Hi,
Have you any photographs?
Replacement for 68008
-
- Font of All Knowledge
- Posts: 4756
- Joined: Mon Dec 20, 2010 11:40 am
- Location: Sunny Runcorn, Cheshire, UK
Re: Replacement for 68008
Not with me. I’m on a road trip in California. I’ll post pics and schematics when I get home.
Re: Replacement for 68008
All in one go without a recharge is the impressive thing.Dave wrote:This is the longest thing I have ever typed on my phone!

-
- ROM Dongle
- Posts: 45
- Joined: Thu Jan 27, 2022 6:02 pm
Re: Replacement for 68008
This is actually my initial idea, with a 68HC001Dave wrote:YesDerek_Stewart wrote:Hi,
Could the 68SEC000 be connected to the QL 68008 socket with an appropriate DIP to QFP adapter board, with the possibility of using it own un-contended ram?

I have been following your comments carefully and I am glad to hear that a good exchange of ideas has taken place. I also want to thank that this thread can be the starting point for a project to come to fruition, especially to Dave and Nasta for their high technical level. However, there are many concepts that are beyond my knowledge and experience, so I will have to study it in depth and document myself better.
Anyway, the 68020 has been mentioned here as a replacement for the 68008. I don't think this one totally fits my idea: a cheap replacement, easy to find, with almost no need for glue logic and probably with improved performance, so I think we should focus in other mentioned CPUs, such as 68HC001, 68EC000 or 68SEC000.
Regards.
-
- Font of All Knowledge
- Posts: 4756
- Joined: Mon Dec 20, 2010 11:40 am
- Location: Sunny Runcorn, Cheshire, UK
Re: Replacement for 68008
Hi Javier,javier2112 wrote:This is actually my initial idea, with a 68HC001Dave wrote:YesDerek_Stewart wrote:Hi,
Could the 68SEC000 be connected to the QL 68008 socket with an appropriate DIP to QFP adapter board, with the possibility of using it own un-contended ram?!
I have been following your comments carefully and I am glad to hear that a good exchange of ideas has taken place. I also want to thank that this thread can be the starting point for a project to come to fruition, especially to Dave and Nasta for their high technical level. However, there are many concepts that are beyond my knowledge and experience, so I will have to study it in depth and document myself better.
Anyway, the 68020 has been mentioned here as a replacement for the 68008. I don't think this one totally fits my idea: a cheap replacement, easy to find, with almost no need for glue logic and probably with improved performance, so I think we should focus in other mentioned CPUs, such as 68HC001, 68EC000 or 68SEC000.
Regards.
I am in the process of drawing a schematic for connection of a 68SEC000 to the 68008 socket, with a possible internal expansion connector.
I guess that Dave has done all this, so it is a little personal development and education for myself, I will post updates.
The point about the Super Gold Card, I thought that is a solution, but the Super / Gold Card rely on a custom Altera Chip that is obsolete, and worse that this, the PLD code seems to only work with one type of Altera chip. So that being said, maybe a good idea to start from again and produce something that everyone understands and is well documented.
Regards,
Derek
Derek
Re: Replacement for 68008
One thing which has not been mentioned is that running a faster CPU with it's private faster RAM and possibly other faster bus attached devices, requires a set of buffers/latches between the CPU bus and the QL bus. The faster signals the faster clocked CPU generates while accessing faster RAM etc should not be let out on the QL bus because it will wreak havoc on signal integrity and possibly even cause damage to some faster components due to excessive signal ringing.
Both the GC and the SGC use buffers between their on-board RAM and peripherals and the QL bus for this very reason. The buffers only drive the QL side of the bus when the GC/SGC is actually accessing stuff on the motherboard, at which point they are already slowed down to QL speed.
On the other hand, running the CPU much faster regarding clock speed but limiting the bus speed to QL standards does not require buffers (only a state machine to force 68008-like timings). That being said, a faster CPU will generate sharper edges on signals when their state changes, which is what causes ringing. Some sort of series termination may be required (low value resistors in series with the signals to reduce edge steepness and thus also ringing).
Both the GC and the SGC use buffers between their on-board RAM and peripherals and the QL bus for this very reason. The buffers only drive the QL side of the bus when the GC/SGC is actually accessing stuff on the motherboard, at which point they are already slowed down to QL speed.
On the other hand, running the CPU much faster regarding clock speed but limiting the bus speed to QL standards does not require buffers (only a state machine to force 68008-like timings). That being said, a faster CPU will generate sharper edges on signals when their state changes, which is what causes ringing. Some sort of series termination may be required (low value resistors in series with the signals to reduce edge steepness and thus also ringing).
-
- ROM Dongle
- Posts: 45
- Joined: Thu Jan 27, 2022 6:02 pm
Re: Replacement for 68008
I can share my unfinished project in Eagle Cad format for you. It doesn't use internal QL CLK signal (bad choice) nor it doesn't include expansion RAM nor glue logic, but it maybe useful for you.Derek_Stewart wrote:Hi Javier,javier2112 wrote:This is actually my initial idea, with a 68HC001Dave wrote:
Yes!
I have been following your comments carefully and I am glad to hear that a good exchange of ideas has taken place. I also want to thank that this thread can be the starting point for a project to come to fruition, especially to Dave and Nasta for their high technical level. However, there are many concepts that are beyond my knowledge and experience, so I will have to study it in depth and document myself better.
Anyway, the 68020 has been mentioned here as a replacement for the 68008. I don't think this one totally fits my idea: a cheap replacement, easy to find, with almost no need for glue logic and probably with improved performance, so I think we should focus in other mentioned CPUs, such as 68HC001, 68EC000 or 68SEC000.
Regards.
I am in the process of drawing a schematic for connection of a 68SEC000 to the 68008 socket, with a possible internal expansion connector.
I guess that Dave has done all this, so it is a little personal development and education for myself, I will post updates.
The point about the Super Gold Card, I thought that is a solution, but the Super / Gold Card rely on a custom Altera Chip that is obsolete, and worse that this, the PLD code seems to only work with one type of Altera chip. So that being said, maybe a good idea to start from again and produce something that everyone understands and is well documented.
-
- Font of All Knowledge
- Posts: 4756
- Joined: Mon Dec 20, 2010 11:40 am
- Location: Sunny Runcorn, Cheshire, UK
Re: Replacement for 68008
Hi,
I was going to use KiCad, but I can use any CAD package.
My main idea was to connect the new CPU at the same clock speed as the QL, and have it own uncontended ram and leave the video ram to the internal QL ram.
So hopefully the timings for the QL will be able to allow Microdrive and Network to operate correctly,
I was going to use KiCad, but I can use any CAD package.
My main idea was to connect the new CPU at the same clock speed as the QL, and have it own uncontended ram and leave the video ram to the internal QL ram.
So hopefully the timings for the QL will be able to allow Microdrive and Network to operate correctly,
Regards,
Derek
Derek
-
- ROM Dongle
- Posts: 45
- Joined: Thu Jan 27, 2022 6:02 pm
Re: Replacement for 68008
I've sent a PM.Derek_Stewart wrote:Hi,
I was going to use KiCad, but I can use any CAD package.
My main idea was to connect the new CPU at the same clock speed as the QL, and have it own uncontended ram and leave the video ram to the internal QL ram.
So hopefully the timings for the QL will be able to allow Microdrive and Network to operate correctly,
Javi