[an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive]
Home The Hardware Versions Documents Protected Area
 
[an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive]
Overview Picture gallery ROMs, tools Electronics

Hardware informations about the Floppy-Speeder Professional-DOS, Version 1

This section mainly describes the hardware of the Floppy-Speeder Professional-DOS from Mikrotronic. Since it didn't work, when it arrived, the whole hardware was analyzed systematically.
Firmware
In a first step the EPROM of the drive expansion PCB (27128, 16KB) was read out and some interesting strings were found:
The drive identifies itself as: 0x05b7: SPECIAL DOS  1541
On another location of the additional Professional-DOS ROM expansion area, this was found:
0x35d0: PROFES- SIONAL  1541 DOS
0x35f0: WAS MADEWITH THEGENEROUSHELP OF:
0x3610:         UWE        STAHL
0x3630: FRANK     THOMAS        MATTHIAS
0x3650:   KEGEL-    MANN        AFTER AN
0x3670: IDEA OF         KLAUS H. ROREGER
0x3690:         IN 1985.........SEND US
0x36b0: YOUR IDEAS FOR  IMPROVAL........
0x36d0: USE @D- TO MAKE USE OF  RAM A000
0x36f0: TO  BEFF........
These contents lead to the asumption, that this piece of hardware would probably be one of the very first revisions of Professional-DOS or perhaps a prototype of it.
⇒ Find more ROM informations at the ROM and tools section of this area.
Hardware
After retrieving the Professional-DOS hardware, the first task was to documentate the hardware by some high quality scan pictures. This helped reassembling the board, after all work was done.
The following list describes the steps, that were done to analyze and repair the board:
  • All socketed parts have been removed and the complete board was measured out pin by pin (264 pins with a simple multimeter). This way a complete crossconnection table could be created that ended up in a schematic of the hardware. It was a little bit difficult to decide, if there's a connection between some of the pins, because most of the TTL-IC's haven't soldered out yet. Sometimes a resistance became nearly 0 Ω without a direct connection between these two points. This way the fault with the board could be found, because one of the pins of the processor (address line A12) wasn't connected to any other, but logically had to.
    This procedure was a lot of work, nearly 14 long and boring hours. Although the fault was found, the rest of the planned procedure was done nevertheless.
  • All parts have been soldered out, so that some scans of the empty PCB and the bottom side of the parts could be made. As you may know, normally the labels of all the TTL-ICs of Professional-DOS are removed to prevent users from copying the hardware design (as happened with several other speeder systems). Sometimes IC-producers print the type number onto the bottom side of the chips and so it was with one of the smaller ICs. This chip could be identified as a 82S123, which is a 256 Bit PROM (32x8 Bits) also known as 74S188. It seems, that it is used as some sort of programmable logic (5 inputs, A0...A4 and 8 outputs, O1...O8), like a PAL or GAL.
  • All the circuit paths of the PCB have been measured out again. This way the crossconnection table could be checked again and some bugs were removed (wrong numbering because of the many ill oriented parts). This check gave the assurance, that there were not any more faults with the board than the one found earlier.
  • The whole hardware has been reassembled then, but using high precision sockets all the way. This way each part and the whole board can be analyzed again without the need of desoldering the board anymore.
Now the drive board itself was working again. Professional-DOS didn't reach the proposed speed yet, but that may be caused by the fact, that the desired C64 kernal ROM replacement was missing.
⇒ View the documentation steps of disassembling the hardware at the picture gallery of this area.
Electronics
Because another Professional-DOS user asked for help to repair his hardware a deeper look into the hardware was needed. »Someone needs to understand a thing, before he can repair it«. First all of the unknown IC part numbers needed to be unveiled.
  • Two other TTL-ICs were identified, they are both the only 14-pin ICs on the board. The first was simple, because a picture of Professional-DOS in the german C64 magazine 64'er showed a 74LS00. The second one was more interesting. By simply spitting onto the top side of the chip (iiieeeehhh :-) some rests of the labeling could identify it as a 74LS74.
  • After playing around with a picture manipulation program to prepare the scans for this web site another one of the removed labels of one of the TTL-ICs could be identified. In combination with the crossconnection table, it would be safe to say that this chip is a 74LS173.
  • In the meantime all of the unlabelled ICs could be identified as 82S123/74S188, 74LS123, 74LS00, 74LS157, 74LS367, 74LS74 and 74LS173. Each part was tested with a prototyping socket board, some switches, bus drivers (74LS245) and LEDs. An ASCII chart with all known TTL-ICs was very helpful for finding the correct IC types.
    Also the contents of the 82S123-PROM were read out manually.
After these steps some schematics were created, who brought a higher level of understanding of the inner working of Professional-DOS. There's not only the clock frequency doubling system, but some circuit, that is able to automatically divide GCR nibbles, buffer one of the nibbles for one work cycle and decode them via the main ROM without the need for extra 6502 code.
⇒ The results of these steps are collected on the Electronics section of this area.
Reassembly
Later a friend found a little tool named »PROFESIONAL 1541« in his C64 collection. It creates Professional-DOS kernal ROM replacements out of SpeedDOS (+) ROMs. After building such a C64 kernal and flashing it into a new ROM, the Professional-DOS speeder system reached it's full transfer speed. 202 Blocks could be loaded in less than 5 seconds.
⇒ Get it from the ROM and tools section.

 

Wolfgang Moser, 2003-05-25,  ,  ("contact me" form disabled due to too high demand, sorry)