It is currently Thu Mar 28, 2024 5:25 am




 Page 2 of 3 [ 31 posts ]  Go to page Previous  1, 2, 3  Next
Author Message
 Post subject:
PostPosted: Fri May 18, 2007 3:45 pm 

Joined: Fri May 18, 2007 3:41 pm
Posts: 2
I have found that the IOR and IOW lines need to be damped with 100 ohm resistors or you will have problems with IDE drives. These two signals are edge-triggered in the drives - they need to be pretty clean. Put a 100 ohm resistor in series as close to the driver IC as you can. I use 74HCs on my interfaces.

Bob


Offline
 Profile  
 
 Post subject:
PostPosted: Thu May 24, 2007 5:24 am 
User avatar

Joined: Sun Nov 02, 2003 8:15 am
Posts: 1549
Location: netherlands
Hi Bob,

What you mean:
logic>-----100Ohm------>IOR/IOW-drive
or
logic>--+----------------->IOR/IOW-drive
100
|
ground

Cheers,
Sijmen.


Offline
 Profile  
 
 Post subject:
PostPosted: Fri May 25, 2007 6:52 am 

Joined: Sun Jun 11, 2006 4:33 pm
Posts: 82
Good question, where ?


Offline
 Profile  
 
 Post subject:
PostPosted: Thu Jun 28, 2007 11:41 am 

Joined: Fri May 18, 2007 3:41 pm
Posts: 2
I put the resistors in series. You might want to try 50 to 75 ohms, first.

ICpin >> 75 ohms >> IDE interface


Offline
 Profile  
 
 Post subject:
PostPosted: Fri Jun 29, 2007 11:10 pm 

Joined: Sun Jun 11, 2006 4:33 pm
Posts: 82
@Mr-Atari : you understand where put R75 ?


Offline
 Profile  
 
 Post subject:
PostPosted: Sun Jul 01, 2007 12:39 pm 
User avatar

Joined: Sun Nov 02, 2003 8:15 am
Posts: 1549
Location: netherlands
Like Ataribob says:

IC-logic>-----(50-100)Ohm------>IOR or IOW-drive

Cheers,
Sijmen.


Offline
 Profile  
 
 Post subject:
PostPosted: Sun Jul 08, 2007 4:28 pm 

Joined: Sun Jun 11, 2006 4:33 pm
Posts: 82
You tested ?


Offline
 Profile  
 
 Post subject:
PostPosted: Mon Aug 06, 2007 6:58 pm 

Joined: Thu Feb 05, 2004 10:34 am
Posts: 261
Location: NL
Today I've tested a lot things with my MyIDE cart made by Mr. Atari (the one with the HCT ttl's)

I found out I had troubles too on some larger CF cards.

The solution was strange. Perhaps someone here can explain.

The atari it works 100% right is the atari with the metal shielding-plate installed.

Unfortunately most of my ATARI's don't have that metal shield-case anymore. So that is why i kept the trouble.

Is this a mass/ground problem or a disturbance problem?

Thanks
Mariuss


Offline
 Profile  
 
 Post subject:
PostPosted: Tue Aug 07, 2007 4:38 am 

Joined: Mon Jan 23, 2006 10:49 am
Posts: 187
Location: Salzburg, Austria
Hi Marius!
Sysop*Marius wrote:
The atari it works 100% right is the atari with the metal shielding-plate installed.

This could just be a coincidence.

Quote:
Is this a mass/ground problem or a disturbance problem?

The most common reason is a timing problem when writing to the IDE drive / CF card. Due to poor design of the Atari (and sometimes crappy CPUs) the address/data-bus of the CPU is already invalid at the trailing edge of PHI2.

I've built an internal MyIDE interface with a CF adapter by myself (using a GAL instead of HCT/LS chips) and also ran into this problem.

The solution, at least for internal MyIDE interfaces, is very simple: just use PHI0 for generating the IOW signal (but don't change the circuit for the IOR signal, use PHI2 here). This shifts the IOW pulse a little bit earlier into time so that the address/data-bus is still valid at the end of the pulse.

Unfortunately you can't do this with an external MyIDE interface - PHI0 isn't available on the cartridge slot. Other workarounds, like shortening the IOW signal using a HCT123 or so, won't work, too, since the length of the IOW signal is already slightly below the minimum limit.

so long,

Hias


Offline
 Profile  
 
 Post subject:
PostPosted: Tue Aug 07, 2007 12:05 pm 

Joined: Thu Feb 05, 2004 10:34 am
Posts: 261
Location: NL
This is very interesting, but since I'm not a technical engineer, it is not 100% clear to me what to do.

Could you explain this a little more?

On the top of this topic I read also something about resistors. This is not 100% clear to me either.

I might need to know what wires I should connect to what pins. It is no problem to build-in the Interface.

So with your solution I am still able to use MyIDE in an Atari computer without that metal-shield inside? That would make me very glad.

Thanks for your help!
Marius


Offline
 Profile  
 
 Post subject:
PostPosted: Tue Aug 07, 2007 3:01 pm 

Joined: Mon Jan 23, 2006 10:49 am
Posts: 187
Location: Salzburg, Austria
Hi Marius!

Sysop*Marius wrote:
This is very interesting, but since I'm not a technical engineer, it is not 100% clear to me what to do.

Could you explain this a little more?

OK, in fact it's really simple. First, have a look at the original MyIDE schematics (from http://www.mr-atari.com/myide.htm). Also check that this matches your interfaces (if in doubt, ask Sijmen).

Now, cut the connection from the 74F04 pin 10 to the 74F32 pin 10. Pin 10 of the '04 must still be connected to pin 13 of the '32. If you are using a PCB like the one on the MyIDE website (below the schematics), also cut the trace between 74F32 pin 10 and 74F32 pin 13 and then solder a wire from 74F04 pin 10 to 74F32 pin 13 (the trace runs from the '04 first to 74F32 pin 10 and then to pin 13).

Then, locate the PHI0 signal inside the Atari. It's on the CPU pin 37 or, for example in an 800XL, also on the U18 (LS08) pin 6. Connect this signal to the 74F04 pin 9. Then connect 74F04 pin 8 to 74F32 pin 10 and you are done :-)

This modification does the following: it removes PHI2 from the IOW logic and replicates the same signal path (also with the inverter 74F04) for the PHI0 signal and uses this for generating IOW.

Quote:
On the top of this topic I read also something about resistors. This is not 100% clear to me either.

I'd recommend trying the PHI0 mod first. Adding resistors wasn't necessary to fix the MyIDE interface with my XLs (I tested it with some 3 or 4 harddrives and 2 CF cards).

Quote:
So with your solution I am still able to use MyIDE in an Atari computer without that metal-shield inside? That would make me very glad.

I hope so. And I also hope (and think) it was just a coincidence that it worked fine on the Atari with the metal shield.

On my main Atari 800XL the error occured only once every 30-60 minutes or so and it was quite annoying. After the modification I ran the scandisk test for several hours and all is fine since then.

so long & good luck,

Hias


Offline
 Profile  
 
 Post subject:
PostPosted: Wed Aug 08, 2007 5:10 pm 

Joined: Fri May 04, 2007 2:18 pm
Posts: 1
Hello HiassofT!

I installed internal homebrew interface (OS V3.1, chips 74F04 74F32, 74LS245) as suggested Mr. Atari on his web page. For storage I am using Transcend 1GB 80x compact flash card.

I am experiencing problems while detecting compact flash geometry with scandisk. In fact heads and sectors are read correctly (64 sectors, 16 heads), but number of cylinders vary from reading to reading and also quick scan shows some bad sectors.

The test with Verbatim 256 MB compact flash card showed even worse results (all geometry parameters were wrong).

I tried to solve the problem by doing the following:
1. I moved phi2 from CPU pin 39 to ANTIC pin 29 as suggested on http://www.b-pahl.de/atari8bit/myide/myide.html
but with no success - results were the same

2. I experimented with different chips (LS, HCT), but there was no improvement in stability.

2. I did the modification with phi0 as you suggested, but unfortunately there is no improvement (heads and sectors are more or less read correctly but cylinders not).

I would appreciate any help.

I am wondering if someone tried to demultiplex R/W signal with 74LS139 or 138? Maybe even LS245 could be eliminated when using compact flash card.

Gregor


Offline
 Profile  
 
 Post subject:
PostPosted: Wed Aug 08, 2007 6:40 pm 

Joined: Thu Feb 05, 2004 10:34 am
Posts: 261
Location: NL
@Hiassoft:

I will try this soon. By the way (offtopic) I found an old demo of you! A scrolltext in a circle "as seen on amiga" ... with a tune (Queen?) ... Nice to see such old productions from you!

@GZ

I found another very strang solution. Perhaps another coincidence, but you might want to try.

Do you have an external pc powersupply?
Does your CF card adapter has a connector in which you can connect the PC Powersupply?

If yes + yes... try this.

Don't switch the powersupply on. Keep it of.
But connect the CF adapter to the powersupply.

On my system this increases compatibility with a huge factor!!!!

I'm still thinking of some strange grounding or mass problem, but I will try hiassofT's hint.

what powersupply are you using for your Atari, and CF card adapter?
I heard it is best both get their current from the same source.

Some older atari PowerSupplies are getting too old for this. I measured a 4.75V on the powerconnector!!!! Far too low.

Try that grounding trick, and test again. I had the same issue with the detecting of heads/cylinders/sectors!
With the trick it worked immediately fine.

And that is why I searched for an atari with the metal shield installed, which still seems to work.

I hate solutions like these.

I hope hiassofts trick will do it for me!
Thanks
Marius


Offline
 Profile  
 
 Post subject:
PostPosted: Thu Aug 09, 2007 3:01 am 

Joined: Mon Jan 23, 2006 10:49 am
Posts: 187
Location: Salzburg, Austria
Sysop*Marius wrote:
I will try this soon. By the way (offtopic) I found an old demo of you! A scrolltext in a circle "as seen on amiga" ... with a tune (Queen?) ... Nice to see such old productions from you!

Oh, this demo must be a hundred years ago now :-) I remember doing it when a friend of mine bought an Amiga and showed me a nice intro. I thought that the Atari should be able to do it (almost) as well :-)

Now, ontopic again:

@Gregor:
Stick to the original ICs (LS245, F04, F32). Using LS or HCT types for the '04 and '32 won't work, they are too slow. IIRC using a HCT245 instead of the LS245 didn't work, too, but the reason wasn't too obvious to me. The timing specs are (almost) identical but, unfortunately, I don't have a 'scope to check what's really going on.

Replacing the '04/'32 logic with a simple GAL (16V8 with 10 or 15nS) is usually a good idea. With the remaining in/outputs of the GAL you can also buffer the other lines (A0-A2, CCTL) going to the IDE connector. I used a slightly larger GAL and also added code to drive an internal 8-character (debug-)display and to limit the MyIDE address space to 16 bytes (I just see that 8 bytes would have been sufficient, too). Here's the GAL source code in PALASM format:
;PALASM Design Description

;---------------------------------- Declaration Segment ------------
TITLE    MyIDE and debug display
PATTERN 
REVISION 1.0
AUTHOR   Matthias Reichl
COMPANY  HiassofT
DATE     06/04/06

CHIP  myidedis  PALCE22V10

;SIGNATURE HIAS

;---------------------------------- PIN Declarations ---------------
PIN 1           PHI2    COMBINATORIAL   ; INPUT
PIN 2           PHI0    COMBINATORIAL   ; INPUT
PIN 3           RW      COMBINATORIAL   ; INPUT
PIN 4..11       A[0..7] COMBINATORIAL   ; INPUT
;PIN 12         GND

PIN 13          /ENDISP COMBINATORIAL   ; INPUT
PIN 14          /ENIDE  COMBINATORIAL   ; INPUT
PIN 15          TRDIR   COMBINATORIAL   ; OUTPUT
PIN 16          /TREN   COMBINATORIAL   ; OUTPUT
PIN 17          /DISP0  COMBINATORIAL   ; OUTPUT
PIN 18          /DISP1  COMBINATORIAL   ; OUTPUT
PIN 19..21      ADR[0..2] COMBINATORIAL   ; OUTPUT
PIN 22          /IOR    COMBINATORIAL   ; OUTPUT
PIN 23          /IOW    COMBINATORIAL   ; OUTPUT
;PIN 24         VCC

STRING IDESEL '(ENIDE * /A[4] * /A[5] * /A[6] * /A[7])'
STRING DISPSEL '(ENDISP * /RW * A[4] * A[5] * A[6] * A[7])'

;----------------------------------- Boolean Equation Segment ------
EQUATIONS

; transceiver direction
TRDIR = /RW

; transceiver enable
TREN = IDESEL + DISPSEL

; IDE access
IOR  =  RW * PHI2 * IDESEL
IOW  = /RW * PHI0 * IDESEL

; display access
DISP0 = DISPSEL * PHI0 * /A[2];
DISP1 = DISPSEL * PHI0 * A[2];

; address output
ADR[0] =   IDESEL * A[0]
         + DISPSEL * /A[0];
ADR[1] =   IDESEL * A[1]
         + DISPSEL * /A[1];
ADR[2] =   IDESEL * A[2]
         + DISPSEL * /A[3];

Using a jumper I connect ENIDE either to CCTL or the $D1xx output of te '138 inside my 800XL (so I can switch between "internal" and "external" MyIDE). The TREN signal is connected both to the '245 and to the IDE, ADR0..2 is also connected to the IDE.

Removing the '245 is not a good idea, but buffering all other signals going from the Atari to the IDE can be helpful: Especially if you are using a long IDE cable (something longer than lets say 15-20cm) you will place a higher (capacitive) load onto the processor bus. I already have quite some expansions inside my XL and even the debug display didn't work without the '245. Since the signals from the CPU (especially the databus) is already under a quite heavy load, chances of interference increase.

So, in my case, I buffered all signals and used a short (~15cm) cable from my MyIDE PCB to the CF/IDE adapter (which I built into my XL just above the SIO connector so that I can change CF cards easily).

Concerning power supply problems: Yes, they are really an issue. Without proper power supply you'll get every kind of strange problems.

Don't forget the 100nF (ceramic) caps near the 74F04 and 74F32 chips, they are needed. It's also a good idea to place a cap near the '245. Then, have a look at the CF/IDE adapter. It should also have a 100nF plus some 10-100µF near the power connector. If not, add them.

Last thing, both GND and +5V should be properly connected. Ideally, use a single, thicker, lead from the Atari's GND to a single point on the MyIDE PCB from where you run GND wires to the other ICs. If you don't do it right, you may run into the problem of a ground bounce.

BTW: Using PHI2 from the Antic is usually a good idea, but only if you also use PHI0 for generating the IOW signal. Using PHI2 from the Antic for IOW will most certainly give you (at least intermittent) errors.

so long,

Hias


Offline
 Profile  
 
 Post subject:
PostPosted: Fri Aug 10, 2007 9:47 am 

Joined: Thu Feb 05, 2004 10:34 am
Posts: 261
Location: NL
My ready-to-plug grounding (mass) solution

Look at the picture.
The two wires are the same (ground) and connected to the outer ring of that cinch plug.

The centerpin of the cinch plug is not connected (!)

I plug this into the RF (TV OUT) plug of the atari to provide a decent MASS/GROUND to the CF adapter.

It might work for some people? It works great here.

I will try the other modifications later. For now this works great. When I unplug the mass-connector, it fails immediately. Strange eh.

I only have this problem on larger CF cards. The cards of 16, 32 and 64MB don't have this problem (all sandisk).

by the way I still think the metal shield is a good helper, because it provides an extra mass, what might help?

I like any comments.
Thanks
Marius
[/b]


There are images and/or files attached to this posting. You must be logged-in to view or download this content.


Offline
 Profile  
 
Display posts from previous:  Sort by  
 Page 2 of 3 [ 31 posts ]  Go to page Previous  1, 2, 3  Next


Who is online

Users browsing this forum: No registered users and 67 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  

cron