hard disc in a cpc 6128

[FRAGKI-2012]

hello to all and happy new year !!!
I m looking to create a simple ide connector for my 6128 !!!
can anyone  help me ?

[TFM]

Try to use one of the existing solutions. So you can stay compatible to software running on it.


The 8255IDE from Jarek is the fastest btw.

[FRAGKI-2012]

yes but I do not have any schematic to download as in Yarek's site there is nothing that you can download and he does not answer in the e-mail.
thank you for your reply

[MacDeath]

thses days you may skip or forget a true HDD... a better solution should be something which emulate an HDD with an SD/USB flash card...


Those can easily do 8 Gigas minimum which is gigormous ofor a CPC.
And they are easy to swap, change or connect to a Modern PC, and need close to no power supply.


An IDE HDD would be big and heavy and clumbersome.

[Bryce]

As always, the hardware won't be the issue with a HDD, whether it's a real HD, CF, SD or USB. The problem is still going to be making a decent set of documented drivers to make it usable.

Bryce.

[MacDeath]

Future Soft/TFM is always willing to do that for all the most obscure hardwares, why would it change ?

[Bryce]

For real integration, I would want a ROM based firmware that allowed full Dir / Load / Save / Directory manipulation and FAT management (and fully documented). I don't think this exists yet. Any solutions I've seen so far were a long way away from this. Or is there something out there I'm not aware of?

Bryce.

[Munchausen]

TFM - is Jarek's solution faster than the symbiface?

[rpalmer]

Munchausen,Jareks interface or any that uses an 8255 is not going to be faster than the other (like mine).
The symbiface version uses a CPLD to communicate with the attached device, but data to the interface is still limited by the CPU and the VGA interaction.
In terms of the real world speed of data transfer of a single sector is going to be a lot faster than what a floppy will do, so for a CPC it would handle files a lot quicker.

rpalmer

[TFM]

TFM - is Jarek's solution faster than the symbiface?

Yes, of course it is! While the SF2 and RPalmers solution transfers data via one I/O port the 8255IDE from Yarek is using four I/O ports, which is a genius solution. 


The SF has a maximum data transfer rate of about 162,7 kb/second. While the 8255IDE makes 186 KB/second. These values are for the drivers I made, so other drivers may be slower. But you won't get it more quick.


@RPalmer and other who doubt my words: Please read the hardware description of the mentioned products in the CPCWiki. I explained how my drives work couple of times somewhere in the forum. Further I own all that hardware and tested it intensively on a real CPC.

[HAL6128]

For real integration, I would want a ROM based firmware that allowed full Dir / Load / Save / Directory manipulation and FAT management (and fully documented). I don't think this exists yet. Any solutions I've seen so far were a long way away from this. Or is there something out there I'm not aware of?

Bryce.

...would be BonnyDOS a solution? It's ROM based, documented, but not a FAT type like MS is using.

[TFM]

Yarek did provide a patched Amsdos ROM which supports his 8255IDE interface (which is connectors for hard discs and SD cards). But I haven't had the time to check that all in detail.

[rpalmer]

TFM,

The 8225 version of an IDE interface uses one of the ports to provide control lines to the attached IDE device which means actual transfers are manually controlled by the CPC. This means that all transfers are going to be slower than other interface types where a CPLD or PAL/GAL manages the control lines of the IDE interface.

I am not doubting the merits of speed between Yarek and the SF, but for those interfaces which manually control the device the transfers will ALWAYS be slower.

Also, the 186K/s is a figure based on the routines theoretical maximum and not its real max.
If TFM doubts this, goto "http://8bit.yarek.pl/interface/yamod.ide8255/ide8255.txt"; which details the max transfer rates for Yareks IDE controller

rpalmer

[steve]

I get a "not found" message, would you check the address please.

[Bryce]

...would be BonnyDOS a solution? It's ROM based, documented, but not a FAT type like MS is using.

Using a custom format instead of FAT32 is a major drawback, especially if the device is using a CF card which could otherwise easily be managed and populated on a PC.

@Steve: Remove the inverted comma after txt and the link works.

Bryce.

[TFM]

TFM,

The 8225 version of an IDE interface uses one of the ports to provide control lines to the attached IDE device which means actual transfers are manually controlled by the CPC. This means that all transfers are going to be slower than other interface types where a CPLD or PAL/GAL manages the control lines of the IDE interface.

I am not doubting the merits of speed between Yarek and the SF, but for those interfaces which manually control the device the transfers will ALWAYS be slower.

Also, the 186K/s is a figure based on the routines theoretical maximum and not its real max.
If TFM doubts this, goto "http://8bit.yarek.pl/interface/yamod.ide8255/ide8255.txt";;;;; which details the max transfer rates for Yareks IDE controller

rpalmer

Oh it's really great that you tell me about a piece of hardware which actually really works with 186 kb/s in my hands.


It uses four ports an not one!!! How often do I have to say that.


Routine (working and tested) looks like...


...
INI
INI
INI
INI
LD B,E
...
...
This block will be repeated 128 times to get the maximum speed of 186 KB/Second. AS TOLD: TESTED!!!

The hard-disc is so much faster than the CPC, that the CPC will NOT wait for it! And it has a buffer! So what you call manually controlled really does not matter. Not even 0.1%




And btw: Don't bitch at others hardware. Get your own card running!

[TFM]

Using a custom format instead of FAT32 is a major drawback

Exactly.

[Munchausen]

Well it sounds like there isn't a great difference in speed anyhow. IDE in symbos does seem very fast with my symbiface.

[MacDeath]

FAT32

then a slim16 ?

[rpalmer]

TFM,

I was not "bitching" about anyones hardware - if came across as such then accept my apology - I am happy to say the 8255 IDE I created is not as fast as those that have a dedicated chip to control the IDE device read/write lines. Also my IDE 8255 does indeed work!.

I was only saying/inferring that an interface which uses 1 of the 3 ports on the 8255 to provide control lines to an IDE device cannot transfer data to the IDE device as fast as those which have the dedicated chip (i.e SF with CPLD or using a GAL/PAL) to provide the read/write control lines independent of the 8255.

rpalmer

[Bryce]

@rpalmer: What software are you using to control the HDD with your device. Is it an existing OS or have you written your own drivers / modified existing drivers? What commands are available? And is the software available to download anywhere? Will it also work with a CF card?

Bryce.

[Munchausen]

rpalmer, I can't find info on HDOS anywhere, but I've read that you made a driver for the symbiface (untested), FAT16/32 and HDOS, and also that you sell 8255 IDE expansions. Can you tell me more about this stuff? It would be great if my symbiface could work with IDE and FAT in DOS!

[rpalmer]

Munchausen,

I have indeed produced drivers and a DOS to handle the devices attached.

The attached devices I tested with were both old and new IDE drives as well as some CF cards. I would also like to note that I don't have a symbiface and I was supplied with its drivers from prodatron to which I modified to suit HDOS (see drivers source).

The DOS enhancement allows for the use of FAT16/32 devices as used by modern computers.

You can find the current released version of the documentation attached (also included are the drivers & source).

rpalmer

[Bryce]

Am I right in thinking that this is just the driver ROM? Where can I get the main HDOS ROM?

Bryce.

[rpalmer]

Bryce,

I have the current release version available as a ROM.

rpalmer