New motherboard for the Commodore 1541 floppy drive, Part II

Introduction

Some time ago I had already talked about my idea of a new system board for the 1541 from Commodore. As I announced at that time, this was not my actual goal, but only an intermediate step.

Today I want to report about the next “milestone”, and how the state of the current board is.

Things have been happening again and for some time now I have been testing a new prototype.

Flashback

As I already wrote here, my first step was to completely rebuild the controller board of the 1541, but as a 1:1 replica. I succeeded without any problems and the board has been working in one of my floppies for years without any problems.

What started out as a crazy idea, then developed more and more into a small evolution.

In the meantime I had not only created a replacement for the gate array, but also a switchable solution for IEEE-488 and the CBM bus (“IEC bus”) with the TriMod adapter. Furthermore I had also developed replacements for the two linear regulators, which become just handwarm with lower power dissipation .

So why not include all these things?

Changes

Almost all of the TTL technology I have now moved to a CPLD from Xilinx. This saved an immense amount of space.

Then, as already mentioned, I replaced the two voltage regulators with two switching regulators. I have designed these for the double line, plus reserve. Price-wise, this made no difference, but as a result, they get just hand-warm when the floppy drive is running.

So no significant waste heat is produced, which of course also saves energy.

Of course, I followed up the question about the general problem of clocked switching regulators and the associated problems due to the ripple.

I was all the more surprised when I made measurements on original boards. Unfortunately, in my opinion, Commodore has committed a design error here, as the two ground rails for 12V and 5V only at a single point (near the rectifier) together has led.

This then leads to ground loops, which drives the ripple extremely upwards. And the values are so high that even the cheapest switching regulators from China no longer affect this ripple.

Nevertheless, I had taken care during the development of my power supply to keep the ripple extremely low. And this I also succeeded in doing.

And since I couldn’t get around SMD components anyway, I started to change the whole board to SMD, where it was possible.

The next generation of the 1541

And so the new system board was created, which received many changes and some new features. And I think it can be seen:

Next Generation Mainboard

Features list

  • Interfaces
    • CBM bus in the form of the two round DIN sockets
    • Connection for parallel cable (for SpeedDOS or DolphinDOS)
    • GPIB (IEEE-488) interface (switchable to CBM bus and parallel cable)
  • Write / read amplifier
    • Adapted system from the 1541-II. All analog technology has migrated into one chip.
  • Power supply
    • Completely new designed power supply by high efficient switching regulators
  • Size
    • Saving of almost the whole TTL technique, and replacement by a CPLD
    • Board has shrunk to about half, compared to the so called “Short-Board”.
  • Other
    • RAM extended to 32kByte
    • 64kByte EEPROM for space of 4 Kernals of 16kByte size each
    • Switchable between 1541 and 2031 operation
    • Status LEDs

Installation in the 1541

The installation is very simple. Disconnect all cables from the old system board, remove the screws, replace the board with the new one, tighten the screws and reconnect the cables.

The connections for the CBM bus are still in the same place as on the original board.

Only for the parallel cable and the IEEE-488 cable an opening must be filed into the housing. But this is not very difficult with a ribbon cable.

Since the old metal bracket and heatsink block for the linear regulators is now missing, no mounting is possible on the right side. The board is now only held by 3 screws.

Especially when plugging in the cable for the IEEE-488 bus, special care is needed to avoid bending the board at this point.

Preview

Probably many will now ask themselves why more gimmicks were not integrated, such as electronic disk punch, track display, etc.

As I mentioned at the beginning, this board is just another intermediate step.

Therefore the topic is not finished at this point. There will be another continuation…

5 2 votes
Article Rating
Subscribe
Notify of
guest
4 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments

4
0
Would love your thoughts, please comment.x
()
x