Apple II/II+/IIe IDE Drive Interface

The following document describes the current status of an IDE Drive Interface for the Apple II/II+/IIe.

The project is not complete.  The interface shown here allows the IDE Drive registers to be accessed from the Apple II.  The outstanding item is the completion of the required EPROM so ProDOS recognizes the card as a block device.  Work on this EPROM is currently in progress.

Features of this interface are:

1. Uses easily available off the shelf parts.  Component part numbers are shown for DigiKey (US) and Jaycar Electronics (New Zealand/Australia).
2. Uses easily available IDE 3.5" hard disks (note that this interface is not compatible at a wiring level for 2.5" IDE drives found in notebook computers).  Drives that have been used with the interface are listed below.
3. For compatibility with Apple II operating systems the interface uses 256 byte sectors so a drive will only store half it's rated capacity (standard IDE drives use 512 byte sectors).  I understand that ProDOS has a limit of 32mb/volume and two volumes per slot.  This suggests that using a drive any larger than 120mb (60mb with 256 byte sectors) is not going to be of much benefit.  Given the cost of the interface and the drives I felt this was an acceptable compromise.  It also made the driver modifications a lot simpler!
4. The interface does not use DMA for maximum performance.  The design objectives were to create a simple interface that was reliable, easy to build and did not use custom logic.   

Things still to be done:

1. I am currently working on the EPROM with the required code to act as a ProDOS block device.  The interface uses ports $C0x0..$C0x7 mapping to the 7 IDE drive registers (where x = 8+slot number for the interface card)..

Building the Interface:

1. The following GIF shows the schematic for the interface.  Click on the image for a more detailed version.  Save it to your machine and print it from your favorite graphics program to get a workable schematic.

2. Note:  There is an issue with the schematic.  The /IOW and /IOR lines at the IDE connector are reversed.  This will be corrected in the next version.
2. The drive can be built on prototype board if you are fairly confident with point to point construction (I build the prototype this way and it has performed well) . 
3. The interface with an IDE drive can be mounted internally in the Apple II but take care with loading limitations for the power supply.  I have built my interface and drive in an external box with it's own power supply.  The Apple System Bus signals are transferred to the interface via a 50 way ribbon cable.

Interface Parts List 

Power Supply Parts List (not shown in schematic)

Apple II System Bus Attachment Cable (not shown in schematic)

Notes:  

1. Schematic does not show the power supply circuit or the 0.1uf Monolithic Capacitors.

Frequently Asked Questions

How does the IDE interface tell if the Apple is talking to it?

The Apple II unlike the TRS-80 doesn't require the same level of address decoding because the Device Select and I/O Select lines are enabled for a specific slot only when the memory address range for that slot is selected ($CnXX for I/O Select and $C0nX (n=slot+8).  

This means that the unlike the IBM PC architecture where a card will have the same memory and port addresses regardless of the slot they are plugged into the memory addresses of an Apple card change based on the slot. 

Device Select (Pin 41 on the Apple bus) is connected to the IDE interface "enable" pin.

The IDE enable pin is Pin 37 on the IDE connector (/CS0) 

How does the IDE drive know which address you are trying to read? 

For the purposes of what we are doing the IDE drive has 8 registers. 

The register address is controlled by the DA0/DA1/DA2 pins on the IDE connector.  Notice how these are tied to the A0,A1,A2 lines on the Apple address bus.  The drive will only use the data on DA0/DA1/DA2 if /CS1 is low which is tied to the apple Device Select line.  

If the interface was connected to slot 7 by accessing memory addresses $C0F0..$C0F7 we can read and write the IDE drive registers and pass data via the data bus.

The following pins on the Apple bus are used:     

• Pin 18    R/W    (high =read, low=write)     
• Pin 20    I/O Strobe     
• Pin 41    Device Select   

To derive the IDE /IOR signal the following occurs.       

The following signals will exist on the Apple bus:           

• Pin 18    High         
• Pin 20    Low         
• Pin 41    Low       

Pin 18 is passed through the inverter U2A to convert from high to low.  The output of inverter U2A and Pin 20 and Pin 41 pass through U1A and if all are low then the output of U1A is high.     The output of U1A is passed through the inverter U2B and connected to IDE Pin 25 /IOR.   

To derive the IDE /IOW signal the following occurs.

The following signals will exist on the Apple bus:         

• Pin 18    Low         
• Pin 20    Low         
• Pin 41    Low     

Pin 18 and Pin 20 and Pin 41 pass through U1B and if all are low then the output of U1B is high.  The output of U1B is passed through the inverter U2C and connected to IDE Pin 23 /IOW.  

The truth table is as follows:

The short answer is 'harder than is worthwhile unless you want to rewrite the operating system (or write your own!).

The long answer is....

To implement an IDE interface using 512 byte sectors requires the use of a 16 bit data bus.  (The sectors are actually 256 16 bit words.)  Because the Apple II has only an 8 bit data bus we can create the same effect by using buffers and reading/writing high and low bytes from separate I/O memory addresses. 

The issue is not really what is technically possible but what is worth the effort from the software/driver standpoint.  

For both the TRS-80 and the Apple II the operating systems deal with 256 byte sectors.  To change this to use 512 byte sectors or treat a 512 byte sector as two logical 256 byte sectors was more work that I had time for.... to use 512 byte sectors you would probably have to rework a lot of the operating system and to use two logical sectors would require extra I/O to read the full 512 bytes, update the 256 bytes that had changed and then write back all 512 bytes.  This would slow the interface considerably.  

My objective was to get a basic interface running so I could add Hard Disk support two two machines that I had acquired.... I was running out of 5.25" floppy disks.... and the interface has to be simple and reliable.... and easily reproducable.  Certainly if you have the time and inclination you could try extending this.

What's next....

If you have any questions about the interface feel free to send me an email.  My email address is [email protected].  I also keep a close watch on the comp.sys.apple2 newsgroup and the Delphi Apple II forums under the JAQAS alias.

I am looking at the possibility of having some PCB's for the interface made if there is sufficient demand.  If you are interested in a hard disk for your Apple II and don't feel confident building the interface without a PCB or would like a built up PCB please send me an email and I will see how the numbers work out.

(C)opyright 2000, Andrew Quinn