would be awesome! Please do, eventually.

[alexh]

Originally Posted by AwoLStill

@Kipper2k
replace the IDE with an edge connector for an expansion card...

The only expansion that makes sense (to me) would be a 2nd clock port.

Question is would anyone want to run any clock port expansions with just a EC020@14MHz and 8MiB Fast?

[indigolemon]

Originally Posted by alexh

The only expansion that makes sense (to me) would be a 2nd clock port.

Personally I think SPI is a great idea, as it opens up a whole world of cheap peripherals - unlike the clock port which relies on specifically designed hardware. Spidi already has an off the shelf SPI mp3 decoder working with his Wicher accelerator.

If anything, it would be good if someone designed a clockport to SPI adapter so we can have a standard to work to that isn't 100% amiga specific.

[solidcore]

Originally Posted by alexh

The only expansion that makes sense (to me) would be a 2nd clock port.

Question is would anyone want to run any clock port expansions with just a EC020@14MHz and 8MiB Fast?

Yeah but anything heavy would weigh the k1208 down way too much lol. Another slot stacked perhaps should be at the PCI to a1200 end than the card itself for a 2nd board or something. regarding the k1208 itself, the SPI modules will be okay but expect supporting brackets if no trap door or resting on the trap door shut with perhaps some insulation to prop it up a bit as the k1208 already weighs enough to sag slightly.

[Daedalus]

*mumble*Too many projects...*mumble* A clockport adaptor should be pretty easy to implement - I'd love to give it a go! But I've got too many projects on the go already and not enough time for them all... Someone else has already developed an I2C host for the clockport which is a similar proposition. Key for any of these buses though is the software - there really needs to be a standard implementation (e.g. spi.device) that would mean that the MP3 decoder for the Wicher 500 would also work with the same driver on the A1200 with a clockport adaptor, and in the A2000 with a Zorro adaptor. The clockport itself missed that boat by not having a clockport.device from the get-go, and as such, drivers for peripherals are a mess.

So please, if an SPI interface is included, try to collaborate with other implementations to ensure software compatibility across the board.

Sorry for the slightly OT rant

[Daedalus]

Originally Posted by solidcore

but expect supporting brackets if no trap door or resting on the trap door shut with perhaps some insulation to prop it up a bit as the k1208 already weighs enough to sag slightly.

The A1200 trapdoor has specific moulded features that are intended for supporting trapdoor cards properly. (The same goes for the 500 and 600 too incidentally.) They should be used for any cards that aren't supported by the edge of the main part of the case.

[indigolemon]

Originally Posted by Daedalus

The A1200 trapdoor has specific moulded features that are intended for supporting trapdoor cards properly.

I think the use of the A1200 edge to PCI adapter prevents this doing the job in this case, could be wrong though.

[solidcore]

Originally Posted by Daedalus

The A1200 trapdoor has specific moulded features that are intended for supporting trapdoor cards properly. (The same goes for the 500 and 600 too incidentally.) They should be used for any cards that aren't supported by the edge of the main part of the case.

ahhh i have no trapdoor assumed it was flush , my a1200 was donated as the pics previously show no trap door therefor it sags lol. plenty of room for spi though

Originally Posted by indigolemon

I think the use of the A1200 edge to PCI adapter prevents this doing the job in this case, could be wrong though.

Perhaps freemilk or kipper2k can elaborate on that unfortunately as above no trap door to confirm.

[indigolemon]

Originally Posted by solidcore

Perhaps freemilk or kipper2k can elaborate on that unfortunately as above no trap door to confirm.

PM me your address man, pretty sure I have a spare trapdoor!

[edit] yep, just found it - has a hole cut by previous owner for a cooling fan to poke out of but it's 95% there and solid as is

[solidcore]

Originally Posted by indigolemon

PM me your address man, pretty sure I have a spare trapdoor!

[edit] yep, just found it - has a hole cut by previous owner for a cooling fan to poke out of but it's 95% there and solid as is

PM sent. Anything to make that old a1200 a little more complete haha it came to me very dirty and neglected, love to restore old poorly amigas when applicable.
She had a poorly membrane- replaced that, a missing key -put a new one in, no top case, unfortunately, need to track one down.

[plasmab]

Has nobody looked into the 2nd IDE/Gayle support in AmigaOS? The IDE port is prime for this just needs someone to investigate what address to map things to?

EDIT: There is some code talking about ide_doublers here... https://github.com/spotify/linux/blo...rs/ide/gayle.c

IIRC all you would have to do is update the code in the CPLD to follow all reads and writes to the INTENA/INTENAR registers but only assert PUNT for this register if there was an interrupt pending on the second IDE port. I have zero time to try this.

[plasmab]

If i read that code correct just put the GAYLE_REGS base address at 0xDAA000 and it will work. No masking required (make sure you dont mask DA9000 with a PUNT).

[solidcore]

Plasmab I'll solder the ide header on her and play around thanks for the link and guidance .

[plasmab]

The main limitation will be A14 is not wired to the Port. So basically you'll need to differentiate between the 1208 IDE and A1200 IDE by A12 +(A[23:16] = 8'hDA) (i think - my brain is thinking about DME arcs and flying aeroplanes so double check).

[kipper2k]

the trapdoor cover will close on the a1200 with the K1208 in, the only issue is if you use a 2032 battery instead of a 1216 battery for the RTC clock but that is simply a case of moving it slightly in the next update i do.

BTW, i found another PCB manufacturer, it is cheaper than PCBway.com. I have a board order i just submitted and will know how good their work is next week.

www.jlcpcb.com

[solidcore]

@plasmab thanks again
@kipper2k let's hope no hidden Chinese surprises with this one.

[alenppc]

Originally Posted by kipper2k

BTW, i found another PCB manufacturer, it is cheaper than PCBway.com. I have a board order i just submitted and will know how good their work is next week.

www.jlcpcb.com

I've been using them for all of my stuff for a while now.

[t3v6]

Hi, resubscribed because of this thread.
I am very interested, but have no technical engineering background at all.
My question:
Would the PCMCIA port work with the K1208 (seeing the CD32 doesn´t care about PCMCIA)?

[plasmab]

Originally Posted by t3v6

Hi, resubscribed because of this thread.
I am very interested, but have no technical engineering background at all.
My question:
Would the PCMCIA port work with the K1208 (seeing the CD32 doesn´t care about PCMCIA)?

AFAIK Kipper2k disabled the TF328 style IDE GAYLE so yes, PCMCIA should work.