Skip to content

Archive

Category: Commodore

d2d64_3

I needed a program to transfer C64 disks to .d64 images with a good overview of the process, but more importantly, it had to be fast and with minimial interaction to be used for reading disks in large batches. I ended up making a modified version of Nibread which i decided to call d2d64 so it would not be mixed up with the original Nibread. Nibread is part of the Nibtools utilities by Pete Rittwage at the C64 Preservation Project (http://c64preservation.com/).

d2d64 should not be used for preserving originals, it is only for making backups of your unprotected disks.

 
d2d64 has a new UI with a progressbar and colors to indicate status. It saves the disk as a .d64, defaults to 35 tracks, uses errorinfo if appliciable and will not reset/bump between reads. It also has an option for creating filenames based on the A/B-side of a disk. All you need to do is press Enter between the disksides.
 


I made a video on youtube that shows the process of transferring a disk. The whole video is 41 seconds, including initializing drive, turning disk and reading two disksides. Reading one diskside takes about 15-18 seconds. Hardware is a 1571 with serial cable (-s: SRQ) and a Zoomfloppy. OS is Windows 10.

 

There are two versions of d2d64 available. First one is based on nibtools (with SRQ support) for xum1541/Zoomfloppy users. This is probably the one you want. The second one is an older version based on mnib(predecessor to nibtools) and should be used with XMP/XAP1541 cables(LPT-connected drives). You can scroll down to the “Short history” part of this post for a brief explanation on the hardware differences. The older version I made between 2007-2010 when Zoomfloppy was not available. I decided to include it here in case some of you (like me) still have their XA/XM1541 systems running.
 

Download:


 

If you want to preserve your originals, visit these links below:
Kryoflux: http://www.kryoflux.com/
nibtools: http://c64preservation.com/nibtools

 

Short history
The Commodore drives communicate with serial communication through a DIN-6 plug cable between the drive and computer. For faster speeds, a parallel cable evolved allowing 8-bits to travel in parallel. The drive parallelcable was not a previous standard, but a cable soldered directly to the second VIA-chip on the drive and then connected to the C64 Userport.
 

The 15×1 drives don’t have a standard communication port and therefor you need a special cable to hook the drive to a PC. Early software even transferred files and images through the V.24/RS232-serial protocol.
 

Later on(1992-) came the X1541-cables which provided multiple options to connect your drive to the PC. The drive was connected to the PC through the LPT-parallelport which required exact timing to work. There were incompatibilities with some motherboards which were circumvented by different versions of the 1541-cable. The drive parallelcable was also available for the X1541 cables and added a “P” to the name.

Even later came OpenCBM, based on CBM4Linux, making it possible to communicate with drives under Windows NT/2K/XP with XA/XM1541 cables.

You can read everything about the X1541-cables at Joe Forsters homepage: here
 


My x1541 breakout box from the early days. This connected to the LPT-parallelport on a PC and disk transfering was done in DOS. Old PC-hardware did not work properly with all x1541-cables so i needed a device for testing combinations and different low level components.
 


XMP1541 in the making…
 

USB to the rescue
The LPT-Parallelport became more and more obsolete when the PC hardware evolved. Lots of tinkering with timings and settings was also required to get it working.
 
Till Harbaum worked on an USB adapter called xu1541 until he lost interest in 2007. Nate Lawson, with Wolfgang Moser and Spiro Trikaliotis, continued the project and developed it even further. In early 2009 the xum1541(pronounced “zoom”) was introduced, a full speed USB device with support for parallel transfers. Best known xum1541 implementation is the Zoomfloppy. Today OpenCBM also supports the xum1541.

 
Read more on the following links:
xum1541: http://www.root.org/~nate/c64/xum1541/
Zoomfloppy: http://www.go4retro.com/products/zoomfloppy/

 


 

A collection of disk covers that I have acquired during the years, also known as “Cover Art”. Use arrow keys or mouse scrollwheel to navigate though the covers and press enter or klick on the picture for a larger version.

Most of these covers are photocopies and some of them have had a rough life in the snailmail process. Addresses have been removed due to privacy reasons.

Thanks to Tech for a huge amount of covers.

 

2015-10: Added 183 covers.
2016-02: Added 3 covers by Microboy

 


 

 


 

    A friend of mine, Hedning/Genesis Project, sent me pictures of a 1541-clone that neither of us recognized and we asked around if anyone knew the name of the drive. Moloch contacted us with pictures and said it known in the US by the name “Commander II”. It might been stripped of the labels when imported to Euro-Land and sold as a different product. The Commander II was reviewed in Ahoy issue 16, 1985 (pages. 28-38).

     

    Update 2016-07-24: Hedning was generous and gave me one of his drives, thanks a lot! I added a few more pictures of the drive.

     
    Update: Thanks to Moloch we now have a name for this previously unknown clone. It was sold in the US by the name “Commander II”.

     


     

    The LEDs might been replaced with green/red since some traces were broken. I also removed a home-made deviceswitch due to its bad construction.

     

    The +5V is connected to the large area on this side of the PCB.

     

    The mech is a Chinon F-051.

     

     

    +5V/+12V PSU.

     


     

    Pictures by Hedning

     

     

     


    Serialports are located on the side of the drive.

     


    “Diskett” was a local Commodore computer shop/service center located in Malmö, southern sweden.
    The second drive, with serial 841633, does not have this sticker.

     

     

     

    1541clone_hp64_7

     


     

    Pictures by Moloch

     

     

     

     

FD-148 is a third party floppydrive for the Commodore 64.

 

 


 

 


Remove the screws marked with red and lift the cover.
 


You need to loosen 4 screws to remove the shielding and additional 3 screws to release the PCB.
 


Detach the white connectors and lift the main PCB. There is a ribbon cable beneath the PCB which also needs to be detached. The back plate can now be removed.
 


Remove the 3 screws marked in the picture. There are spacers below that PCB which you need to take care of.
 


Spacers marked with red.
Two voltage regulators(7805/7812) are fastened to the black heatsink to the right.

 


Flip the drive around and remove the 4 screws.
The drive mech is now loose, lift the plastic casing.
 


 


 


 


 

I have repaired a fair amount of drives, and every time I find myself loading a test or a diagnostic software from disk. This is a bit of a contradiction since the drive being tested, or repaired, might not even be able to load a program. I also have a variety of tools for different purposes and wanted to have everything in one place.

 

This is why i created a 1541 Diagnostic Cartridge.


 

The challenge was to get all these tools to fit into a 8K cart. I wanted to keep the hardware simple and a 16K cart would have overwritten the Basic interpreter.

Therefor I have optimized both code and visual on the tools to keep the size down. Some tools are old, others were written from scratch and a few I rewrote in machine code instead of basic. A couple of the tools are still in Basic, but optimized.

The cartridge is an 8K ROM at $8000-$9FFF

You can also use RESTORE to return to the menu.

You may need to send a “I0:” or turn the drive off/on before running further tests if an previous error has occurred.

 

 
Update 2017.09: The Cartridge is also available at the Protovision shop.
 
Update 2015.12: Upon request from their customers, The Shareware PLUS Commodore 64 & 128 Blog asked me if they could offer the 1541 diagnostic cartridge in their product sortiment. The 1541 Diagnostic Cartridge is now available in their eBay shop.

 

 

I had a broken 1571 on which I located the problem to the powersupply.

The PSU is relatively easy to refurbish with new caps and so forth, but I wanted to try out a switching powersupply i had.

The new powersupply is a MeanWell PT-65A and is rated 5VDC/7A and 12VDC/3,2A. It also has a -5VDC/0.7A output which was not used. This specific model was on a €4 sale at a electronics distributor and the current ratings were good, hence the selection.

 

WARNING! LETHAL VOLTAGES ARE PRESENT ON THE POWERSUPPLY.

 

 


1571 with original powersupply.

 

 


Mounting plate.

 


Testmount.

 


A very tight fit, I had to make a cut on the left side.
The PCB is mounted on spacers and the cables were crimped.

 

Finished assembly.

 

 

 

 

 

 

This post spawned out of a discussion at a facebook group http://www.facebook.com/groups/retrodatorer/. Thanks to Jonny Hylander, Fredric QJ Blåholtz and Krister Andersson for ideas and suggestions.

Update 2016-07-20 : Thanks to Marty/Radwar for sending me a huge list of different kernal checksums. The diag586220+ v0.4 is able to identify a staggering 49 different kernals, for example Professional Dos, RapiDos, DolphinDos and Speeddos. Download the v0.4 sourcecode for a full list.

Update 2016-11 : KiWi at www.SX-64.de sent me a version that now works correctly on the SX-64. Scroll down for the download and sourcecode. See his page here.

Update 2017-04 : Added a standalone ROM Checksum program with diag586220+ v0.5 tables.

 


 

The diagnostic 586220 does a checksum on the kernal ROM to verify if it is ok. It will only identify the original CBM ROMs. All other (even good, for example JiffyDos) kernals will be marked as “BAD”.

The 586220 was used since i have done an reassembly/disassembly of it earlier.

Code to calculate the checksum.

Depending on $FF80(Kernel revision) the checksum is either $E0 or $E1. This determines if Kernal is marked as “OK” or “BAD” by the diagnostic test.

I remembered that the C128 diagnostic cart displays the checksum. When running the C128 diagnostic(789010) the checksum on the C64 Kernal is reported as $D4. After investigating the code, which is identical to the one above except for the fact that only $1F00 of memory is checked. This is probably a mistake since the kernal is 8K ($2000 bytes).

We shared ideas about how this would be accomplished, but also noticed that some of the different kernals generated the same checksum. This might be deliberate for them to pass a diagnostic test.

 

All code below is compatible with 64tass (i.e. Turbo Assembler compatible).

The new checksum routine. To make a distinction, the address lowbyte is xor:ed and therefor not resulting in an identical checksum.

 

 

…I made a program that can identify the kernal ROM.

 

A table is used for the checksum and a pointer(.word) to the matching ROM name. You can easily add new checksums and ROMs to the code.

 

 


 

Incorporate the checksum with Diagnostic 586220
 
The goal is to replace the original checksum routine with the code above.

As we can see in the disassembly, the routine for ROM tests are at $890E. Further investigations reveal that code between $890E and $8A44 can be replaced. This area also has the routines to check BASIC and CHARACTER ROM.

 

Checksum routine is located a few lines below $8926.

After inserting the code, you can now assemble(compile) the code with c64tass. I won’t go into details about the code itself, but it is presented and downloadable further down on the page.

 

The output file (.o64) can now be written to an eprom. Skip the two first bytes (0x00, 0x80) which is the loadingadress of the file.
 

You can also test the cart in vice by converting it to a crt. Cartconv.exe is included with the Vice emulator.

 

586220-
586220+ diagnostic, it is now possible to identify the kernal ROM.

 

 

Download package with the sourcecode for diag586220+ and original disassembly of the 586220 diagnostic. Compare these two files to see how the implementation was made.

 

Download package for diag586220+ v0.4 by Marty / Radwar. It identifies 49 different kernals.

 

Download package for diag586220+ v0.5 by KiWi / www.SX-64.de. Fixes an issue with SX-64.

 

Standalone ROM Checksum program with diag586220+ v0.5 tables

 


 

The new checksum code.

Checksums for BASIC and CHAR are also displayed.

 

 

Information on how to build a VU-Sette.

Read more about the “VU-Sette” here: http://blog.worldofjani.com/?p=1127

 

The following articles/documentation have been used as information for this project:
64er 1985/10 p.xx (http://www.64er-online.de/download/1985/index.html)
Svenska Hemdator Nytt 1990-05 p.34 (http://www.selda.se/sub/hemdatornytt/1990_05/)
C2N/1530/1531 Service Manual Preliminary 314002-002 (1984 Oct).

 

 

The VU-Meter


The VU-Meter was pulled out from a “Precision Batterychecker mini” from ebay 😉

 

 

Prototype

An azimuth tape and Winter Games original was used for inspecting the signal.

 


…and more testing.

 

 

Circuit diagram for the VU-Sette.

I made a sketch with the information i had gathered. Connection points for “green” and “purple” can be found at the bottom of this page.

Component list:
IC1: LF356
R1 : 10 KOhm
C1 : 470uF
R2 : 62 Ohm 0,5W
Q1 : 2N3904
Speaker: 8 Ohm 0,5W

You might need to adjust the values to match the make and model of your VU-Meter/speaker. The speaker and the VU-Meter are independent of eachother. Either one can be left out if wanted.

 

Potentiometer R1 is for zero-adjusting the VU-Meter. The tape read-head signal and R1 must be adjusted to match. When the signal peaks, you want the VU-Meter be at the max position. You can use an original game, or a tape recorded with correct adjustment, to find the peak. Note that not all originals have a “good” recording signal to prevent duplicating.

The initial adjustment can be a bit tricky, but when it is done, you will only need to adjust the tape read-head when loading a tape.

I can think of some cases where you want to adjust the sensitivity further; for very bad/old tapes, or that the reference tape had a signal that was “too good” and you need to bring the level down a bit.

 

 


A Veroboard was used for the first build. I also needed to know much room there is inside the datasette for the pcb and VU-Meter. This version does not have the speaker.

 

 

PCB Layout


I ordered a few PCBs to test the concept. In retrospect, I would rather use solderpads for the speaker and for the VU-Meter. The VU-Meter would probably be able to hold the weight of the PCB and wouldn’t need any further mounting. The extra potentiometer snuck into the upper left corner can probably be left out, I wasn’t sure if there was a need for extra adjustment of the VU-Meter.

 

 

 

Assembly


A 22 mm hole for the VU-Meter was drilled 6 mm inwards from the junction.

 

 

Parts for the VU-Meter. The rightmost VU-Meter is from another batterytester i found on ebay. Both work with the VU-Sette, but the leftmost is easier to fit into the datasette.

 

 


VU-Meter with everything cluttered into the bottom left corner. 🙂

 

 


Fully assembled.

 

 

 


 

Finding the signal
The read-signal is amplified and filtered four times before leaving the datasette. Therefor the signal strength will be different depending on which stage you choose to tap the signal from. VU-Sette has been successfully implemented with the PCBs and connection points below. You might need to do some testing to find out which output stage is most suitable for your equipment. YMMW.

 

Datasette with one OP.Amp (PCB: 1530, 1531, NP-136H):


1530

1531

NP-136H

 

 

Datasette with one OP.Amp (PCB: CMR-001):


CMR-001

 

 

Datasette with two OP.Amps (PCB NP-090B):


NP-090B

 

 


 

 

 

 

The C128 Cartridge functionality is different than in the C64. Maximum size has been doubled to 32K and an EPROM socket (U36) has been provided inside the machine in addition to the cartridge port.

An external cartridge is referred as “External(Function ROM)” and the internal EPROM socket is referred as “Internal(Function ROM)”. This will focus on how to create an external cartridge.

A 8-bit machine can only “see” 64K at a time, and the same goes for the C128.

There are two areas in the C128 memory where cartridges can be located; the MID ($8000-$BFFF) and HIGH ($C000-FFFF). A cartridge can be started up in either these 16K areas or single 32K cartridge can be used. It is possible to have both an internal “cartridge” and an external cartridge installed, but the external has preference over the internal.

C128 memory is managed by a chip called the Memory Management Unit(MMU) and does not use the PLA(GAME/EXROM lines) like the C64.

MMU is controlled by its primary registers at $D500- and its secondary registers at $FF00-. The secondary registers are available to make it possible to switch out the I/O area at $D000-DFFF and still be able to access the MMU. This creates a memory “hole” at $FF00-$FF04 which can not be used for code.

Note that both cartridge areas (MID and HIGH) are always switched in together. If you set up a 16K cartridge in MID area and want to access Kernal routines, you will need to bank in Kernal at the HIGH space.

 

Cartridge Autostart
When a C128 is booted, a cartridge presence is checked. First it checks if GAME or EXROM lines are low. If this is the case, a C64 mode cartridge is connected and the machine starts in C64 mode. Otherwise the C128 memory is checked for an identifier(“CBM”) string at $8007 and $c007, this is done both for the external and internal slots. The banks are checked in the following order: External Mid, External High, Internal Mid and Internal High.

 

Code at $E27A- is used to verify if a cartridge is inserted:

 
 

16K Cartridge code example at $8000(MID) with Kernal ROM banked in.

 
Test the code above with WinVICE:

 



C128 16K Cartridge $8000-$BFFF (MID)
 

 
You can add another 27128 and connect RH to /OE on that chip to get 32K.

 

I have two Action Replay 6 clones, one i got from an auction and it had a german ROM and the other is a polish clone called “Atrax 7.3”. Both are probably Atrax clones from the beginning… I have successfully run stock AR6 ROMs in both of these.

Update: Added two more pictures on the other one