Testers wanted for the XpressNet library.

geraldfg95

Post by geraldfg95 » 13.03.2010, 20:23

rjversluis wrote:Hi Gérald,

the same problems as you had with the prev revision?

Hi Rob
With xpressenet ? , yes d'on't work i any revision , only lenz work .

regard
Gérald

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 13.03.2010, 20:25

Hi,

in your prev. trace are many errors like:

Code: Select all

LI: Buffer overflow
Transaction error; Resend.
no response received within 50ms; resend command 1 of 3 times
Do you still have those error using 1440?

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 13.03.2010, 20:27

Hi Gérald,
geraldfg95 wrote: With xpressenet ? , yes d'on't work i any revision , only lenz work .
To get any further I need some more exact reporting.;-)

ron&bram
Posts: 2460
Joined: 11.06.2008, 19:34
Location: Heemskerk, Netherlands

Post by ron&bram » 13.03.2010, 20:44

Hi Rob, Gérald,

In Géralds last trace with byte level enabled I see other strange things also:

Code: Select all

FF FE E4 13 00 0C 80 7B
According to the Lenz manual this message means:

E4 13: Loc speed message with 128 speedsteps
0C: loc address 12
80: Direction forwards, speed 0

The response from the lvz is:

FF FE 61 82 E3 (byte trace dumps the FF FE on a seperate line)

which means command not supported. I find it very strange that a command station reports that a proper Xpressnet loc speed command is not supported. The message itself was received correctly, otherwise the lvz response would have been FF FE 61 81 E1, received message did not pass xor test (by the way, this message is seen often in the trace too).

Best regards, Ronald

ron&bram
Posts: 2460
Joined: 11.06.2008, 19:34
Location: Heemskerk, Netherlands

Post by ron&bram » 13.03.2010, 20:54

One more thing,

The lenz lib sets the flow to cts hardware handshake, also when the sublib is usb:

Code: Select all

SerialOp.setFlow( data->serial, cts );
The xpressnet lib on the contrary always sets the flow to none when the sublib is usb:

Code: Select all

SerialOp.setFlow( data->serial, none );
Could this be an explanation?

Best regards, Ronald

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 13.03.2010, 21:10

Hi Ronald,

good point!
This could explain the buffer overflow message.
I will make this optional although the current LIUSB manual states:

<DE>
- Baudrate: immer 57600 Bit pro Sekunde
- 8 Datenbits, 1 Startbit, 1 Stopbit, kein Paritybit
- kein Handshake
</DE>

ron&bram
Posts: 2460
Joined: 11.06.2008, 19:34
Location: Heemskerk, Netherlands

Post by ron&bram » 13.03.2010, 21:17

Hi Rob,

That it also says in my version of the Lenz manual, but maybe this manual, like many many others, was written with only that other operating system in mind :wink: .

Best regards, Ronald

geraldfg95

Post by geraldfg95 » 13.03.2010, 21:19

rjversluis wrote:Hi,

in your prev. trace are many errors like:

Code: Select all

LI: Buffer overflow
Transaction error; Resend.
no response received within 50ms; resend command 1 of 3 times
Do you still have those error using 1440?
In the rev 1440 i have no errors message , but the time reaction is long .
Accessory clac-clac-clac ..... , i command a loco to level 40 speed : don't move , i move cursor to 41 for example : move , i click on stop : don't stop , reclick : stop ....
the respoce is not in real time , in automatique mode , loco don't stop at the good place .... so is inusable ...

regard
Gérald

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 13.03.2010, 21:32

OK,

the snowy 1442 revision can be downloaded.
Please set the Flow control in the setup to CTS.

If this does not do the trick I'm out of ideas.

geraldfg95

Post by geraldfg95 » 13.03.2010, 22:25

rjversluis wrote:OK,

the snowy 1442 revision can be downloaded.
Please set the Flow control in the setup to CTS.

If this does not do the trick I'm out of ideas.

Congratulations !!!
you fixed it !!!
first tests are good !
I continue tomorrow , an i ask you the result

Regards
Gérald

geraldfg95

Post by geraldfg95 » 13.03.2010, 22:35

oops ....sorry ....
You do not have the required permissions to view the files attached to this post.

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 14.03.2010, 08:57

Hi Gérald,

the last trace was much better, but at power off it was still sending accessory commands which is useless and the LZV does not like it.

I added a flag in the xpressnet library which signals track power. If the power is off all accessory commands are rejected. This should prevent the LZV Busy / Resend loop.

The 1444 snow revision can be downloaded.

Beaucoup succès.

jeanmichel

Post by jeanmichel » 14.03.2010, 10:34

Hi Rob,

1445 is perfect with LI101f and LZV 100. Good job!

Jean-Michel

rjversluis
Site Admin
Posts: 42343
Joined: 10.04.2006, 08:48
Location: Speyer, Germany
Contact:

Post by rjversluis » 14.03.2010, 11:02

Hi Jean_Michel,

tnx for reporting! :-)

Happy Railroading.

geraldfg95

Post by geraldfg95 » 14.03.2010, 16:45

rjversluis wrote:Hi Gérald,

the last trace was much better, but at power off it was still sending accessory commands which is useless and the LZV does not like it.

I added a flag in the xpressnet library which signals track power. If the power is off all accessory commands are rejected. This should prevent the LZV Busy / Resend loop.

The 1444 snow revision can be downloaded.

Beaucoup succès.
Hi Rob
Thanks for your good job ! 8)
With the v1444 is very better for me , only the commande to accesory are repeted any time .(clac-clac-clac-clac ....)
Can you make me the lattest revision for snow leopard please ?
i think we are near of the ultime rev ....

regard
Gérald

Post Reply

Return to “Lenz”