forum.defence-force.org

Wouldn't it be simpler to recreate everything with a Raspberry PI connected with an Arduino Uno for managing the lines to the Oric expansion port ?

Then we could do everything with parts that are both easy to obtain and can be programmed easily as well.

I saw something like this for replacing an HxC and I suppose the RPI is very capable of emulating the Microdisc controller and the disk drive. Another advantage would be that we could also emulate Disk II controllers for 8D DOS and Jasmin controllers for FT-DOS (or whatever is called the Jasmin DOS).

Here are some links of related stuff:
http://www.lemon64.com/forum/viewtopic.php?t=48771
http://hackaday.com/2013/11/26/raspberr ... ppy-drive/
http://www.raspberrypi.org/forums/viewt ... 41&t=34143
http://virtualfloppy.blogspot.fr/2013/0 ... oject.html

Hialmar wrote:Wouldn't it be simpler to recreate everything with a Raspberry PI connected with an Arduino Uno for managing the lines to the Oric expansion port ?

Then we could do everything with parts that are both easy to obtain and can be programmed easily as well.

I saw something like this for replacing an HxC and I suppose the RPI is very capable of emulating the Microdisc controller and the disk drive. Another advantage would be that we could also emulate Disk II controllers for 8D DOS and Jasmin controllers for FT-DOS (or whatever is called the Jasmin DOS).

Here are some links of related stuff:
http://www.lemon64.com/forum/viewtopic.php?t=48771
http://hackaday.com/2013/11/26/raspberr ... ppy-drive/
http://www.raspberrypi.org/forums/viewt ... 41&t=34143
http://virtualfloppy.blogspot.fr/2013/0 ... oject.html

Using two board is simpler for you?
And no it won't be that simple. Even if the 6502 is not running at 400Mhz, there are really tight timing you need to follow.
Retromaster may have done some strange choice on the cumulus, but the CPLD/FPGA is the only way to make is properly.
This is not a matter of horsepower. a rPI is not made for electronic hacking, it's just a media center oriented stuff, and it is really not well suited for beeing embedded. The high level current it need plus the really long boot time is not something that make it really interesting.

And a rPI is about 40€, an arduino is about 20 ro 30€, you are near the current prise of the cumulus. And metadata as he do low volume may not have the best price we can have for such a design.

IMHO, it will be NOT simpler.
All links are for floppy emulators only and in this part Cumulus works well.
The problems are related to the connection with the system bus and
I think this can be fixed even without changes in current Cumulus hardware.
Last but not least - many people are happy with Cumulus -
it's already 'important factor' in the Oric world and can't be easily replaced.

as Godzil and iss says: it is quite easy to replace a floppy drive with a Raspberry Pi. Replacing a controller + a floppy drive is a different task. I think you will find that the Raspberry Pi have too few I/O lines to do this. And the timing of all the signals involved will be hard to get working.

I just wanna pitch in and say as a happy Cumulus user, I'd just like to see more Cumulus' out there being used instead of hacking on rPi stuff. As soon as I turn on an rPi, I don't wanna play with the Oric any more .. but as soon as I turn on the Cumulus+Oric, I wanna play all day! Having a smooth, embedded, functional disk drive on the Oric (at last!) is a real dream. Just get a Cumulus and move on!

There's been many attempts at providing mass storage to the Oric, many many "semi working" prototypes of IDE hard drive interfaces, Compact Flash readers, etc...
As others have said, the problem is never the device, it's the reliability.

As far as I can see, Cumulus is working on about 80% of the machines out of the box, which by Oric standard is quite an achievement.

The more solutions are available, the better, specially if they are all compatible with the existing software.

Typically, for the average user Cumulus is much better than the HxC: It accepts DSK files directly, no file format conversion. It's smooth as running on emulators, except it's the real stuff

If you can provide something as cheap (relatively speaking), with as many (or more) features, and at least as reliable, go ahead, I'm not going to stop you from doing it

Good luck!

Ok ok it's a bad idea.

My point was to try and propose something.

The problem with the Cumulus is its availability.
Apparently we lost metadata so, well, I think I will buy an HxC and connect it to my Cumana controller.

The Cumulus is a very impressive device. I'm testing one right now and would be very pleased to have one myself but it doesn't seem to be happening in any near future so well...

Hialmar wrote:Ok ok it's a bad idea.

My point was to try and propose something.

The problem with the Cumulus is its availability.
Apparently we lost metadata so, well, I think I will buy an HxC and connect it to my Cumana controller.

The Cumulus is a very impressive device. I'm testing one right now and would be very pleased to have one myself but it doesn't seem to be happening in any near future so well...

Nobody is saying it is a bad idea.
The idea actually make sense, but if it does not go through, it's just wasted time and one more good idea going nowhere.

Regarding Cumulus, everything is documented and "open source". Metadata only took what Retromaster did, and managed to get it done, so technically you can build your own, you can improve the design and add features, etc...

Cumulus has many issues, but it has one advantage on everything else: It's mostly working for most people

I've done a rough approximation, for building 100 pieces it would cost around $50-60 per card for the main cumulus PCB (+ mounted components of course)
I suspect the Cumulubus won't cost more than $10 for pcb + parts

Of course without VAT.

BUT for this we need to pay for the complete batch, aka 100 * ($60 + $10) = ~$7000

100 is the minimum I can get with the manufacturer I usually use. I can look with some friends what solution they use for lower volume, but the cost may be higher.

Also, doing such a mass production (not really massive TBH) we should solve all the cumulus problem before, and the screen may need to be changed to a more easy part to source (but maybe a bit more expensive..)

I hope he is just busy, not that we lost him....

As a side note you can always get one of the Microdisc compatible controllers. You need a floppy drive or an HxC to work with it, but much better than nothing.

I have one from Silicebit and it works like a charm...

I was wondering wether metadata would have the motivation to build any more cumulus's (cumuli?) it looked like hard work assembling them all by hand, testing them etc. So I'm not surprised we have "lost" him and I am grateful for the work he has done. It would be nice to have them mass produced (well, you know what I mean) and I think there was sufficient interest shown to be able to support a modest run (even at a higher cost). The problem is getting them made in the first place and that means everyone who wants one stumping up the cash beforehand...maybe not enough people would be prepared to do this to reach the 7K mark?

If I needed to pay up front to enable a production run of a revised Cumulus to take place I would be prepared to do that. Not sure how everyone else feels about that though.

I'm willing to pay up to a 100 € for this.

I would pitch in but at the moment I'm a little more excited about the idea of building a new Oric from scratch with new PCB's .. if that happens, then maybe what we should consider is *adding* a Cumulus-like setup to the new PCB's so that all future Oric's just come with an SD card slot.

ibisum wrote:I would pitch in but at the moment I'm a little more excited about the idea of building a new Oric from scratch with new PCB's .. if that happens, then maybe what we should consider is *adding* a Cumulus-like setup to the new PCB's so that all future Oric's just come with an SD card slot.

+1