The goal of this document is to create a comprehensive interpretation of LocoNet Personal specification. Its first version is roughy the same as original specification, but subsequent versions would add meaning to badly formulated paragraphs and clarify unclear points. The style should also be improved.
-
LocoNet Personal Use Edition 1.0 SPECIFICATION: https://www.digitrax.com/static/apps/cms/media/documents/loconet/loconetpersonaledition.pdf
-
LocoNet Personal Use Edition 1.0 Extension: http://embeddedloconet.sourceforge.net/SV_Programming_Messages_v13_PE.pdf
This is the definition of the protocol used by Digitrax products that communicate on the Long distance version the LocoNet® network. This LocoNet Personal Use Edition 1.0 information is provided solely for non-commercial private use by Digitrax customers. No rights are conveyed for the commercial use of this information, and Digitrax Inc. is not able to provide technical support for private use. Digitrax conveys no warranty for this information and incurs no obligations for its use or incorrect usage. Possession of this information signifies acceptance of these conditions of usage. LocoNet is a registered trademark of Digitrax Inc.
LocoNet® is a "PEER to PEER" distributed network system on which all devices can monitor the network data flow. The network is event driven by different devices in time, and is not polled by a centralized controller in normal operation. LocoNet is a powerful power decentralized and "scalable" distributed system. This is similar to the way the worldwide telephone system works — i.e. there is no worldwide "master control program" and all telephone central offices follow a strict set of "rules" for worldwide access. The Internet operates in a similar "distributed” manner.
The access technology is Carrier Sense Multiple access with Carrier Detect, CSMA/CD. This type of network technology is the overwhelming choice for connecting computer LAN’s around the world. The first implemented 10M bps version of Ethernet has been upgraded to 100M bps and now has even 1,000M bps extensions. The CSMA/CD is the basic physical or media access layer that allows multiple devices to interoperate and exchange data efficiently.
LocoNet® uses CSMA/CD techniques to arbitrate and control network access. On top of this physical layer, LocoNet® specifies a higher level of message protocol that gives efficient management of data structures for operations in the model railroad environment. A sensible bit rate was chosen as a good compromise between ease of wiring, with no need for terminations or complex distribution rules, and fast latency or rapid response time. We avoided a simple "specsmanship" of specifying faster bit rates to distinguish us from other system manufacturers' capabilities, and instead chose a fundamentally new system architecture we call LocoNet®.
Since LocoNet® is a distributed system where each device can determine the urgency of its access, it is not easy to compare in a one-to-one manner with older technology such as polled bus systems. We can compare the explosive growth of commercial computer LAN’s, and even the Internet itself, to understand that all new "state-of-the-art" data processing systems are based on network type topology. The old idea of a central "mainframe" and "Main Control Program" was state of the art for the 60’s and 70’s and has yielded to new technologies and ideas.
Comparing "raw" bit rates is not meaningful in this context since the strength of a true network system is in allowing multiple queued access without requiring polling or data flow controlling overheads. The real strengths of a LAN become apparent when a large number of devices need fast data access, or low latency, to transfer requests or state change information. With a LocoNet® implementation, a modest bit rate is sufficient on a large railroad layout to give realistic operation. Note that a typical LocoNet® implementation allows us to achieve about 98% of network traffic capacity with less than 1% collisions.
Another decisive advantage of an optimally designed LAN is the ability to overlay or add new capabilities and thus "scale" the system in features as well as size. Not having to modify a "master control program" each time we wish to reconfigure or expand system features is a very real advantage.
An example of this is the way the Digitrax "Big Boy" can be expanded with DT100’s and get a network "fast clock", even though the DT200 master did not itself support this. The soon to be released Digitrax Tetherless throttles (both Radio and IR) can take advantage of the network environment to add a number of different types of features among many interfaces on the same LocoNet ® network. LocoNet® is configured to allow all types of data traffic to flow on a single wiring scheme. This means you do not have to run separate Throttle, Booster, Feedback and PC wiring for the system. Since LocoNet ® is based on LAN technology, a large LocoNet® layout can use Bridging, Routing and Fault isolation techniques as used on commercial networks to expand to a large physical extent.
To allow the addition of multiple independently operating PC’s around a large physical layout using simple access hardware, the LocoNet® was "slowed down" and simplified somewhat over a "raw hardware" driven system capability. This trade off was made to ensure ease of attachment of PC’s, and we at Digitrax feel that once the power of "computer assisted modeling" was appreciated, a single PC would rapidly prove inadequate, as task complexity and demands increased. Also, we feel it is important that PC tasks can be independent, modular and spread out among many different PC’s around the layout. In particular being able to use slower old "AT" 286’s and 386’s running DOS and Windows 3.1 etc., was important, since requiring a dedicated new "Wintel PC" costing over US$2,000 seems like a very expensive proposition for what is essentially a hobby!
The normal LocoNet® state is IDLE, with no data traffic unless a device has information to send. With no traffic flow, the network is RFI quiet.
The full implementation of LocoNet® uses a 6 pin USOC RJ12 style TELCO connector. The network is designed to operate "daisy chained" on unterminated 26 AWG 3 pair cable or flat 6 conductor type 120 ohm impedance ribbon cable. It can be cabled in numerous variations. It is designed to be tolerant of cabling environment. It is permissible for the individual wires to loop back on themselves, noting that the 2 Rail_sync lines are of opposite phase and cannot be connected. The connections are balanced to minimize RFI. The connections may be branched in any combination to yield a "Star" or "Bus" or any combination thereof. Only a single LocoNet® current termination is needed and is typically supplied by the system master.
Pinouts for the RJ11/6 connector are:
-
RAIL_SYNC- white
-
SIGNAL GROUND
-
LOCONET-
-
LOCONET+
-
SIGNAL GROUND
-
RAIL_SYNC+ blue
Using typical UL6010 6 conductor 26AWG Telephone flat ribbon cable, a network may typically have a total parallel cable length of up to 2000 ft (609 m), with no point-to-point length exceeding 1000 ft (304 m). This is using the "standard" Long Distance LocoNet® termination of a 15 milliamp positive current source on pins 3 and 4 of the connectors. The capacitance of 2000 ft of this cable is approximately 68 nF, and the loop resistance of a pair of these stranded 26AWG conductors is 80 ohms per 1000 ft.
The maximum amount of parallel cable is limited by the requirement that the minimum RISE TIME in the region of +2 Volts to +7 Volts, is 0.35 Volts per microsecond. The minimum FALL TIME in the same region should be greater than 0.75 Volts per microsecond.
LocoNet® is a "Wired-Or" multiple access linear network using CSMA/CD techniques. Repeaters, network buffers and isolators can be implemented. For the current single ended implementation, the 2 LocoNet signals, + and -, are paralleled and the RJ11 cable connections become polarity insensitive, wire resistance is lowered and connection reliability is enhanced.
Single ended voltage levels and characteristics are:
-
High = 1 = "MARK" : LOCONET+/- voltage above +4.0 Volts with respect to ground conductors.
-
Low = 0 = "SPACE" : LOCONET+/- voltage below +4.0 Volts with respect to grounds.
-
The data should be received with 1.0 Volt of HYSTERESIS centered on +4.0 Volts.
-
Maximum LOCONET+/- high voltage is +24V and nominal is +12V.
-
Minimum receiver input impedance is 47 Kilohms, measured from pins 3&4 to pins 2&5 (GND).
-
The transmitter is OPEN COLLECTOR to SIGNAL GROUND and should be able to sink 50 milliamps in the "ON" state at no more than 1.6V, and withstand 35 Volts in the OFF state.
-
One single device shall provide the "Wired-Or" pull-up for the LOCONET+/- signals. Typical termination is performed by the packet generating "MASTER" and is a 15 milliamp current source from +12V.
-
Loconet devices may draw up to 15 mA from the RAIL_SYNC +/- lines whenever the voltage is greater than 7V. The unloaded voltage is between 12V and 26V max. It is general practice to provide a LOCAL current limited copy of the closest track voltages, to pins 1&6 of Throttle jacks around the layout. In this case the master "backbone" copy of RAILSYNC +/- is not on the Throttle jack.
-
The RAIL_SYNC+/- are a low power copy of the DCC data to be transmitted to the rails. The signals may be received by a differential receiver and boosted to drive the rails.
-
A device with a separate power supply isolated from Loconet may connect to the LOCONET+/- pins 3&4 and SIGNAL GROUND pins 2&5 with a just 2 wires.
-
To use a 1/4" (6.3 mm) stereo 3-pin plug, the SIGNAL GROUND should be connected to the Sleeve, the LOCONET +/- connected to the Tip, and the Sleeve may be connected as a power source. The power supplied to the Sleeve MUST be a CURRENT SOURCE (from +12V to +26V) and be limited to 20 milliamps maximum, because the Plug shorts the Tip and Ring when initially inserted.
|
ℹ️
|
There must be a typo, Sleeve is mentioned twice, but Ring is not mentioned. Anyway, these jacks are not in use since early 2000s. |
LocoNet® data is sent in normal asynchronous format using 1 START bit, 8 DATA bits and 1 STOP bit. The 8 bit data is transmitted LSB first. The bit times are 60.0 uSecs or 16.66 KBaud +/-1.5%. A PC serial "COM" device can use the convenient rate of 16.457 KBaud. This corresponds to a divisor of 07 for the standard NS8250 UART chip or equivalent used by most compatibles. Bytes may be transmitted "back-to-back", with a Start bit immediately following the Stop bit of the previous character. Normal network "IDLE" is the "MARK" voltage state. Data is sent half duplex and transmitters process the transmit echo to monitor network collisions.
CARRIER DETECT (CD) for fundamental network access timing may utilize simple RC time constant "one-shots". CD becomes active immediately on any detection of network in the SPACE state. It then times out for 20 bit times or 1.2 milliseconds as the CD BACKOFF time and goes inactive. CD jitter of up to 180uS is acceptable and helps ensure even statistical network access with minimal collisions.
All transmitters are responsible for detecting TRANSMIT COLLISIONS on a 1 bit or whole echo-byte basis. If a TRANSMIT collision is detected the TRANSMITTER will force a line BREAK of 15 BIT times with a Low or "SPACE" on LocoNet®, and decrement the Transmit Attempt count. (The device can attempt the next acess at the same Priority, or change it by some small amount, depending on an internal Phase reference, if the delay from Network free to Siezure is greater than 2 uS).
All receivers will process the BREAK as bad data framing and reset Message parsers The network is then free to re-arbitrate access. Any message that has format or framing errors , data errors or is a fragment caused by noise glitches and does not completely follow the MESSAGE FORMAT will be ignored by ALL receivers, and a new OPCODE will be scanned for re-synchronization.
To SEIZE access to the LocoNet® a device shall wait for the CD BACKOFF time to elapse from the last space level seen on LOCONET+/-. The "MASTER" device may at this time seize the network immediately upon seeing CD has "released". All other devices add additional time delays before being allowed to attempt NETWORK SEIZE. Throttles and other devices will always wait a minimum of another 6 bit times or 360uS MASTER delay before being allowed to attempt a network seize or access.
On the first attempt to access the network to transmit new input information, a device will add a further PRIORITY delay of up to 20 bit times. If network access is not gained after the priority delay, due to seizure/usage by another device, the PRIORITY delay is decremented by 1 bit time for the next access attempt, which may occur after the current message or fragment ends. In this way all devices may be queued in priority, and none may seize the network in priority over the MASTER, which often returns acknowledgments and other information based on a previous request message.
A device shall make at least 25 Transmit Attempts before deciding Message Transmit failure.The Transmit Attempts must include attempting Network access for at least 15 milliseconds per access attempt.
A BUSY opcode is included to allow the master to keep the network active whilst it is performing a task that requires a response, and entails a significant processing delay, i.e. it can ensure no new requests are started until it has responded to the last message. In addition to the BUSY opcode, the master may simply add 15 bit BREAK sequences to the network to delay any new messages starting until it has completed and responded.
Individual device types may have their access tailored by setting different maximum and minimum PRIORITY delays. In particular, SENSOR type devices may have initial Priority of 6 or less, so they can broadcast messages to the network in a timely manner.
To provide the greatest protection against network bandwidth being wasted due to repeated collisions a device should assert the SPACE of the start bit of the message OPCODE within 2 microseconds of determining that its access delays have elapsed and the network is still free. This has the effect of improving the COLLISION aperture uncertainty for a transmit collision. If the transmitting device detects a transmit collision either by bad TRANSMIT ECHO or a TRANSMITTED 1 bit being forced to 0 on LOCONET, it will initiate the 15 bit BREAK sequence to flag all devices that data is bad.
A simple "COM" port on a PC may access the network by a more direct method.
The protocol has been encoded so that a PC may watch the LocoNet® message dialog and infer that the network is free because
the last message decoded does not imply a follow-on response, so that the network is immediately free for a new message dialog.
In this situation, the PC may immediately seize the network before the CD BACKOFF time has elapsed.
This allows the PC to pre-empt all other devices and completely control the LocoNet® to the level desired. Note that the message <81><7e> is a "time burner" NOP code sent by a Master to restart the CD Backoff timers, and hence keep the network busy in a hardware sense.
This <81> opcode should thus be simply stripped and ignored.
Several PC’s may share access to LocoNet® by subdividing the 20 bit CD BACKOFF delay into priority windows for access. They are responsible for detecting transmit COLLISIONS by checking their TRANSMIT ECHO data and watching a CARRIER DETECT to see if a PC transmit "window" is active already, before attempting to transmit.
If the LOCONET+/- signal remains at a fixed SPACE (low) level for more than 100 milliseconds, a DEVICE will assume a DISCONNECT state is in effect. From this DISCONNECT state or initial start-up state a device will wait a 250 millisecond STARTUP backoff before attempting to access the network. A device will not need to reset its internal state upon DISCONNECT and re-connection, but if it is maintaining a SLOT in the refresh stack it will be required to check the SLOT status matches its internal state before re-using any SLOT. If a device diconnects from LocoNet® and so does not access or reference a slot within the system PURGE time, the master will force the unaccessed SLOT to "COMMON" status so other system devices can use the SLOT. The typical purge time of a DT200 operating as a Master is about 200 seconds. A good "ping" or Slot update activity is about every 100 seconds, i.e. if a user makes no change to a throttle/slot within 100 seconds, the throttle/device should automatically send another speed update at the current speed to reset the Purge timeout for that Slot.
All LocoNet® communications are via multi-byte messages. The "MASTER" is defined as the device that is maintaining the refresh stack for DCC packet generation and is actively generating the DCC track data. Refresh of information is typically only performed for MOBILE decoders. Stationary type decoders are not refreshed and individual IMMEDIATE commands are sent out to the track as requested.
The MASTER is only privileged in respect to performing the task of maintaining the locomotive REFRESH stack and generating DCC packets. In this way other network transactions may occur that the MASTER does not need to be involved with or understand , as long as they follow the MESSAGE PROTOCOL and timing requirements. i.e. Other devices may have a dialog on the network without disturbing or involving the "MASTER".
Devices on LocoNet® monitor the MESSAGES, check for format and data integrity and parse good messages to decode if action is required in the context. Devices such as Throttles, Input Sensors , Computer interfaces and Control panels may generate LocoNet® messages without needing prompting or polling by a central controller.
Devices frequently will be added and removed from an operating LocoNet ®. The devices and protocol are tolerant of electrical and data transients. The format chosen gives a good degree of data integrity, guaranteed quick network-state synchronization, high data throughput , good distribution of access to many competing devices and low event latency. Also , the devices may be operated without need for unique ID or other requirements that can make network administration awkward.
The data bytes on LocoNet® are defined as 8 bit data with the most significant bit (transmitted last in the 8 bit octet) as an OPCODE flag bit. If the MS bit , D7, is 1 the 7 least significant bits are interpreted as a network OPCODE . The opcode byte may only occur once in a valid message and is the first byte of a message. All the remaining bytes in the message must have a most significant bit of 0 , including the last CHECKSUM byte. The CHECKSUM is the 1’s complement of the byte wise Exclusive Or of all the bytes in the message, except the CHECKSUM itself. To validate data accuracy, all the bytes in a correctly formatted message are Exclusive Or’ed. If this resulting byte value is "FF" hexadecimal, the message data is accepted as good.
The OPCODES may be examined to determine message length and if subsequent response message is required. Data bits D6 and D5 encode the message length. D3=1 implies Follow-on message/reply:
D7 D6 D5 D4 -- D3 D2 D1 D0 (Opcode Flag) 1 0 0 F D C B A Message is 2 bytes, including Checksum 1 0 1 F D C B A Message is 4 bytes, inc. checksum 1 1 0 F D C B A Message is 6 bytes, inc. checksum 1 1 1 F D C B A Message in N bytes, where next byte in message is a 7 bit byte count.
The A,B,C,D,F are bits available to encode 32 opcodes per message length.
The model of the MASTER refresh stack is an array of up to 120 read/write refresh SLOTS. The slot address is a principal component and is generally the second byte or 1st argument of a message to the master. The refresh SLOT contains up to 10 data bytes relating to a Locomotive and also controls a task in the Track DCC refresh stack. Most mobile decoder or Locomotive operations process the SLOT associated with the Locomotive to be controlled. The SLOT number is a similar shorthand ID# to a "file handle" used to mark and process files in a DOS PC environment. Slot addresses 120-127 ARE reserved for System and Master control.
Slot #124 ($7C) is allocated for read/write access to the DCS100 programming track, and the format of the 10 data bytes is not the same as a "normal" slot. See programmer_track.
To request a MOBILE or LOCOMOTIVE decoder task in the refresh stack, a Throttle device requests a locomotive address for use (opcode [OPC_LOCO_ADR] <BF>,<loco adr hi>,<loco adr lo>,<chk> ).
The Master (or PC in a Limited Master environment) responds with a SLOT DATA READ for the SLOT, (opcode <E7>…)>, that contains this locomotive address and all of its state information.
If the address is currently not in any SLOT, the master will load this NEW locomotive address into a new SLOT (speed=0, FWD, Light/Functions OFF and 128 step mode) and return this as a SLOT DATA READ.
If no inactive slots are free to load the NEW locomotive address, the response will be the Long Acknowledgment (opcode <B4>) with a "fail" code, 0.
Note that regular "SHORT" 7 bit NMRA addresses are denoted by <loco adr hi>=0.
The Analog, Zero stretched, loco is selected when both <loco adr hi>=<loco adr lo>=0. <loco adr lo> is always a 7 bit value.
If <loco adr hi> is non-zero then the Master will generate NMRA type 14 bit or "LONG" address packets using all 14 bits from <loco adr hi> and <loco adr lo> with <loco adr hi> being the most significant address bits.
Note that a DT200 Master does not process 14 bit address requests and will consider the <loco adr hi> to always be zero.
You can check the <TRK> return bits to see if the Master is a DT200.
The throttle must then examine the SLOT READ DATA bytes to work out how to process the Master response.
If the STATUS1 byte shows the SLOT to be COMMON, IDLE or FREE ("NEW" in original spec) the throttle
may change the SLOT to IN_USE by performing a NULL MOVE instruction, opcode ([OPC_MOVE_SLOTS]
<BA>,<slotX>,<slotX>,<chk>) on this SLOT. This activation mechanism is used to guarantee proper SLOT usage interlocking in a multi-user asynchronous environment.
If the SLOT return information shows the Locomotive requested is IN_USE or UP-CONSISTED (i.e. the SL_CONUP, bit 6 of STATUS1 =1 ) the user should NOT use the SLOT. Any UP_CONSISTED locos must be UNLINKED before usage! Always process the result from the LINK and UNLINK commands, since the Master reserves the right to change the reply slot number and can reject the linking tasks under several circumstances. Verify the reply slot # and the Link UP/DN bits in STAT1 are as you expected. The throttle will then be able to update Speed./Direction and Function information. Whenever SLOT information is changed in an active slot , the SLOT is flagged to be updated as the next DCC packet sent to the track. If the SLOT is part of linked CONSIST SLOTS the whole CONSIST chain is updated consecutively.
If a throttle is disconnected from the LocoNet®, upon reconnection (if the throttle retains the SLOT state from before disconnection) it will request the full status of the SLOT it was previously using. If the reported STATUS and Speed/Function data etc., from the master exactly matches the remembered SLOT state the throttle will continue using the SLOT. If the SLOT data does not match, the throttle will assume the SLOT was purged free by the system and will go through the setup "log on" procedure again.
With this procedure the throttle does not need to have a unique "ID number". SLOT addresses DO NOT imply they contain any particular LOCOMOTIVE address. The system can be mapped such that the SLOT address matches the LOCOMOTIVE address within, if the user directly Reads and Writes to SLOTs without using the Master to allocate Locomotive addresses
Active Locomotives (including Consist TOP) SLOTS may be released for assignment to BT2 throttles in the "DISPATCH" mode. In this case a BT2 operating in its normal mode will request a DISPATCH SLOT that has been prepared by a supervisor type device. This is included for club-type operations where simpler throttles with limited capabilities are given to Engineers (Operators) by the Hostler or Dispatcher.
To DISPATCH PUT a slot, perform a SLOT MOVE to Slot 0. In this case the Destination Slot 0 is not copied to, but the source SLOT number is marked by the system as the DISPATCH slot. This is only a "one-deep stack".
To DISPATCH GET, perform a SLOT MOVE from Slot 0 (no destination needed).
If there is a DISPATCH marked slot in the system, a SLOT DATA READ (<E7>,..) with the SLOT information will be the response.
If there is NO DISPATCH slot, the response will be a LONG ACK (opc <B4>,..) with the Fail code, 00.
(still in definition stage)
Immediate codes may be sent to the Master by a device. These are converted to DCC packets and sent as the next packet to the rails. They are not entered into any refresh stack. These are available in a system based on the DCS100/"Chief".
Opcodes for access to an auxiliary Service mode Programming Track are included. These requests are not entered in the main DCC packet stream .
Note that several confusing expansions and opcode sequences have been stripped from this LocoNet ® version. An experimenter who implements this protocol correctly should have no problems running on a LocoNet® that has other expanded features. Again, we recommend resisting the temptation to "optimise" or take shortcuts with this protocol since it will lead to guaranteed future problems with your hardware and software.
All Copyrights and rights reserved, Digitrax 1997.
|
ℹ️
|
Any opcodes shown here in itallics are not finalised and are informational only. Do not use. All other OPCODES and states are reserved for future expansion. |
LocoNet® Personal Use version definitions 1.0
DRAFT DEFINITIONS October 16, 1997 SUBJECT TO REVISION
| Opcode | Byte | Description | Follow-on message? Response opcode | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
0x85 |
FORCE IDLE state, B’cast emerg. STOP |
NO |
|||||||||
0x83 |
GLOBAL power ON request |
NO |
|||||||||
0x82 |
GLOBAL power OFF request |
NO |
|||||||||
0x81 |
MASTER busy code, NUL |
NO |
|||||||||
0xBF |
REQ loco ADR |
<E7>[OPC_SL_RD_DATA] |
|||||||||
DATA return If address is not found, master puts address in free slot and sends [OPC_SL_RD_DATA]`<E7>……` If there is no free slot, [OPC_LACK] with argument 0 is returned ( |
|||||||||||
0xBD |
REQ SWITCH WITH acknowledge function (not DT200) |
||||||||||
REQ SWITCH function <SW1> = <0,A6,A5,A4 - A3,A2,A1,A0> - 7 LS bits of address. A1, A0 select 1 of 4 input pairs in a DS54 <SW2> = <0,0,DIR,ON - A10,A9,A8,A7> - control bits and 4 MS bits of address. DIR=1 for Closed,/GREEN, =0 for Thrown/RED ON=1 for Output ON, =0 FOR output OFF Response is:
|
|||||||||||
0xBC |
REQ state of SWITCH |
||||||||||
Request a switch to have specific state. <SW1> and <SW2> are same as in [OPC_SW_ACK]
|
|||||||||||
0xBB |
Request SLOT DATA/status block |
<E7>[OPC_SL_RD_DATA] |
|||||||||
Request slot data / status block |
|||||||||||
0xBA |
MOVE slot SRC to DEST |
<E7>[OPC_SL_RD_DATA] or [OPC_LACK] |
|||||||||
Move SRC to DEST If <SRC> slot is not in IN_USE state, clear SRC. SPECIAL CASES:
Return Fail [OPC_LACK] code if illegal move |
|||||||||||
0xB9 |
LINK slot ARG1 to slot ARG2 |
<E7>SLOT READ |
|||||||||
Make slot SL1 slave to slot SL2. Master LINKER sets the SL_CONUP/DN flags appropriately Reply is return of Slot status <0xE7>. Inspect to see result of Link invalid Link will return Long Ack Fail |
|||||||||||
0xB8 |
;UNLINK slot ARG1 from slot ARG2 |
YES <E7>SLOT READ |
|||||||||
;<0xB8>,<SL1>,<SL2>,<CHK> UNLINK slot SL1 from SL2 ;UNLINKER executes unlink STRATEGY and returns new SLOT# ; DATA/STATUS of unlinked LOCO . Inspect data to evaluate UNLINK |
|||||||||||
CODES 0xB8 to 0xBF have responses |
|||||||||||
OPC_CONSIST_FUNC |
0xB6 |
;SET FUNC bits in a CONSIST uplink element |
NO |
||||||||
;<0xB6>,<SLOT>,<DIRF>,<CHK> UP consist FUNC bits ;NOTE this SLOT adr is considered in UPLINKED slot space |
|||||||||||
0xB5 |
WRITE slot stat1 |
NO |
|||||||||
|
|||||||||||
0xB4 |
Long acknowledge |
NO |
|||||||||
|
|||||||||||
OPC_INPUT_REP |
0xB2 |
General SENSOR Input codes |
NO |
||||||||
; <0xB2>, <IN1>, <IN2>, <CHK> <IN1> =<0,A6,A5,A4- A3,A2,A1,A0>, 7 ls adr bits. A1,A0 select 1 of 4 inputs pairs in a DS54 <IN2> =<0,X,I,L- A10,A9,A8,A7> Report/status bits and 4 MS adr bits. "I"=0 for DS54 "aux" inputs and 1 for "switch" inputs mapped to 4K SENSOR space. (This is effectively a least significant adr bit when using DS54 input configuration) "L"=0 for input SENSOR now 0V (LO) , 1 for Input sensor >=+6V (HI) "X"=1, control bit , 0 is RESERVED for future! |
|||||||||||
0xB1 |
Turnout SENSOR state REPORT |
NO |
|||||||||
<SN1> =<0,A6,A5,A4 - A3,A2,A1,A0>, 7 ls adr bits. A1,A0 select 1 of 4 input pairs in a DS54. <SN2> =<0,1,I,L - A10,A9,A8,A7> Report/status bits and 4 MS adr bits. This <B1> opcode encodes input levels for turnout feedback. I=0 - "aux" inputs (normally not feedback), =1 - "switch" input used for turnout feedback for DS54 ouput/turnout address (encoded by A0-A10). L=0 - input level is 0V (LO), =1 - input level > +6V (HI). Alternately: <SN2> =<0,0,C,T - A10,A9,A8,A7> Report/status bits and 4 MS adr bits. This <B1> opcode encodes current OUTPUT levels. C=0 - "Closed" output line is OFF, =1 - "Closed" output line is ON (sinks current). T=0 - "Thrown" output line is OFF, =1 - "Thrown" output line is ON (sinks current).
|
|||||||||||
0xB0 |
REQ SWITCH function |
NO |
|||||||||
<SW1> =<0,A6,A5,A4 - A3,A2,A1,A0>, 7 LS address bits. A1,A0 select 1 of 4 input pairs in a DS54. <SW2> =<0,0,DIR,ON - A10,A9,A8,A7> Control bits and 4 MS address bits. DIR=1 - Closed/GREEN, =0 - Thrown/RED. ON=1 - Output ON, =0 - Output OFF. If command fails, immediate response of [OPC_LACK] There are special values for SW2 for stationary broadcast and stationary interrogation.
|
|||||||||||
"A" class codes
|
|||||||||||
OPC_LOCO_SND |
0xA2 |
SET SLOT sound functions |
NO |
||||||||
OPC_LOCO_DIRF |
0xA1 |
SET SLOT dir,F0-4 state |
NO |
||||||||
OPC_LOCO_SPD |
0xA0 |
SET SLOT speed |
NO |
||||||||
e.g. |
|||||||||||
<reserved> |
|||||||||||
0xEF |
WRITE SLOT DATA, 10 bytes |
||||||||||
SLOT DATA WRITE, 10 bytes data /14 byte MSG |
|||||||||||
0xE7 |
SLOT DATA return, 10 bytes |
NO |
|||||||||
SLOT DATA READ, 10 bytes data / 14 byte MSG If STAT2.2=0, EX1/EX2 encodes an ID# if STAT2.2=1, the STAT2.3=0 means EX1/EX2 are ALIAS ID1/ID2 are two 7 bit values encoding a 14 bit unique DEVICE usage ID:
|
|||||||||||
0xE5 |
Move 8 bytes peer to peer, SRC→DST |
NO |
|||||||||
SRC/DST are 7 bit args. DSTL/H=0 is broadcast message. SRC=0 is MASTER SRC=0x70-0x7E are reserved. SRC=7F is throttle msg xfer, <DSTL><DSTH> encode address (ID), <0><0> is throttle broadcast. ;<PXCT1>=<0,XC2,XC1,XC0 - D4.7,D3.7,D2.7,D1.7>, ;XC0-XC2=ADR type CODE-0=7 bit Peer TO Peer addresses ; 1=><D1>is SRC HI,<D2>is DST HI ;<PXCT2>=<0,XC5,XC4,XC3 - D8.7,D7.7,D6.7,D5.7> ;XC3-XC5=data type CODE- 0=ANSI TEXT string,balance RESERVED
|
|||||||||||
0xED |
Send n-byte packet (to DCC bus) immediately |
yes LACK |
|||||||||
<REPS> D4,5,6=number of bytes in packet,D3=0(reserved); D2,1,0=repeat count. <DHI> = <0,0,1,IM5.7 - IM4.7,IM3.7,IM2.7,IM1.7> - high bits of packet.
;Not limited MASTER then [OPC_LACK]=<B4>,<7D>,<7F>,<chk> if CMD ok ;IF limited MASTER then Lim Masters respond with <B4>,<7E>,<lim adr>,<chk> ;If internal buffer BUSY/full respond with <B4>,<7D>,<0>,<chk> (NOT IMPLEMENTED IN DT200) |
|||||||||||
|
ℹ️
|
<E7> for slot 0 read return master config information. |
-
SLOT NUMBER: 0-7FH. 0 is special SLOT, 070H-07FH reserved by DIGITRAX.
-
D7 SL_SPURGE 1=SLOT purge en, ALSO adrSEL (INTERNAL use only) (not seen on NET!)
D6 SL_CONUP CONDN/CONUP: bit encoding-Control double linked Consist List
11 LOGICAL MID CONSIST, Linked up AND down
10 LOGICAL CONSIST TOP, Only linked downwards
01 LOGICAL CONSIST SUB-MEMBER, Only linked upwards
00 FREE locomotive, no CONSIST indirection/linking
ALLOWS "CONSISTS of CONSISTS". Uplinked means that Slot SPD number is now SLOT adr of SPD/DIR and STATUS of consist, i.e. is an Indirect pointer. This Slot has same BUSY/ACTIVE bits as TOP of Consist. TOP is loco with SPD/DIR for whole consist. (top of list).
D5 SL_BUSY
D4 SL_ACTIVEBUSY(slot is allocaed by some throttle)/ACTIVE(slot data is sent to track): bit encoding for SLOT activity
11 IN_USE - refreshed
10 IDLE - not refreshed, allocated but can be given to new throttle, see this section
01 COMMON - refreshed, can be given to new throttle
00 FREE - not refreshed, can be given to new throttle
D3 SL_CONDN Shows other SLOT Consist linked INTO this slot, see SL_CONUP
D2 SL_SPDEX
D1 SL_SPD14
D0 SL_SPD283 BITS for Decoder TYPE encoding for this SLOT:
011 send 128 speed mode packets
010 14 step MODE
001 28 step. + Generate Trinary packets for this Mobile ADR
000 28 step/ 3 BYTE PKT regular mode
111 128 Step decoder, Allow Advanced DCC consisting
100 28 Step decoder, Allow Advanced DCC consisting
-
SLOT LOCO ADR: LOCO adr Low 7 bits (byte sent as ARG2 in ADR req opcode <BF>)
-
SLOT SPEED (byte also sent as ARG2 in SPD opcode <A0> )
0x00 SPEED 0 STOP inertially
0x01 SPEED 0 EMERGENCY stop
0x02→0x7F increasing SPEED, 0x7F=max speed
-
SLOT DIRF byte: (byte also sent as ARG2 in DIRF opcode <A1>)
D7-0 always 0
D6-SL_XCNT reserved, set 0
D5-SL_DIR 1=loco direction FORWARD
D4-SL_F0 1=Directional lighting ON
D3-SL_F4 1=F4 ON
D2-SL_F3 1=F3 ON
D1-SL_F2 1=F2 ON
D0-SL_F1 1=F1 ON
-
TRK byte: GLOBAL system/track status.
D7-D4 Reserved
D3 GTRK_PROG_BUSY 1=Programming track in this Master is BUSY.
D2 GTRK_MLOK1 1=This Master implements LocoNet 1.1 capability, 0=Master is DT200
D1 GTRK_IDLE 0=track is paused, B’cast EMERG STOP.
D0 GTRK_POWER 1=DCC packets are on in master, global power up
-
D3 1=expansion IN ID1/2, 0=ENCODED alias
D2 1=expansion ID1/2 is NOT ID usage
D0 1=this slot has SUPPRESSED ADV consist
-
Locomotive address high 7 bits. If this is 0 then Low address is normal 7 bit NMRA SHORT address. If this is not zero then the most significant 6 bits of this address are used in the first LONG address byte (matching CV17). The second DCC LONG address byte matches CV18 and includes the Adr Low 7 bit value with the LS bit of ADR high in the MS postion of this track adr byte.
ℹ️a DT200 MASTER will always interpret this as 0. -
SLOT SOUND: Slot sound/ Accesory Function mode II packets. F5-F8. (byte also sent as ARG2 in SND opcode)
D7-D4 reserved
D3-SL_SND4 F8
D2-SL_SND3 F7
D1-SL_SND2 F6
D0-SL_SND1 F5 1 = SLOT Sound 1 function 1 active (accessory 2)
-
EXPANSION RESERVED ID1: 7 bit ls ID code written by THROTTLE/PC when STAT2.4=1
|
ℹ️
|
This could be STAT2.3 instead of STAT2.4 (as STAT2.4 is not mentioned anywhere, while STAT2.3 seems to mean "treat ID fields as ID") |
Note that a 3 byte DCC track packet configured as <sync>,<1011-1111>,<1000-D c b a > <ecb> is a DCC Broadcast Address to Stationary decoders.
Broadcast LocoNet Switch adr is then <SW2>=<0,0,a,D-1,1,1,1>, <SW1>=<0,1,1,1-1,0,c,b>
The DCC packet <sync>,<1011-1111>,<1100-D c b a> <ecb> is an Interrogation for all DS54’s.
This causes a 2 LocoNet <B1> messages encoding both Output state and Input state, for each sensor adr a/b/c encodes.
Interrogate LocoNet Switch adr is <SW2>=<0,0,a,1-0,1,1,1>, <SW1>= <0,1,1,1-1,0,c,b>.
This is generated by DCS100 at power ON, and scans all 8 inputs of all DS54’s.
The programmer track is accessed as Special slot #124 (0x7C). It is a fully asynchronous shared system resource.
To start Programmer task, write to slot 124. There will be an immediate LACK acknowledge that indicates what programming will be allowed. If a valid programming task is started, then at the final (asynchronous) programming completion, a Slot read <E7> from slot 124 will be sent. This is the final task status reply.
<0xEF>,<0E>,<7C>,<PCMD>,<0>,<HOPSA>,<LOPSA>,<TRK>;<CVH>,<CVL>,<DATA7>,<0>,<0>,<CHK>
This OPC leads to immediate LACK codes:
-
<B4>,<7F>,<7F>,<chk> Function NOT implemented, no reply.
-
<B4>,<7F>,<0>,<chk> Programmer BUSY, task aborted, no reply.
-
<B4>,<7F>,<1>,<chk> Task accepted, <E7> reply at completion.
-
<B4>,<7F>,<0x40>,<chk> Task accepted blind NO <E7> reply at completion.
Note that the <7F> code will occur in Operations Mode Read requests if the System is not configured for and has no Advanced Acknowlegement detection installed.. Operations Mode requests can be made and executed whilst a current Service Mode programming task is keeping the Programming track BUSY. If a Programming request is rejected, delay and resend the complete request later. Some readback operations can keep the Programming track busy for up to a minute. Multiple devices, throttles/PC’s etc, can share and sequentially use the Programming track as long as they correctly interpret the response messages . Any Slot RD from the master will also contain the Programmer Busy status in bit 3 of the <TRK> byte.
A <PCMD> value of <00> will abort current SERVICE mode programming task and will echo with an <E6> RD the command string that was aborted.
<PCMD> Programmer command: defined
-
D7 = 0
-
D6 — Write/Read, 1 — Write, 0 — Read
-
D5 — Byte Mode, 1 — Byte operation, 0 — Bit operation (if possible)
-
D4 — TY1 Programming Type select bit
-
D3 — TY0 Prog type select bit
-
D2 — Ops Mode, 1 — Ops Mode on Mainlines, 0 — Service Mode on Programming Track
-
D1 = 0 — reserved
-
D0 = 0 — reserved
Type codes:
| Byte Mode | Ops Mode | TY1 | TY0 | Meaning |
|---|---|---|---|---|
1 |
0 |
0 |
0 |
Paged mode byte Read/Write on Service Track |
1 |
0 |
0 |
0 |
Paged mode byte Read/Write on Service Track |
1 |
0 |
0 |
1 |
Direct mode byteRead/Write on Service Track |
0 |
0 |
0 |
1 |
Direct mode bit Read/Write on Service Track |
x |
0 |
1 |
0 |
Physical Register byte Read/Write on Service Track |
x |
0 |
1 |
1 |
Service Track- reserved function |
1 |
1 |
0 |
0 |
Ops mode Byte program, no feedback |
1 |
1 |
0 |
1 |
Ops mode Byte program, feedback |
0 |
1 |
0 |
0 |
Ops mode Bit program, no feedback |
0 |
1 |
0 |
1 |
Ops mode Bit program, feedback |
<HOPSA> Operations Mode Programming — 7 High address bits of Loco to program, 0 if Service Mode
<LOPSA> Operations Mode Programming — 7 Low address bits of Loco to program, 0 if Service Mode
<TRK> Normal Global Track status for this Master, Bit 3 also is 1 WHEN Service Mode track is BUSY
<CVH> High 3 BITS of CV#, and ms bit of DATA.7 <0,0,CV9,CV8 - 0,0, D7,CV7>
<CVL> Low 7 bits of 10 bit CV address. <0,CV6,CV5,CV4-CV3,CV2,CV1,CV0>
<DATA7>Low 7 BITS OF data to WR or RD COMPARE <0,D6,D5,D4 - D3,D2,D1,D0> ms bit is at CVH bit 1 position.
(if saw LACK <B4>,<7F>,<1>,<chk> code reply at task start)
<0xE7>,<0E>,<7C>,<PCMD>,<PSTAT>,<HOPSA>,<LOPSA>,<TRK>;<CVH>,<CVL>,<DATA7>,<0>,<0>,<CHK>
<PSTAT> Programmer Status error flags. Reply codes resulting from completed task in PCMD
| D7-D4 |
reserved |
| D3 |
1 = User Aborted this command |
| D2 |
1 = Failed to detect READ Compare acknowledge response from decoder |
| D1 |
1 = No Write acknowledge response from decoder |
| D0 |
1 = Service Mode programming track empty — No decoder detected |
This <E7> response is issued whenever a Programming task is completed. It echos most of the request information and returns the PSTAT status code to indicate how the task completed. If a READ was requested <DATA7> and <CVH> contain the returned data, if the PSTAT indicates a successful readback (typically =0). Note that if a Paged Read fails to detect a successful Page write acknowledge when first setting the Page register, the read will be aborted, showing no Write acknowledge flag D1=1.
The system FAST clock and parameters are implemented in Slot 123 (0x7B). Use <EF> WR_SL_DATA to write new clock information; Request to read slot 0x7B (<BB><7B>..), will return current System clock information, and other throttles will update to this SYNC. Note that all attached display devices keep a current clock calculation based on this SYNC read value, i.e. devices MUST not continuously poll the clock SLOT to generate time, but use this merely to restore SYNC and follow current RATE etc. This clock slot is typically "pinged" or read SYNC’d every 70 to 100 seconds , by a single user, so all attached devices can synchronise any phase drifts. Upon seeing a SYNC read, all devices should reset their local sub-minute phase counter and invalidate the SYNC update ping generator.
<0xEF>,<0E>,<7B>,<CLK_RATE>,<FRAC_MINSL>,<FRAC_MINSH>,<256-MINS_60>,<TRK><256-HRS_24>,<DAYS>,<CLK_CNTRL>,<ID1>,<1D2>,<CHK>
| <CLK_RATE> |
0=Freeze clock, 1=normal 1:1 rate, 10=10:1 etc, max VALUE is 128(0x7F) to 1 |
| <FRAC_MINSL> |
FRAC mins hi/lo are a sub-minute counter, depending on the CLOCK generator |
| <FRAC_MINSH> |
Not for ext. usage. This counter is reset when valid <E6><7B> SYNC msg seen |
| <256-MINS_60> |
This is FAST clock MINUTES subtracted from 256. Modulo 0-59 |
| <256-HRS_24> |
This is FAST clock HOURS subtracted from 256. Modulo 0-23 |
| <DAYS> |
number of 24 Hr clock rolls, positive count |
| <CLK_CNTRL> |
Clock Control Byte
D6 — 1=This is valid Clock information, 0=ignore this |
| <ID1>,<1D2> |
This is device ID last setting the clock. |
[END]