DCC++ and extended accessory decoders

[aleopardstail]

I have been reading the DCC++ documentation, I can see the "a" command to send an activate bit to an accessory decoder via address and sub address - the values indicate this is the basic accessory decoder (9 bit address + 2 bit sub address - the 3rd bit of the sub address doesn't seem to be mentioned)

 

what I can't see is how to send the extended accessory decoder command, i.e. 11 bit addressing with 5 bit data

 

is this here and I'm being dense or is this not currently supported?

[Crosland]

 

CAn't help with the extended address, but...

 

13 hours ago, aleopardstail said:

I have been reading the DCC++ documentation, I can see the "a" command to send an activate bit to an accessory decoder via address and sub address - the values indicate this is the basic accessory decoder (9 bit address + 2 bit sub address - the 3rd bit of the sub address doesn't seem to be mentioned)

 

Is there a de-activate command, or another parameter to the A command? The third bit of the sub address selects between the two outputs of the (one of four) pair selected by the 2 bit sub-address. For things like solenoid point motors this corresponds to which coil will be activated for normal/reverse, straight/diverging, on/off or however you want to describe the state of a turnout.

[aleopardstail]

5 hours ago, Crosland said:

 

CAn't help with the extended address, but...

 

 

Is there a de-activate command, or another parameter to the A command? The third bit of the sub address selects between the two outputs of the (one of four) pair selected by the 2 bit sub-address. For things like solenoid point motors this corresponds to which coil will be activated for normal/reverse, straight/diverging, on/off or however you want to describe the state of a turnout.

 

yeah the "a" command takes address, sub address then a single bit for the state of the device.

 

weirdly, its the 9 bit accessory address, which makes sense, then a 2 bit sub address - not the three bit the protocol has, apparently by convention it is assumed devices are in pairs (e.g. to make connecting crossovers easy) which I guess happens here, will be interesting once I hook a sniffer up to see what actually comes out.

 

just doesn't seem to be a direct way to set the 11 bit extended decoders, though there is an indirect way of "work out the four bytes yourself and you can send them" which is probably workable if a bit awkward to experiment with, guessing JMRI manages that internally was just wondering for driving via an Arduino serial monitor when testing things out

 

cheers anyway, has cleared up a few things on the basic one

[Crosland]

13 hours ago, aleopardstail said:

weirdly, its the 9 bit accessory address, which makes sense, then a 2 bit sub address - not the three bit the protocol has, apparently by convention it is assumed devices are in pairs (e.g. to make connecting crossovers easy) which I guess happens here,

 

It's nothing to do with cross-overs, the pairs relate to the two sides of a solenoid point motor for controlling a single turnout. 

[aleopardstail]

1 hour ago, Crosland said:

 

It's nothing to do with cross-overs, the pairs relate to the two sides of a solenoid point motor for controlling a single turnout. 

ohhh, now that makes sense, so setting a binary "1" activates one, then a binary "0" activates the other

 

fair enough, obviously not all devices need that but for a basic controller that makes a lot of sense driving solenoids via a CDU or similar

[RAF96]

I thought the sub addresses were to do with the likes of setting feathers and multi-aspect signals.

[Crosland]

1 hour ago, RAF96 said:

I thought the sub addresses were to do with the likes of setting feathers and multi-aspect signals.

 

Not in a basic accessory packet. They are as described above.

 

Signal aspects are controlled using the 5-bit data field in an extended accessory packet. I.e. the type of packet aleopardstail is trying to send.