Faster IEEE routines

classic Classic list List threaded Threaded
18 messages Options
Reply | Threaded
Open this post in threaded view
|

Faster IEEE routines

Ruud
Hallo allemaal,


Because I have problems with the ZoomFloppy, I decided to create my
own tool to create images of a floppy by connecting the drive
directly to a PC.

I'm using the original commands to access the drive and so far it is
looking good. But it takes about 15 minutes to create a D82 image.
That would mean about 2.5 minutes for every 160 KB = a 1541 floppy.
And that is not so fast IMHO.

So my first question is: are there any numbers available how long it
takes a ZoomFloppy to create an image of the various IEEE drives?

My second question: I was wondering on forehand if there are faster
methods available to retrieve data from an IEEE floppy so I won't
need to invent the wheel twice?

Thank you for any input!


--
   
Kind regards / Met vriendelijke groet, Ruud Baltissen
www.Baltissen.org







       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Jim Brain
On 5/24/2017 11:43 AM, [hidden email] wrote:
> Hallo allemaal,
>
>
> Because I have problems with the ZoomFloppy, I decided to create my
> own tool to create images of a floppy by connecting the drive
> directly to a PC.
I know you are just doing your thing, but this is so disheartening to
read...

> I'm using the original commands to access the drive and so far it is
> looking good. But it takes about 15 minutes to create a D82 image.
> That would mean about 2.5 minutes for every 160 KB = a 1541 floppy.
> And that is not so fast IMHO.
>
> So my first question is: are there any numbers available how long it
> takes a ZoomFloppy to create an image of the various IEEE drives?
You should search the archives. I am almost certain I timed an SFD1001
here a while back to respond to an earlier inquiry on the same topic.
>
> My second question: I was wondering on forehand if there are faster
> methods available to retrieve data from an IEEE floppy so I won't
> need to invent the wheel twice?
Write your own parallel routines using the IEEE lines as conduits for
your non IEEE protocol.

Jim

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Spiro Trikaliotis
In reply to this post by Ruud
Hello Ruud,

* On Wed, May 24, 2017 at 06:43:24PM +0200 [hidden email] wrote:
 
> Because I have problems with the ZoomFloppy, I decided to create my
> own tool to create images of a floppy by connecting the drive
> directly to a PC.

[...]

>                                                        so I won't
> need to invent the wheel twice?

somehow, I do not understand you.

You have a problem with your ZF. I send you a list of things to check in
order to find out what the problem is.

Instead of answering (and, possibly, helping us to find problems that we
have to circumvent), you create your own project and reinvent the wheel.

I have one question: What do you do if you have a power outage in your
area? Do you ask others how to build a coal-fired power plant, so you do
not have to reinvent the wheel, and then you build your own nuclear
plant?

I am speechless.

Regards,
Spiro.

--
Spiro R. Trikaliotis
http://www.trikaliotis.net/

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Stein
That's a little harsh...

Maybe his priority at this time is to archive some disks instead of spending his time troubleshooting his ZF setup?

Maybe he just prefers the challenge and satisfaction of building his own nuclear plant where he intimately knows every pipe and every cable even if it's not as good as the commercial versions, especially since he already has the basic plant in his basement?

Not to disparage the excellent ZF, but sometimes this is how things move forward...

I've certainly done the same thing myself on occasion when something frustrated me, whether the flaw lay with the item in question or with my thinking.

m


----- Original Message -----
From: "Spiro Trikaliotis" <[hidden email]>
To: <[hidden email]>
Sent: Wednesday, May 24, 2017 12:57 PM
Subject: Re: Faster IEEE routines


> Hello Ruud,
>
> * On Wed, May 24, 2017 at 06:43:24PM +0200 [hidden email] wrote:
>
>> Because I have problems with the ZoomFloppy, I decided to create my
>> own tool to create images of a floppy by connecting the drive
>> directly to a PC.
>
> [...]
>
>>                                                        so I won't
>> need to invent the wheel twice?
>
> somehow, I do not understand you.
>
> You have a problem with your ZF. I send you a list of things to check in
> order to find out what the problem is.
>
> Instead of answering (and, possibly, helping us to find problems that we
> have to circumvent), you create your own project and reinvent the wheel.
>
> I have one question: What do you do if you have a power outage in your
> area? Do you ask others how to build a coal-fired power plant, so you do
> not have to reinvent the wheel, and then you build your own nuclear
> plant?
>
> I am speechless.
>
> Regards,
> Spiro.
>
> --
> Spiro R. Trikaliotis
> http://www.trikaliotis.net/
>
>       Message was sent through the cbm-hackers mailing list

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
In reply to this post by Spiro Trikaliotis
On 5/24/17 9:57 AM, Spiro Trikaliotis wrote:
> Instead of answering (and, possibly, helping us to find problems that we
> have to circumvent), you create your own project and reinvent the wheel.

Ruud's last message does read as if his effort is a response to his recent
ZoomFloppy issues, but he's actually had his own project for a few years now:

http://www.baltissen.org/newhtm/cbmpc.htm

Regards,
Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
In reply to this post by Jim Brain
On 5/24/17 9:52 AM, Jim Brain wrote:
> Write your own parallel routines using the IEEE lines as conduits for your non
> IEEE protocol.

I suggest that any new protocol using the IEEE-488 lines should be designed to
be compatible with the 75160/75161 transceivers.  This will allow it to be
used on the CBM-II computers and also the modern PetSD-based devices.

On the PET/CBM computers, the IEEE-488 lines use MC3446 transceivers.  The
lines can be controlled independently and used in nonstandard ways.  I believe
this is true of the drives also, but I haven't checked all of them.  The
CBM-II computers use the 75160/75161.  They are more restrictive.

The SSE SoftBox is a peripheral designed for PET/CBM only.  It uses the
IEEE-488 data lines in an interesting way.  It has a mode where six of the
eight data lines are used as outputs while the remaining two are used as
inputs.  When I ported the code to CBM-II, I found that the CBM-II hardware
wasn't capable of doing this because of the different transceivers.

Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Jim Brain

How did you resolve the 6/2 versus 8/0 or 0/8 issue as part of the conversion, or did that issue force the effort to stop?


Jim

Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
On 5/24/17 2:02 PM, Jim Brain wrote:
> How did you resolve the 6/2 versus 8/0 or 0/8 issue as part of the conversion,
> or did that issue force the effort to stop?

I did get the SoftBox running with the CBM-II machines.  I came up with a hack
for the transceiver issue, which I explained here:

https://github.com/mnaberez/softbox/blob/9957d6d85172937fec3eee97fdb46429e4e959cd/cbm/cbm2.asm#L233-L309

It depends on careful timing to know when the SoftBox is in a certain state.
This is possible because the SoftBox doesn't use interrupts.  I don't like
this solution but it does allow the unmodified SoftBox to be used with the
CBM-II machines.  It would be better to modify the SoftBox so its protocol is
compatible with the 75160/75161.  I also disassembled the SoftBox BIOS so I
might make that change at some point.

Regards,
Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Jim Brain
On 5/24/2017 4:30 PM, Mike Naberezny wrote:

> On 5/24/17 2:02 PM, Jim Brain wrote:
>> How did you resolve the 6/2 versus 8/0 or 0/8 issue as part of the
>> conversion, or did that issue force the effort to stop?
>
> I did get the SoftBox running with the CBM-II machines.  I came up
> with a hack for the transceiver issue, which I explained here:
>
> https://github.com/mnaberez/softbox/blob/9957d6d85172937fec3eee97fdb46429e4e959cd/cbm/cbm2.asm#L233-L309 
>
>
> It depends on careful timing to know when the SoftBox is in a certain
> state. This is possible because the SoftBox doesn't use interrupts.  I
> don't like this solution but it does allow the unmodified SoftBox to
> be used with the CBM-II machines.  It would be better to modify the
> SoftBox so its protocol is compatible with the 75160/75161.  I also
> disassembled the SoftBox BIOS so I might make that change at some point.
Impressive.

Did you decide on whether to create a reproduction of this unit?  It
sounds like a cool thing to utilize, but I assume they are scarce.

Jim


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Francesco Messineo
On Thu, May 25, 2017 at 4:52 AM, Jim Brain <[hidden email]> wrote:

> On 5/24/2017 4:30 PM, Mike Naberezny wrote:
>>
>> On 5/24/17 2:02 PM, Jim Brain wrote:
>>>
>>> How did you resolve the 6/2 versus 8/0 or 0/8 issue as part of the
>>> conversion, or did that issue force the effort to stop?
>>
>>
>> I did get the SoftBox running with the CBM-II machines.  I came up with a
>> hack for the transceiver issue, which I explained here:
>>
>>
>> https://github.com/mnaberez/softbox/blob/9957d6d85172937fec3eee97fdb46429e4e959cd/cbm/cbm2.asm#L233-L309
>>
>> It depends on careful timing to know when the SoftBox is in a certain
>> state. This is possible because the SoftBox doesn't use interrupts.  I don't
>> like this solution but it does allow the unmodified SoftBox to be used with
>> the CBM-II machines.  It would be better to modify the SoftBox so its
>> protocol is compatible with the 75160/75161.  I also disassembled the
>> SoftBox BIOS so I might make that change at some point.
>
> Impressive.
>
> Did you decide on whether to create a reproduction of this unit?  It sounds
> like a cool thing to utilize, but I assume they are scarce.
>

if a good schematic (and a dump of the TTL prom) is available, I might
like to make such a reproduction (particularly if someone wants to
help condensing most of the glue logic into a cpld).

Frank IZ8DWF

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Ruud
In reply to this post by Ruud
Hallo Spiro and Jim,


> somehow, I do not understand you.
> ....
> I am speechless.

I am terribly sorry if I upset both of you in some way but that
wasn't meant at all. I have my CBM-PC project for many years:

    http://www.baltissen.org/newhtm/cbmpc.htm

(as Mike already pointed out)

But this is a machine specific project. So it was just a matter of
extracting the imaging piece and creating an independent tool that
can run on almost any PC and laptop with a LPT and COM port. That
took less than half an hour and it was even finished before I read
your answers.

And FYI, the used routines have already been developed more than 10
years ago. So I certainly didn't reinvent the wheel in this case.

Spiro, regarding your questions: I have a family and I just cannot
do those things I like to when ever I want. But I planned to handle
them for today. OK? :)

Really: no offence meant and none taken.


--
   
Kind regards / Met vriendelijke groet, Ruud Baltissen
www.Baltissen.org






       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Ruud
In reply to this post by Ruud
Hallo Mike,


> I suggest that any new protocol using the IEEE-488 lines should be
> designed to be compatible with the 75160/75161 transceivers. This
> will allow it to be used on the CBM-II computers and ...

I don't realy understand this. I used CBM-HD connected to CBM and
CBM-II computers without any problem. I also tried CBM-PC with the
2031LP, the only drive with 75160/75161 AFAIK: no problems as well.

IEEE-488 is a protocol and should be hardware independent. Same for
RS-232. This enables me to connect a C64 to a PC and exchange data
without the fact that a machine has to worry about what hardware the
other machine is using.


> It has a mode where six of the eight data lines are used as
> outputs while the remaining two are used as inputs.

When doing that then the SoftBox is not using the IEEE protocol
anymore because IEEE requires the 8 data lines either input or
output and not using a weird mix. And on its turn you cannot blame
the CBM-II not understanding this weird behaviour.


--
   
Kind regards / Met vriendelijke groet, Ruud Baltissen
www.Baltissen.org






       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Jim Brain
In reply to this post by Francesco Messineo
On 5/25/2017 4:20 AM, Francesco Messineo wrote:
>
> if a good schematic (and a dump of the TTL prom) is available, I might
> like to make such a reproduction (particularly if someone wants to
> help condensing most of the glue logic into a cpld).
I'd not mind converting the logic (I use opportunities like this to
continue to learn HDL.)  A good schematic would be a useful first step,
though.

Jim


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
In reply to this post by Jim Brain
On 5/24/17 7:52 PM, Jim Brain wrote:
> Did you decide on whether to create a reproduction of this unit?

I think it's a good idea but I'm not currently working on a reproduction
SoftBox.  All of the information that would be needed can be found on my
project page and the GitHub repository.

http://mikenaberezny.com/hardware/pet-cbm/sse-softbox-z80-computer/
https://github.com/mnaberez/softbox

Regards,
Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Francesco Messineo
On Thu, May 25, 2017 at 6:56 PM, Mike Naberezny <[hidden email]> wrote:

> On 5/24/17 7:52 PM, Jim Brain wrote:
>>
>> Did you decide on whether to create a reproduction of this unit?
>
>
> I think it's a good idea but I'm not currently working on a reproduction
> SoftBox.  All of the information that would be needed can be found on my
> project page and the GitHub repository.
>
> http://mikenaberezny.com/hardware/pet-cbm/sse-softbox-z80-computer/
> https://github.com/mnaberez/softbox

the schematic looks not really complete and I can't find a dump of the
74S small PROM. It might be easy to reverse engineer the missing parts
(including the PROM) but since there're actual units around, would be
easier if someone could complete the schematic and dump the PROM,
isn't it?

Frank

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
On 5/25/17 10:00 AM, Francesco Messineo wrote:
> the schematic looks not really complete and I can't find a dump of the
> 74S small PROM.

It is an original schematic, not a reverse engineered one.  It is complete but
it is unclear in one area around the middle.  That mostly affects some address
lines and the UART section.  Those lines are not difficult to work out.  A
dump of the PROM is available in the GitHub repository.  Martin
Hoffmann-Vetter worked out the equations and they can be found in the
repository as well.

Regards,
Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Francesco Messineo
On Thu, May 25, 2017 at 7:28 PM, Mike Naberezny <[hidden email]> wrote:

> On 5/25/17 10:00 AM, Francesco Messineo wrote:
>>
>> the schematic looks not really complete and I can't find a dump of the
>> 74S small PROM.
>
>
> It is an original schematic, not a reverse engineered one.  It is complete
> but it is unclear in one area around the middle.  That mostly affects some
> address lines and the UART section.  Those lines are not difficult to work
> out.  A dump of the PROM is available in the GitHub repository.  Martin
> Hoffmann-Vetter worked out the equations and they can be found in the
> repository as well.

Ok now I've found it, it's in the ls287 folder, I was only looking on
bios folder, my bad :)


Very well, it's already a good start!
Frank

       Message was sent through the cbm-hackers mailing list
Reply | Threaded
Open this post in threaded view
|

Re: Faster IEEE routines

Mike Naberezny
In reply to this post by Mike Naberezny
On 5/25/17 12:16 AM, [hidden email] wrote:
> When doing that then the SoftBox is not using the IEEE protocol
> anymore because IEEE requires the 8 data lines either input or
> output and not using a weird mix. And on its turn you cannot blame
> the CBM-II not understanding this weird behaviour.

This conversation started with you asking for a way to have faster transfers
over the IEEE-488 bus.  Jim replied with "Write your own parallel routines
using the IEEE lines as conduits for your non IEEE protocol."  I replied to
that and suggested any new protocol should take the 75160/75161 into account.
I gave the SoftBox as an example of a protocol that did not.

Regards,
Mike

--
Mike Naberezny ([hidden email]) http://6502.org


       Message was sent through the cbm-hackers mailing list