Terrible Fire Accelerators

Originally Posted by 8 Bit Dreams

Sure, U can get one from me for 5€ + shipping

Brilliant, thank you! I'll send you a PM.

[alenppc]

Originally Posted by plasmab

Not sure what the postage works out for you but..

https://rover.ebay.com/rover/0/0/0?m...2F172657814756

Thanks, I ordered a few from them. Hopefully they will be working (supaduper mentioned they sell lots of defective parts).

[Homer79]

Thank you, Thank you, Thank you
Ordered new CPLD's from digikey, stupidly quick delivery btw, took the
old ones off, soldered the new ones on, programmed with the pi.
TADA she boots to the kickstart screen.
I dont have a working gotek drive yet to check anything, but is there a
download of diagrom that doesnt need compiling? As I just want to be
sure everything is working.

long story short, adeleparts2010 memory ok, CPLD's bad

[Homer79]

Originally Posted by alenppc

Thanks, I ordered a few from them. Hopefully they will be working (supaduper mentioned they sell lots of defective parts).

Thats where i got mine from and its fine

[plasmab]

Originally Posted by Homer79

Thank you, Thank you, Thank you
Ordered new CPLD's from digikey, stupidly quick delivery btw, took the
old ones off, soldered the new ones on, programmed with the pi.
TADA she boots to the kickstart screen.
I dont have a working gotek drive yet to check anything, but is there a
download of diagrom that doesnt need compiling? As I just want to be
sure everything is working.

long story short, adeleparts2010 memory ok, CPLD's bad

Yay.

[supaduper]

Originally Posted by alenppc

Thanks, I ordered a few from them. Hopefully they will be working (supaduper mentioned they sell lots of defective parts).

I have not tried those CPU`s

But I can confirm I ordered 10 of these below and they ran fine, tested on my A1200 socketed 030/50 accelerator and 9 out of ten of them will run at 50mhz no problem so fill yer boots


https://www.ebay.co.uk/itm/1pcs-MC68...53.m2749.l2649

** Edit

I see someone bought 20 after the link I put here, please tell us how you go with them

[plasmab]

Think I just figured out why my new board is crashing. The buffer is preventing the “auto vector” data from rom getting to the CPU. I think I might emulate this in the ram cpld and speed up interrupts to full bus speed.

[Homer79]

In an effort to get my friend a TF530 as well, only problem is he cant solder.
I occured to me get an edge connector, solder on int2, chop the rest of
the pins off and fill the back of it with resin to prevent shorts. Making the
TF530 plug and play to anyone. I dont know if it would be worth making them
available so more people can have a TF530. Ive heard of people butchering
A500 trying to attach the TF530.

[PR77]

Originally Posted by plasmab

Think I just figured out why my new board is crashing. The buffer is preventing the “auto vector” data from rom getting to the CPU. I think I might emulate this in the ram cpld and speed up interrupts to full bus speed.

@Plasmab, how will the RAM CPLD know the address for the INT vectors? Aren't all the 680x0 interrupts vectors initially at fixed address at the start of ROM (irrespective if the overlay signal is asserted)? Does the 68030 support both vectored and non-vectored?

[plasmab]

Originally Posted by PR77

@Plasmab, how will the RAM CPLD know the address for the INT vectors? Aren't all the 680x0 interrupts vectors initially at fixed address at the start of ROM (irrespective if the overlay signal is asserted)? Does the 68030 support both vectored and non-vectored?

The Amiga emulates auto vectors by having a byte sequence at the end of rom. I was simply going to issue the same data as the rom does for those addresses. But faster.

[PR77]

Originally Posted by plasmab

The Amiga emulates auto vectors by having a byte sequence at the end of rom. I was simply going to issue the same data as the rom does for those addresses. But faster.

I just had a look at the disassembly of the exec.library and it seems that the base address of the KS ROM is "non-traditional" is the sense of having the first 0x3FF bytes reserved for IRQ vectors. I guess the engineers needed to make use of every available byte in ROM.

[alenppc]

Originally Posted by Homer79

In an effort to get my friend a TF530 as well, only problem is he cant solder.
I occured to me get an edge connector, solder on int2, chop the rest of
the pins off and fill the back of it with resin to prevent shorts. Making the
TF530 plug and play to anyone. I dont know if it would be worth making them
available so more people can have a TF530. Ive heard of people butchering
A500 trying to attach the TF530.

Yeah, also butchering their TF530s and sending them back to me broken

The whole edge connector sounds like overkill, but I suppose it could be used for hopeless cases. Is it available on places like digikey?

[plasmab]

Originally Posted by PR77

I just had a look at the disassembly of the exec.library and it seems that the base address of the KS ROM is "non-traditional" is the sense of having the first 0x3FF bytes reserved for IRQ vectors. I guess the engineers needed to make use of every available byte in ROM.

It’s nothing to do with the vectors in ram/rom. It’s to do with the last 10 bytes in the rom. When an interupt happens the cpu puts the interrupt number on A1-A3 and sets all other address bits high. Because the Amiga answers with DTACK rather than VPA the CPU reads which vector to go to from ROM.

The sequence of bytes is

00 18 00 19 00 1A 00 1B 00 1C 00 1D 00 1E 00 1F

At the very top of rom

[Homer79]

It's available from mouser, not very expensive https://www.mouser.co.uk/ProductDeta...vmHJv2sA%3d%3d

Originally Posted by alenppc

Yeah, also butchering their TF530s and sending them back to me broken

The whole edge connector sounds like overkill, but I suppose it could be used for hopeless cases. Is it available on places like digikey?

[elenachen]

Originally Posted by plasmab

It’s nothing to do with the vectors in ram/rom. It’s to do with the last 10 bytes in the rom. When an interupt happens the cpu puts the interrupt number on A1-A3 and sets all other address bits high. Because the Amiga answers with DTACK rather than VPA the CPU reads which vector to go to from ROM.

The sequence of bytes is

00 18 00 19 00 1A 00 1B 00 1C 00 1D 00 1E 00 1F

At the very top of rom

When I read something like that, I get the feeling I've got a lot of things to learn for the ATX4000 project

[plasmab]

Originally Posted by elenachen

When I read something like that, I get the feeling I've got a lot of things to learn for the ATX4000 project

Not really. Just wire up everything the way it is on the 4000 and you’re good.

[elenachen]

Originally Posted by plasmab

Not really. Just wire up everything the way it is on the 4000 and you’re good.

Well, if it would end with the first board

Modding and those thing come after the prototype. I'm scared and excited about that part

[indigolemon]

Originally Posted by alenppc

Yeah, also butchering their TF530s and sending them back to me broken

The whole edge connector sounds like overkill, but I suppose it could be used for hopeless cases. Is it available on places like digikey?

Isn't INT2 available on one of the CIA's? If people can pop the 68000 to fit the TF530, could they pop the ODD CIA into a riser that provided a header for it (it's on pin 21 I think)? Or even just a socket riser that has a flylead pre-attached?

[plasmab]

Originally Posted by elenachen

Well, if it would end with the first board

Modding and those thing come after the prototype. I'm scared and excited about that part

This is why I started with accelerators.. I want to motherboards but there is a massive learning curve

[jediknight]

plasmab any thoughts of bulding a new 1200 motherboard?? copy original , what would you charge?