Thanks Jan, I am aware there are different "Technology Versions" in a number of the 60xx products. And Yes it would be nice to have list from Marklin Identifying
- Versions
- Functionality
- Identification of product (Serial numbers)

re the 6043 Memory I quote the following (I hope Tom wont mind)

Specifically The bit in red helps explain the note in the 60128 manual page 17. But as this only applies to Type B variants it is important to understand which version you are attempting to configure.

Memory Types A and B
There are two versions of the Memory unit. ... They Can be identified by the panel lights that come on when the module is started up.
With version "A" the "Extern" light is on ..
With version "B" the "Off" light is on.

Version "A" is faster and allows for timing sequences to be set between commands as described in the booklet. Simply push Keyboard switches that are not connected to any decoders to causes short pauses between commands. Each keyboard switch can be pushed a maximum of 8 times. This version would be better for many small layouts since it is faster and commands are not likely to be complicated for the quicker speed of transmission

Version "B" fits in with the larger layouts and is slower in its transmission speed. It is possible to change the timing variables between commands in a different fashion. An entire series of commands in a route can be spaced with timing pauses by pushing "Extern " and then one of the buttons A1-C8 before pushing "End" at the conclusion of a route. A1 is the fastest timing pause and C8 would be the slowest. A single pause between commands can be entered by pushing "Off" then the pause variable between A1 and C8 followed by the next command in the series. Be sure to push "End" at the end of the route.

Version "B" of the Memory will also only request a route once when there is constant contact on a track. This usually happens when a car is parked on a track contact section and the signal repeatedly comes through the s88 to the Memory. With version "B" the request is made once until changes occur on the layout (example: another route is requested). Then the constant contact will make the request again.


As to the interlocking discussion, the DIP switches position 3&4 at the back of the unit enable or disable this.

Jos,

Without testing I can conceive that a 6043 set for interlocking may interfere with the rest of the bus to monitor/enforce keyboard operations.
What are the settings of you DIP switches?

Separately, have you attempted programing routes by using multiple presses of the same button. This feature is good for sticky (old M-Style signals) and uncouplers. Up to 8 presses of the same key will cause an extended pause of up to 3 seconds.

Hi Jos.

My 60128 is @ 1.0 and CS2 is @ 1.24 (package 1.3.0(1) )
- I have not seen the update message that you describe


My 6021 is "One Blink" and 6043 is Type B (Extern = OFF on powerup)

From my observations with the LocoNet adapter, Things work as expected ...
- keyboard, memory , loconet generated changes work correctly AND their status is reflected within the CS2 keyboard display
- CS2 generated changes work as long as you only ever use accessories on the same page (group of 16) as the last command from the 6021 bus device

Examples. (With LocoNet adapter) Note:6021 I2C us only addresses 1 to 256 - up to 320 not possible
- fresh power-up - ALL commands issued from any 6021 bus device.
- - CS2 keyboard display on any page - Commands work and are displayed correctly

- fresh power-up - NO commands issued from any 6021 bus device.
- - ANY CS2 keyboard change on ANY panel causes CS2 to STOP

- CS2 Keyboard on the same page (group of 16) as the last 6021 command
- - e.g. CS2 on page 2 (17-32) and last 6021 device command was in the range 17-32
- - ANY CS2 keyboard change ON THIS PAGE works.

- CS2 Keyboard on the different page (group of 16) as the last 6021 command
- - e.g. CS2 on page 2 (17-32) and last 6021 device command was NOT in the range 17-32
- - ANY CS2 keyboard change causes STOP

Conclusion:
There appears to be some "locking" based on the old keyboard (16) boundaries. While It is probable that Marklin could fix this issue with some firmware updates, I feel this is unlikely unless there is some symptom that I haven't found that relates to Marklin products as well as the LocoNet-6021 adapter.

Therefore ... My suggestions fwiw ...
If you do not have Loconet adapter and do not intend to explore it.
- The 60128 offers ROI on an existing 6021. WITH the caveat of checking for component interoperability (As per JOS's experience with one type of Memory)

If you DO have a Loconet-6021 adapter for your 6021.
- Buy a Loconet to PC adapter and link the 2 systems with a small software program instead of spending the money on a 60128.

Do not have Loconet and DO intend to explore it.
- Then spend your money on a Loconet-6021 and LocoNet-PC adapter (or equivilent - e.g. Intellibox) instead of a 60128

Hi again,

Well it's a long time I was busy working on my problem that 2 out of 3 memory devices did not work correctly in combination with the connect-6021 and CS2.

Unfortunately I did not have much time last months to spent time with my modell railroad.
but...

I finally got a reply from Marklin (with excuses for the long delay) saying that the two memories that were not working properly had the need for a microprocessor update. So I sent them in and last week I received the two memories back.

I have tested these 2 upgraded memories and they are now working perfectly.
So besides the A en B versions there have been 'quiet' updates to the microprocessor programming I suspect.

The only drawback is that my third memory that did work ok exhibits now the same problems that it only switches the lights on an attached keyboard and nothing happens on the CS2 when used together with the two updated memories. So i'm going to check with Marklin but I guess it also has to be updated.....

b.t.w., the upgrade costs are about 50 euro's per memory.