cemexdan Posted December 31, 2013 Share Posted December 31, 2013 Hi all, I have just installed RM on a windows 8 pc, no problems at all. I then installed e-link and although the process didn't quite agree with the Hornby instructions once again I seem to have it installed and connected. I have drawn my track plan point control and all, once again very impressed. My question is this, when I go into set up I have selected eLink in the drop down and set baud rate to 19200, this is also set on the com port too. When I close the settings box it asks me the following "you have specified a Hornby eLink although you have set a baud rate which is not 115200" "are you sure you want to keep this?" What is the correct baud rate, does anyone know? Link to comment Share on other sites More sharing options...
Fishmanoz Posted December 31, 2013 Share Posted December 31, 2013 115200 as per the v1.55 manual in Setup. Link to comment Share on other sites More sharing options...
cemexdan Posted December 31, 2013 Author Share Posted December 31, 2013 Thank you I will change the setting on the port and RM, next question how do I set the number of a set of points, all mine say 0000 Link to comment Share on other sites More sharing options...
cemexdan Posted December 31, 2013 Author Share Posted December 31, 2013 Silly me, I have now found how to change point numbers, just need 5 accessory decoders now! Link to comment Share on other sites More sharing options...
Fishmanoz Posted December 31, 2013 Share Posted December 31, 2013 Well done Dan, keep answering your own questions and you'll soon be answering other's too. And by the way, wasn't being abrupt above, just economical with words. Link to comment Share on other sites More sharing options...
Fishmanoz Posted December 31, 2013 Share Posted December 31, 2013 That should be others' shouldn't it. Link to comment Share on other sites More sharing options...
Gregd99 Posted January 5, 2014 Share Posted January 5, 2014 I have never been able to figure out what baud rate actually does when implementing a virtual serial port over a USB connection. Can anyone enlighten me? Link to comment Share on other sites More sharing options...
Fishmanoz Posted January 6, 2014 Share Posted January 6, 2014 I'd assume it is setting a data transfer rate that the controller can handle from RM. Link to comment Share on other sites More sharing options...
96RAF Posted January 6, 2014 Share Posted January 6, 2014 I would have thought it was best set to the highest possible rate that the controller and RM could use to reliably talk to each other without skipping any commands or feedback. Obviously the higher the rate set then the more responsive the program would be whilst risking skipped commands. The slower the rate set the more reliable the communication but expecting a more lethargic response to commands. I have no idea how USB compares to a serial port in this respect, nor if connecting to a USB 3.0 port gives any advantage over USB 2.0 or even USB 1.0. Link to comment Share on other sites More sharing options...
Graskie Posted January 6, 2014 Share Posted January 6, 2014 I always thought the baud rate was how fed up you're getting with your railway! Seriously, though, I think Fishy is correct - speed of data transfer. Link to comment Share on other sites More sharing options...
Fishmanoz Posted January 6, 2014 Share Posted January 6, 2014 I would assume that the default rate for each controller is the maximum the controller can handle reliably and that the controller is the limiting factor, not the USB port and cable. So I'm not surprised to see the much newer design eLink can handle a much higher rate than Elite, speed increases being the way of the world in electronics. That said and given the higher eLink rate, it is also not surprising that it needs a good quality cable to support that rate. Link to comment Share on other sites More sharing options...
frankstewart Posted January 7, 2014 Share Posted January 7, 2014 hi all, just louded an Rm onto windows 8pc it was a nightmare ,had to put it in to a pc shop to get it done however its done ,now will start to do the other things ie programs,i no it will take some time but with all your help i mae make it a bit faster ,just got my first answer and thats the baud rate ,cheers for that one Link to comment Share on other sites More sharing options...
Fishmanoz Posted January 7, 2014 Share Posted January 7, 2014 Welcome to the forums Frank. And it gets easier after a while too, all the best with it. Link to comment Share on other sites More sharing options...
frankstewart Posted January 7, 2014 Share Posted January 7, 2014 cheers and thanks for that ,hope the sore heads are gone ,just got the sound to work on a class 66 ,now going to scratch it and repeat just for practice will be watching for more tips and pass on any cheers Link to comment Share on other sites More sharing options...
Gregd99 Posted January 8, 2014 Share Posted January 8, 2014 Fishmanoz said: I would assume that the default rate for each controller is the maximum the controller can handle reliably and that the controller is the limiting factor, not the USB port and cable USB implements a packetised protocol with a frame/packet period of 1mS (USB1.1) or 125uS (USD2.0). The elite/elink must be able to receive packets at that rate. The xpressnet protocol that runs to the elite (and I assume the elink) is a poll and response protocol. The PC is the master (sends poll) and the elink/elite is the slave (responds to poll). Xpressnet implements a mechanism for the slave to say "I am not able to handle the poll I just received". The master needs to interpret that response. so..... the control flow mechanisms seem to be in place and the bit rate on the link seems to be fixed. this is what led to my original question "what does baud rate mean on a serial over USB link?" I am interested to understand as I suspect that there is a simple answer that I can't see. Link to comment Share on other sites More sharing options...
96RAF Posted January 8, 2014 Share Posted January 8, 2014 Gregd99 said: Fishmanoz said: .... Xpressnet implements a mechanism for the slave to say "I am not able to handle the poll I just received". The master needs to interpret that response.... This could explain why now and again a command fails to execute during a program, but next time I run it it is OK. A pity that when the poll is negative RM doesn't automatically send the command again until it gets a positive response, to avoid a complex program sailing on and the locos getting all out of sync. I use the Elite as my RM loco controller and eLink for my points. Link to comment Share on other sites More sharing options...
Gregd99 Posted January 8, 2014 Share Posted January 8, 2014 RAF96 said: This could explain why now and again a command fails to execute during a program, but next time I run it it is OK. I use the Elite as my RM loco controller and eLink for my points. The elite has an extra "feature" where it sometimes just does not respond to a poll. it seems that the poll is lost before it is processed. eg a setThrottle command is sent but never received by the loco. there is no ack or nack sent back to the PC. in my jmri setup I detect the lack of response and resend the command. missed commands can sometime be quite a problem - eg throw points, stop loco... - and so I had assumed that RM detected this situation and resent the command. Perhaps it does not. You will probably recall there have been some threads suggesting commands are sent twice from RM. I do not know to what extent the elink exhibits the same behaviour as the elite. Link to comment Share on other sites More sharing options...
96RAF Posted January 8, 2014 Share Posted January 8, 2014 Gregd99 said: RAF96 said: This could explain why now and again a command fails to execute during a program, but next time I run it it is OK. I use the Elite as my RM loco controller and eLink for my points. The elite has an extra "feature" where it sometimes just does not respond to a poll. it seems that the poll is lost before it is processed. eg a setThrottle command is sent but never received by the loco. there is no ack or nack sent back to the PC. in my jmri setup I detect the lack of response and resend the command. missed commands can sometime be quite a problem - eg throw points, stop loco... - and so I had assumed that RM detected this situation and resent the command. Perhaps it does not. You will probably recall there have been some threads suggesting commands are sent twice from RM. I do not know to what extent the elink exhibits the same behaviour as the elite. @Greg99 As I stated some time back Rocrail echoes that the eLink fires off a 5 x command for points at least, so 2x command from the Elite would seem reasonable as a safety net. Unless Hornby elects to promulgate the protocol for the Elite and eLink we are virtually stuffed as to understanding why the kit behaves as it does. Other manufacturers such as Lenz publish such data so why is Hornby so shy. Surely any 'secret squirrel' stuff will be protected by patent. Such data in the public domain would allow 3rd Parties to integrate better with Hornby kit. Rocorail gives Hornby kit the finger and advises users to buy compliant kit, thus avoiding problems. Link to comment Share on other sites More sharing options...
Gregd99 Posted January 8, 2014 Share Posted January 8, 2014 Robbie, The messages between the elite and RM are expressnet. I had assumed, and now think correctly based on your rocrail comment, that the messages are also xpressnet between the elink and RM. The spec is available @ http://www.lenzusa.com/1newsite1/Manuals/xpressnet.pdf There are a couple of gotchas where Hornby have misinterpreted a few messages but this has been figured out and is supported in jmri. the misinterpretation is to do with point numbering and then with status messages towards the PC. Another issue is the elite propensity to lose messages:-( as mentioned above. It seems that some additional preamble messages have been inserted with the elink and RM. this currently means that jmri will not operate the elink. I have not had to time to reverse engineer the messages and really wonder what the purpose was in inserting them. Surely not to make it hard for other apps to communicate with the hornby kit? Your comments re the elite and rocrail are reflected in the jmri forum where the hornby kit is seen as low quality and non compliant. I really think that this is unfortunate as but for a few small glitches I think the elite is very good. I really wish the glitches could be addressed but they are apparently not a priority for Hornby. Link to comment Share on other sites More sharing options...
96RAF Posted January 9, 2014 Share Posted January 9, 2014 Gregd99 said: Robbie, The messages between the elite and RM are expressnet. I had assumed, and now think correctly based on your rocrail comment, that the messages are also xpressnet between the elink and RM. The spec is available @ http://www.lenzusa.com/1newsite1/Manuals/xpressnet.pdf There are a couple of gotchas where Hornby have misinterpreted a few messages but this has been figured out and is supported in jmri. the misinterpretation is to do with point numbering and then with status messages towards the PC. Another issue is the elite propensity to lose messages:-( as mentioned above. It seems that some additional preamble messages have been inserted with the elink and RM. this currently means that jmri will not operate the elink. I have not had to time to reverse engineer the messages and really wonder what the purpose was in inserting them. Surely not to make it hard for other apps to communicate with the hornby kit? Your comments re the elite and rocrail are reflected in the jmri forum where the hornby kit is seen as low quality and non compliant. I really think that this is unfortunate as but for a few small glitches I think the elite is very good. I really wish the glitches could be addressed but they are apparently not a priority for Hornby. @Greg The Lenz docu is just what I would have expected Hornby to have produced in support of their kit, seeing as it is stated some kit is NMRA compliant. I note particularly the Lenz statement on Xpressnet timing constraints which seems to be where some of the skipped command problems could originate. Appreciating that a product must pay its way these days it would be nice if maybe Mr Hornby could allocate a few quid to allow one of the support techs to write up a compliancy statement against the various NMRA/Xpressnet/RailCom/etc specs even if only in a tick box format. Surely that effort should be part of any product introduction procedure. At least then tech savvy users and 3rd party concerns would have a steer about how the kit behaves and could act accordingly. Nuff said - I will start a new thread about Xpressnet and RailCom, etc... Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.