Apple IIe with HDDD and HxC

[Jeff]

Hi. The Apple2's DISK II GCR flux reversal rate pulse is 4μS. Considering no more 2 consecutive 0's can be stored with Apple2' GCR that makes maximum pulse width of 8 μS. Where did you derive 12 μs from?

The Apple II and Macintosh GCR (6-and-2 encoding) use 4us, 8us and 12us intervals.

The old Apple II GCR (5-and-3) doesn't use the 12us interval, but this format is far to be the most common.

[george]

12 μS means three consecutive zeros. Where do you see in 6-2 GCR three consecutive zeros?

Code: Select all

6-and-2 encoding table hide
6-bit value	GCR code[27][24][26][25][22]
hex	bin	bin	hex
0x00	00.0000	1001.0110	0x96
0x01	00.0001	1001.0111	0x97
0x02	00.0010	1001.1010	0x9A
0x03	00.0011	1001.1011	0x9B
0x04	00.0100	1001.1101	0x9D
0x05	00.0101	1001.1110	0x9E
0x06	00.0110	1001.1111	0x9F
0x07	00.0111	1010.0110	0xA6
0x08	00.1000	1010.0111	0xA7
0x09	00.1001	1010.1011	0xAB
0x0A	00.1010	1010.1100	0xAC
0x0B	00.1011	1010.1101	0xAD
0x0C	00.1100	1010.1110	0xAE
0x0D	00.1101	1010.1111	0xAF
0x0E	00.1110	1011.0010	0xB2
0x0F	00.1111	1011.0011	0xB3
6-bit value	GCR code[27][24][26][25][22]
hex	bin	bin	hex
0x10	01.0000	1011.0100	0xB4
0x11	01.0001	1011.0101	0xB5
0x12	01.0010	1011.0110	0xB6
0x13	01.0011	1011.0111	0xB7
0x14	01.0100	1011.1001	0xB9
0x15	01.0101	1011.1010	0xBA
0x16	01.0110	1011.1011	0xBB
0x17	01.0111	1011.1100	0xBC
0x18	01.1000	1011.1101	0xBD
0x19	01.1001	1011.1110	0xBE
0x1A	01.1010	1011.1111	0xBF
0x1B	01.1011	1100.1011	0xCB
0x1C	01.1100	1100.1101	0xCD
0x1D	01.1101	1100.1110	0xCE
0x1E	01.1110	1100.1111	0xCF
0x1F	01.1111	1101.0011	0xD3
6-bit value	GCR code[27][24][26][25][22]
hex	bin	bin	hex
0x20	10.0000	1101.0110	0xD6
0x21	10.0001	1101.0111	0xD7
0x22	10.0010	1101.1001	0xD9
0x23	10.0011	1101.1010	0xDA
0x24	10.0100	1101.1011	0xDB
0x25	10.0101	1101.1100	0xDC
0x26	10.0110	1101.1101	0xDD
0x27	10.0111	1101.1110	0xDE
0x28	10.1000	1101.1111	0xDF
0x29	10.1001	1110.0101	0xE5
0x2A	10.1010	1110.0110	0xE6
0x2B	10.1011	1110.0111	0xE7
0x2C	10.1100	1110.1001	0xE9
0x2D	10.1101	1110.1010	0xEA
0x2E	10.1110	1110.1011	0xEB
0x2F	10.1111	1110.1100	0xEC
6-bit value	GCR code[27][24][26][25][22]
hex	bin	bin	hex
0x30	11.0000	1110.1101	0xED
0x31	11.0001	1110.1110	0xEE
0x32	11.0010	1110.1111	0xEF
0x33	11.0011	1111.0010	0xF2
0x34	11.0100	1111.0011	0xF3
0x35	11.0101	1111.0100	0xF4
0x36	11.0110	1111.0101	0xF5
0x37	11.0111	1111.0110	0xF6
0x38	11.1000	1111.0111	0xF7
0x39	11.1001	1111.1001	0xF9
0x3A	11.1010	1111.1010	0xFA
0x3B	11.1011	1111.1011	0xFB
0x3C	11.1100	1111.1100	0xFC
0x3D	11.1101	1111.1101	0xFD
0x3E	11.1110	1111.1110	0xFE
0x3F	11.1111	1111.1111	0xFF

[Jeff]

See "Minimum/Maximum distance between transitions in time" on this page :

https://extrapages.de/archives/20190102 ... notes.html

FM : 4us/8us
MFM : 4us/8us
GCR C64 : 4us/12us
GCR Apple (5-and-3) : 4us/8us
GCR Apple (6-and-2) : 4us/12us

Also visible on the disks dumps :

gcr6a2.png (31.44 KiB) Viewed 5831 times

[george]

If you are right there must be a logical explanation to it. Can you provide it especially when Apple2's GCR does not allow more than 3 consecutive zeros? I am not familiar with your software at all and the screenshot is not comprehensible. Moreover it does not come form a logic analyzer or a scope...

[Jeff]

If you are right there must be a logical explanation to it. Can you provide it especially when Apple2's GCR does not allow more than 3 consecutive zeros? I am not familiar with your software at all and the screenshot is not comprehensible. Moreover it does not come form a logic analyzer or a scope...

The three horizontals lines are the timings intervals coming out from the drive. You will get the same thing from an oscilloscope if you read an GCR 6 and 2 disk.

[george]

This is a mere observation without knowledge and real understanding why this happens if this 12 μs pulse indeed appears at all and is not a bug in your software/hardware instead.

[Jeff]

This is a mere observation without knowledge and real understanding why this happens if this 12 μs pulse indeed appears at all and is not a bug in your software/hardware instead.

More reading needed maybe ? :

http://mirrors.apple2.org.za/ground.ica ... ii.nibbles

The bit time is nominally approximately 4 microseconds.

Code: Select all

           interval between pulses

                    fast wr    slow wr
data     nominal    slow rd    fast rd
-----    -------    -------    -------
11        4 us       4.7 us     3.4 us
101       8 us       9.4 us     6.8 us
1001      12 us     14.1 us    10.2 us
10001     16 us     18.8 us    13.6 us

1001 -> 2 zeros -> 12 us.

Maybe the 5-and-3 8us max interval between pulses assumption is wrong since i have not implemented this yet.
But there is no doubt with the 6-and-3 encoding at all : The interval between pulses used are 4, 8 and 12 us (4us per bit).
This is observed on the dump, and emulation is working quite well too on the real machine (without the HDDD) : https://www.youtube.com/watch?v=MaR9noqb0JE

And the explanation is quite simple : 4us per bit.

1 bit interval : 1*4us = 4us
2 bits interval : 2*4us = 8us
3 bits interval : 3*4us = 12us

I don't know what you need more to understand.

If this simple explanation isn't enough for you, as you said yourself, take an oscilloscope to see it, or ask the others dealing with the Apple GCR format :

https://www.cbmstuff.com/forum/showthre ... 336#pid336
https://www.bigmessowires.com/floppy-emu/

To come back to the HDDD A2 purpose :
The normal PC/shugart are not supporting this 12 us interval. The AGC/Automatic gain control filter is not tuned the same way and false pulses can be generated with the 12us intervals because the AGC gain is going too high. The HDDD A2 add a FM clock to reduce this max interval and make the HD PC drives compatible with the DD Apple II GCR format.

[george]

OK, thank you, I must have understood the 12μs measurement (it encompasses the leading 1 bit). The DISK II drive amp based on MC3470 also generates random data due to AGC when no data is read from the RW head.

What is the time resolution of the HxC? Do you think it could handle the a2r 125nS disk image format in order to "play" it? https://applesaucefdc.com/a2r/

[Jeff]

OK, thank you, I must have understood the 12μs measurement (it encompasses the leading 1 bit). The DISK II drive amp based on MC3470 also generates random data due to AGC when no data is read from the RW head.

What is the time resolution of the HxC? Do you think it could handle the a2r 125nS disk image format in order to "play" it? https://applesaucefdc.com/a2r/

The A2R support is planed in the hxc software. should be there quite soon.

[george]

The A2R support is planed in the hxc software. should be there quite soon.

This would be a giant step into the Apple2 world. Even the author of a2r is lacking abilities to build a "player" for his own format. It would be interesting to discover what percentage of all those already made a2r images would be accurate enough to run the imaged software.

[george]

Hi. Three months later are there any updates on this?

[Jeff]

Hi. Three months later are there any updates on this?

on what exactly ?

[george]

The A2R support is planed in the hxc software. should be there quite soon.

[Jeff]

The A2R support is planed in the hxc software. should be there quite soon.

This is in the beta version since some months.

[george]

Hi. Where can the latest beta (or any beta) that has the .a2r support be downloaded?