| Author: | Nate (guest: search) |
|---|---|---|
| Date: | Tue, Mar 18th, 2008 @ 15:49 ( . ) |
I have revisited my approach to implementing a successor to the xu1541. I am now experimenting with a prototype board that will support mnib and be a lot cheaper. The goal is to start simple with xu1541 compatibility + mnib support. Then, things like 1541 emulation with a d64 onboard can be added. No promises yet because things are very early but I should have more time to work on it in April. I just wanted to start this thread to solicit input from Pete, Womo, and others. |
 | Author: | Womo (guest: search) |
|---|---|---|
| Date: | Wed, Mar 19th, 2008 @ 17:37 ( . ) |
On 03/18/2008 @ 15:49, Nate wrote :
I am now experimenting with a prototype board that will support mnib and be a lot cheaper. woahah, hey, Nate, you already gave us some hints then and there, but this sounds very promising. Well, Spiro always wanted to add an 8K RAM with SPI interface. Although I found such a chip recently, it is not generally available. With only adding such a chip to the programming interface (MOSI, MISO and SCK), extending the existing xu1541 would be super easy. I, on the other hand, prefer the "integrated way", selecting a microcontroller that is fast enough to do the parallel burst protocol without buffering. The relatively new Atmel AT90USBxxxx chips would be fast enough and more or less compatible to existing firmware source. With Atmel's USBKey there is a cheap development board available... There are more solutions thinkable and I'm curious what your's is looking like. Don't you want to post some pictures? |
| Author: | Nate (guest: search) |
|---|---|---|
| Date: | Thu, Mar 20th, 2008 @ 02:58 ( . ) |
On 03/19/2008 @ 17:37, Womo wrote :
On 03/18/2008 @ 15:49, Nate wrote :
: I am now experimenting with a prototype board that will support mnib and be a lot cheaper. : : : woahah, hey, Nate, : : you already gave us some hints then and there, but this sounds very promising. Well, Spiro always wanted to add an 8K RAM with SPI interface. Although I found such a chip recently, it is not generally available. With only adding such a chip to the programming interface (MOSI, MISO and SCK), extending the existing xu1541 would be super easy. I don't like SPI since it's slow and an external, expensive part. : I, on the other hand, prefer the "integrated way", selecting a microcontroller that is fast enough to do the parallel burst protocol without buffering. Yes, that is my approach. : The relatively new Atmel AT90USBxxxx chips would be fast enough and more or less compatible to existing firmware source. With Atmel's USBKey there is a cheap development board available... Yes, I'm using an AT90USBxxx and the MyUSB library. It provides a nice wrapper around the built-in "DMA" for USB transfers. We only need to hit 40 KB/s (5000 xfers/s) to match a 300 RPM drive. Should be no problem. The only question is how small we can break the transfers down so the cheapest part will work. I will get it operational with the highest performance part and then see how to cost reduce. : There are more solutions thinkable and I'm curious what your's is looking like. Don't you want to post some pictures? I'm using the AT90USBKEY demo board for prototyping. Some things have come up with work so I won't be able to do more on this until April. But I wanted to let you two know my direction. |
 | Author: | Pete Rittwage (registered user: 558 posts ) |
|---|---|---|
| Date: | Wed, Mar 19th, 2008 @ 21:37 ( . ) |
I, for one, salute our upcoming USB-compatible overlords. No, really. It would be great to be able to put the crazy parallel compatibility problems behind us. That's most of the questions I get about MNIB these days. |