Re: AW: AW: [DDL-ML] Programmiergleis

["Robert Nieuwland" <rnieuwl@wxs.nl>]

Hello harald

I always had problems with my Lenz boosters and the Lenz Switch decoders.
The problem I already discribed here once.
Sent 100 time switch_bend and Switch_straight to a switch and it will do
only about 75 times. The K87 from marklin didn't have this problem.

Yoe wrote
>----- Original Message -----
>From: "Harald Barth" <haba@pdc.kth.se>
>Ich würde folgendes ausprobieren: Ändre in erddcd.c so daß
>nmra_dev_termios (statt B19200) B38400 enthält. Ändre dann bevor du
>erddcd startest mit setserial spd_cust divisor 7 (genaues steht sicher
>in man setserial). Das ändert den tty so, daß beim Anruf von 38.4k
>eigntlich die Spezialgeschwindigkeit mit dem divisor 7 eingestellt
>wird. Das müßte dann die bessere Frequenz erzeugen. Doch ist bei
>diesem Test natürlich das Märklin protokoll nicht drin (-M off), da es
>die "richtige" 38.4k verwendet.

So I thought - wy not give it a try.
And no problems any more. So this does seem to inprove the whole nmra output
from erddcd. For the first time I can use it not only for the trains but for
the switches as well.
The setserial command is the right one.

Thanks and hopefully this will help others as well

Torsten

Perhaps it is posible to make it a command line option to set the serial
nmra frequency to 38400. I don't have to change the source code each time
there is a new version.

Thanks already if you will give it a thought.

Rob Nieuwland