THEMA: Which is maximum length of s88 bus cable?

As i see, probably s88 bus  is not so popular for  getting occupancy information.
Therefore there are no answers.

------
Vadim

In N-Gauge the S88-Bus is not very popular, Tams Elektronik build better S-88-components:

http://www.tams-online.de/download/index_uk.html

Hallo,
2inch lenght,
from terminal to terminal.

Use shielded wires.


Hans Hans

Hi,

the design of the old parallel s88 Bus is problematic, because it simply uses 5 Volts TTL levels on asymmetric lines, mostly used in conjunction with high-impedance metal-gate CMOS circuits (HEF/CD 4K series) and long cables within a noisy environent. This is no good.

As you mention network cables, i.e. RJ45, I guess you are talking about s88-N.

My suggestions:
- use 12 Volts supply (check for compatibilty with the modules)
- use FTP or SFTP cabling, connect the shield of all cables, but connect it to GND on one end of the bus only
- follow the rules of Cat.5 cabling and do not make T or Star connections (no splitters)

Regard,
Peter

Hans,

are you sure? 2 inches is 50,8 mm. This does not even reach from one terminal to another...

Grüße, Peter W.

Thanks.

Which electronic method for getting  occupancy is the best and more popular for N-Scale?

-------
Vadim

3 Standard-Systems: Sx-Bus, RS-Bus from Lenz, Loconet

Thanks.
But these standarts is too complicated for home making, because it probably requires application of microprocessors and learning assembler ...

-------
Vadim

Let us make an conclusion:

For example, I plan to  use 12V Viessmann  N 5233 s88 Ruckmelders (16 modules), and Viessmann Commander, which have 12V s88 bus.

Does 25 Meters Shielded Cable RG45  work with these 16  Ruckmelders connected consequently in Chain along all 25 m cable?


-------
Thanks,
Vadim

Vadim,


I know you fear struggling with microprocessors. But trust me. There are ABSOLUTELY NO assembler skills necessary for download a HEX file from some web site and flashing the data onto a PIC micro, for example.

Take this into consideration:
http://users.telenet.be/deloof/


I'd suggest to ask Viessmann directly. Their Commander is brand new design, and proprietarly exotic. With long cables, the clock and rise/fall times on data lines are critical.

5233 implements the very classic circuit using two HEF4014 chips, input/output impedances and connectors do not match RG45.

Regards, Peter

Thanks, Peter.

As to Microprocessors:
There are plenty of systems for occupancy detection (Zimo, Lenz, Digitrax, Roco, ESU....etc )I cannot evaluate which system is more cost effective and sufficiently universal for supporting different hardware and software.  

s88 - has absolutely clear logic of operation, and it's simply to make& solder  it without computer, programming, learning  and choosing apropriate processor ...etc.  

As to RG45 and impedance:

The  length of cable is too small (25 M )in comparison with wawelength of highest  frequency of s88.  Therefore shielded cable will work as capacitance. (there is no wawe effects)
As i know HC CMOS chip series can be easily loaded by upto 1 NanoFarade which is equivalent to 25 m of such cable.


------
Thanks
Vadim

Hi Vadim

Take it easy.

Please tell me, how many occupyncy modules u will need?
Please tell me where to place them in your room?
Now tell me the longest distance between 2 modules?
Please tell me, if it is possible to separate your modules in three separated lines?

Ok, then we can decide to use S-88 N Bus with 5 V.
We can decide to use HSI-88 or compatible hardware.
We can decide to use power modules in the lines if necessary.

There are many expensive systems on the market. Maybe some are better in theory.
But if you only want to check occupancy on your N-railroad system, you should use the simple and good working S-88 N System...

Greets
Thomas

Hi Vadim,

okay, I have found some data - this site:
http://www.opendcc.de/s88/s88_n/s88-timing_e.html
talks about 30 meters of length being maximum.

But notice HEF4014 is made of metal-gate CMOS, it is not HC (which only will work up to 6 volts). Therefore output drive impedances are much higher, and the 4014 also does not have bus driver outputs.

The wave length is not the problem. The major problem here is the long signal rise and fall times due to slow logic circuits, escpecially on the clock edges, with respect to the length of the line and the resulting duration until the reflected (false) signal comes back to the input. However 12 volts compensates for this, because HEF becomes faster at higher supply voltages - as output resistance is approx. constant, the higher the voltage difference,  the more current the output FET will be able to drive into the capacitive load.

Regards, Peter

Thanks, Peter.
This is quite useful information!
I ll take in mind it!

--------
Thanks,
Vadim