Wouldn't it be great if there was a Raspberry Pi on a Zorro card that could function as:
1.. Amiga RTG graphics card, like Picasso IV.
2.. Amiga Audio card
3.. Amiga NIC.
4.. USB controller
5.. SD hard drive.

Did I miss anything?

_________________
Men who have girlies in their avatars are Girliemen!

@Darth_X

How will you interface Zorro to Pi? i2c? SPI? UART?

Can you get documentation into the Broadcom chip to write AmigaOS drivers at that point?

_________________
All glory to the Hypnotoad!

will be more simple do it with an FPGA chip ;)

_________________
I love Amiga and new hope by AmigaNG
A 500 + ; CDTV; CD32;
PowerMac G5 Quad 8GB,SSD,SSHD,7800gtx,Radeon R5 230 2GB;
MacBook Pro Retina I7 2.3ghz;
#nomorea-eoninmyhome

@billt

Quote:

billt wrote: @Darth_X How will you interface Zorro to Pi? i2c? SPI? UART? Can you get documentation into the Broadcom chip to write AmigaOS drivers at that point?

Why not directly interface to the broadcom chip? With an FPGA?

_________________
Men who have girlies in their avatars are Girliemen!

@Darth_X

It would be an engineering problem.
The Raspberry Pi doesn't have enough free GPIO pins to be controlled by a Zorro slot.
Someone needs to read the schematic.

They're cheap enough that you can buy them as a stand alone product. Why would someone need it in their Amiga? Why burn out two products at the same time?

@Darth_X

Maybe its possible, but my guess is that is not going to be as fast as it should, you need to make software that translate commands from one OS to another OS.

So what you need up with is some thing like.

Amiga to Amiga Drivers to Rpi-io software for Amiga to Rpi hardware to Rpi-io software for Linux to Linux drivers to hardware.

Its lot of work writing the drivers the interface protocol for the RPI, plus its will sub optimal due all the hardware and software layers between the program and the hardware.

PCIe expection bus for old Amiga computers makes more sense, but really I think it be better if some made a MiniMig with PCIe expection onboard, plus add the Pin outs for floppy controller.

Last edited by NutsAboutAmiga on 10-Sep-2014 at 05:08 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@Chuckt

Quote:

The Raspberry Pi doesn't have enough free GPIO pins to be controlled by a Zorro slot.

It has wrong volt levels as well I belive. The lack of IO pins is not really a issue you can multiplex some pins to add more with some tri-state shift registers.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@Darth_X

Quote:

Why not directly interface to the broadcom chip? With an FPGA?

That is of course what goes in the middle between Zorro and Pi.

But let me rephrase my question...

What on the Pi are you going to connect up to the Amiga (FPGA->Zorro, whatever) The Pi board has limited connectivity for the Amiga to send controls and data into it... Only a few choices, and only serial formats as I can see.

Stuff inside the Broadcom chip is going to be all AMBA bussed stuff (APB, AHB, and/or AXI bussing) which is analogous to the classic Amiga 680x0 bus. We do not have access to that. There's no EBI bus available. If we can use the full GPIO hader as a parallel bus, understand that this would also be very slow. I once made an APB bus work over a pair of 32bit GPIO ports, and while functional, it was crazy slow. And it required almost all of those 64 pins. I don't see that many on Pi. Back that down to 16bit or even 8bit to fit Pi's header, and your going that much slower. And I'm nto sure if this has all of the GPIO capabilities the chip I worked with had.

Which brings us back to documentation. Can we have chip docs for the Pi? Or is this going to be a very slow interface into a Pi running Python to interpret what we are trying to tell it to do and then react? Sounds like a lot of latency.

_________________
All glory to the Hypnotoad!

@NutsAboutAmiga

Voltage shifters are way easier to deal with than other issues in here.

_________________
All glory to the Hypnotoad!

Rather than use a PI for this, it would make more sense to use one of the FPGA with DRAM dev boards out there and program the FPGA with a Zorro bus interface and all the other gizmos you'd want to have on the Amiga.

_________________
It's got 32 bits and it uses them all. It's overclocked, watch the bouncing ball!

@billt

Cool

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@JimS

Will work but will be unstandard, as soon as chips are out dated they are out dated, is not better support a newer bus standard and then some one just port drivers from Linux.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@billt

It doesn't have enough GPIO pins.

http://anycpu.org/forum/viewtopic.php?f=17&t=150

Quote:

billt wrote: @Darth_X Quote: Why not directly interface to the broadcom chip? With an FPGA? That is of course what goes in the middle between Zorro and Pi. But let me rephrase my question... What on the Pi are you going to connect up to the Amiga (FPGA->Zorro, whatever) The Pi board has limited connectivity for the Amiga to send controls and data into it... Only a few choices, and only serial formats as I can see. Stuff inside the Broadcom chip is going to be all AMBA bussed stuff (APB, AHB, and/or AXI bussing) which is analogous to the classic Amiga 680x0 bus. We do not have access to that. There's no EBI bus available. If we can use the full GPIO hader as a parallel bus, understand that this would also be very slow. I once made an APB bus work over a pair of 32bit GPIO ports, and while functional, it was crazy slow. And it required almost all of those 64 pins. I don't see that many on Pi. Back that down to 16bit or even 8bit to fit Pi's header, and your going that much slower. And I'm nto sure if this has all of the GPIO capabilities the chip I worked with had. Which brings us back to documentation. Can we have chip docs for the Pi? Or is this going to be a very slow interface into a Pi running Python to interpret what we are trying to tell it to do and then react? Sounds like a lot of latency.

@Chuckt

new raspberry may have

_________________
maybe i am dyslexia

An Apple a day keep bill gates away

@NutsAboutAmiga

I think DarthX, in his original post, listed just about everything you'd want to add to a classic 68K Amiga. Once those capabilities are added, I don't see any reason to add any newer chips. I don't really see the need to add even these features using a PC bus (PCI-e or whatever). You have to write software to drive the PCI bus in addition to driving the cards themself. Do it right, and you can put all of Darth's examples on a single card.

_________________
It's got 32 bits and it uses them all. It's overclocked, watch the bouncing ball!

@mcbone

Quote:

mcbone wrote: @Chuckt new raspberry may have

I posted the link for the schematic of the New Raspberry PI. It doesn't have enough pins. The new Raspberry Pi barely has enough pins for VGA:

http://hackaday.com/2014/09/10/raspberry-pi-gets-vga-dual-screen-support/

We talked about this issue on 6502.org and the lack of pins is why they won't use it.

@NutsAboutAmiga

The voltage levels are not an issue.

http://www.savagecircuits.com/content.php?85-Mixed-Voltage-Systems-Interfacing-5V-and-3-3V-Devices

@Darth_X

1. Amiga RTG Graphics Card using an RPi,

Not plausible, there is a distinct lack of connection options for this.

2. Amiga Audio Card,

Why even run it as this in the first place?
use ssh over a network connection, push the file to the RPi and command it to play it.
Then the Amiga is free of the task entirely.

3. Amiga NIC, *this* is one function which *may* be workable based on the older,
"EtherBridge" device setup on Aminet (but using the RPi instead of a custom built A2x86)

4. USB Controller,
****SPEED**** breaks this into a problem,
Are you wanting the RPi to host the devices or the Amiga? as you would be forwarding
from the RPi to the Amiga and needing a driver on each machine forwarding to the other.

5. SD Harddisk (again... access limitations, see 1 & 2)

the only way I see any of the above even being remotely functional would be to have
a hardware FPGA on the Zorro bus presenting as a custom Expansion Device set,
and Enabling a particular (each ConfigDev is one) function entry for each RPi connected.

the RPi's will be slow to respond at best so it would make more sense to push things over to the RPi as much as possible.

even in the best case... any generic Ethernet card with a Hub would be more effective and cheaper to maintain than any of this.

Buy an RPi with SD card, Install BerryBoot and a Linux,
Build a PC or Mac and install a Linux,

Install OpenSSH on the Linux with a personal login,

Install and setup GnuPG keychains, Create a Personal PGP key.

Set the "Slave RPi"(/"Slave PC") to host OpenSSH and let the Amiga login with the PGP key as mandatory.

use Ethernet to talk to the slaves from the Amiga using your Network stack of choice,

Use the /"ssh -c "Some Command Here"/" option to tell the slave what to do,
Use the "scp : :" to transfer files from one node to another within the personal network.

Do you need to develop the Zorro FPGA?, no.
Do you need to hard-wire the RPi or PC into the Amiga at all, no.

Is the RPi/PC Slave able to be given commands from the Amiga, yes.
Are you limited to 1 slave? no, How many ? ***thousands*** are connectable given time, money
and the ability to house them all.

Are you limited to an RPi as the only "slave"? with hardware=yes, with an IPv4 private network = no, you can have any device that hosts and accepts ssh commands behind a GPG key level of security to give the commands in the first place.