Some years ago George Senkler and I adapted the instruction we could find on programming an mFx decoder with the Intellibox.

I wrote up how to do it at http://layout.mixmox.com/1/mfx-Intellibox

George has however, 4 Märklin locos with mFx decoders that simply refuse to go into programming mode using the steps we produced. Two locos worked just fine on his system and George brought the errant 4 to my layout today and we got the same results with my Intellibox.

Now having been out of the loop for a few years I am wondering what types of mFx decoders are out there? Did Märklin introduce something to ensure we cannot use the IB to do this?

The instructions with the locos describe the technique of programming the decoders using a 6021 but since George wishes to program addresses above 80 we want an Intellibox solution.

Since Mafi explained that holding down the speed knob on a 6021 was simply sending change direction commands I decided to add mFX decoder programming capabilities to my decoder software and I was hoping to try that out today, but since we could not get the decoders to behave as expected (even manually) it was no test for the software. [The software does all the checking of Special Options and loco protocols for address 80 and the register and new value addresses, sends the direction codes to get into programming mode, and will also send the commands to set the new value.]

thanks Tom

we also tried to read the manufacturer ID using the programming track.

One loco returned 255 and 'pretended' to allow CV writes but was in reality not changing the address.

Two of the others simply refused to respond to reads on the PT.

However, the steps we determined and used before were on the mainline (not the programming track) but are otherwise very similar. We did not try those on the PT track. I need to get him back over here with his errant locos....

Thanks Tom,



yet the method of getting them into programming mode is to send a direction change for address 80 and the method of writing to a CV is to send direction change for address =CV and address=new value.

So if this is true, then this would suggest that as soon as it gets into programming mode, the very next direction change command on address 80 would make it think CV80 is to be written to, and thus it would exit?

I note however that they say the current address or address 80 ! So perhaps ESU ones must use the current address and Trix ones address 80....


I understand the instructions perfectly thanks, I need to get the mFX locos here again to try them out using the programming track.

I suspect that what they are really doing though, is actually mainline programming and using the PT relay to simply provide power while sending the address 80 direction change commands.

This is the same (in theory) as pushing the loco off the insulated third rail while sending the direction change commands - providing power to the loco so that it gets the direction change very soon after getting power.

What does differ is doing the reset before hand - I wonder if that stops/clears any non Motorola commands going out....

If you have more details anywhere else it would be appreciated though.

I decide to look at the instructions that Tom linked to, at the IB even though the locos are not on hand.

It seems that the steps given do not make sense with my IB. (running 1.203 yes its old but I have never needed any of the new stuff)


Says I should press [mode] and then [4] to enter programming mode.

Well on my IB, pressing mode will cycle through the modes, s88 - programming - keyboard.

Doing this in s88 mode would switch to programming mode and the 4 does nothing.
Doing this in programming mode takes me keyboard mode and then [4] will switch a turnout to red.
Doing this in keyboard mode switches me to s88 mode and I am not sure what [4] does there.

Perhaps there is a special shortcut to programming mode in later versions of the IB software...

So knowing how to get to programming mode by myself, I look at the next step:



Nope, when my IB is in programming mode, the [menu] key displays a selection of DCC, Selectrix, Uhlenbrock, etc. decoders. It does not switch the PT relay. Anyway, I have SO16=0 to prevent the PT from having track current.

Their step 7 also assumes that pressing the [menu] key again will switch the relay again.

It seems that they have assumed a certain level of IB firmware.
It seems they have assumed that I have the PT set to run track voltage.
It seems they have assumed that the relay is initially on (so they switch it off in step 6)

Meanwhile it seems all this is attempting to do is provide track power while address 80 is repeatedly having its direction changed. This is also achieved by sliding the loco off some masking tape that is isolating the center rail. It is a good idea to use the PT relay and I may alter my software to use that approach and switch the relay off and on using software, once I get a working mfx loco to test stuff.

If anyone can comment on steps 5 and 6 it would be appreciated.

The IB starts with a blank refresh buffer, sending out no other loco commands than address 80. Maybe this is the essential difference.

Both address 80 and current address should work for both types of mfx decoders.
I always use address 80 coz that one is set to "Motorola old" permanently. Rarely the first attempt fails and I simply try again.

Your conclusions are correct: they use the programming track relay to provide power to the loco at a later stage.
[Menu][4] is a shortcut to get to programming mode, so I presume it's a new feature your software doesn't have yet (1.55 and 2.00 are available for download, but 2.00 should require a registration key to install - 1.55 is free for all).

Pushing the loco onto the track is theoretically the same, but may be slightly different practically if transmission errors occur.
For mfx decoders the main track is used for programming.

I have made a few observations regarding programming of mfx decoders:

1. There are three different ways of sending a direction change:
a) old Motorola direction change commmand
b) Explicit new Motorola direction change command (two versions: forward and backward)
c) New Motorola speed+direction command.

I know that c) won't work for programming. b) works, but you need to change the direction (i.e. alternate between the "forward" and "backward" command). I haven't tested a) for programming. Maybe your IB is configured in a way that uses a) or c) for direction change.

2. The non-ESU decoders may not recognize Motorola commands immediately if they were running mfx or analog last time. This may cause them to not respond to programming. This is cured by running them with Motorola commands with their current address for a few seconds. I have not observed this problem with ESU decoders.

Also generally it seems to be more difficult to get the non-ESU decoders into programming mode. But I have never experienced that it has been impossible. I have just had to apply 2) above and try several times to get it to work.