Originally Posted by: efel 
Hi,
I have read this topic, but "period" and "pulse width" are not still clear to me.
Assuming I want to drive a 74490 point motor, using:
switching function :17
pulse width: 30%
Period : 250ms
will the electrical signal be: 20 pulses of 3.75 ms at a 80Hz frequency?
This signal will not be stopped before 250ms even if the controller delivers a -say- 100ms control duration?
Now, what would be the electrical signal if the switching function is set to 18 instead of 17?
Thanks in advance for your help
Fred
18 "detects" end stop cutout and stops driving power to the turnout
For functional use with a "good" 74491. But quite OK for those without cutout switches - just set a realistic period
Summary - leave the m83 with its defaults ....
Pulse width is just a geeky way of saying "brightness" or "% of power"
By varying the PWM value from 0 to 255 you
- vary a lamps brightness from 0 to 100% (of the available maximum)
- provide a weaker or stronger "oomph" to your turnout motors.
The inference being, that M-Track mechanisms are not as efficient as C-Track mechanism...
However this may also be perceived as a "work around" for C-Track mechanism failures with
- poorly designed "bounce back" detention. Less power means less chance of bouncing back and leaving the tongue in an intermediate state.
- underrated end cutoff switches. Less power
MAY reduce the amount of contact arcing causing totally failure of the switch. The logic here is that any AC pulse must pass through "0" and thus facilitate the suppression of continued arcing.
for 30% of 255 use a value of 80.
But I suspect you are better of leaving it at the factory default of 100%(i.e. 255) to remove confusion, and address the turnout motor issue directly.
If you had a newer 74491 with 2 working end cutoff switches - you may want to use Mode 18
Period relates to the duration of the 'overall' pulse to the device.
(During the period, the actual signal may be pulsing to give the appropriate brightness/power)
OR
if you were using one of the fancy non-solinoid functions like Blinking, MARSing, etc. then the period is the duration between each blink.
For the florescent emulation the period is the time it splutters before settling to "ON". For Zoom it is the time to fade on and fade off.
Typically this has been around 1/5th to 1/4th of a second (200-250ms) for a device with a magnetic/solenoid device.
- In the 6021 days you could just hold your finger down longer if you wanted a device to energise longer (good example was an uncoupler)
- with the CS2, the controller definition for an individual accessory address could define and the period (energised) time. HOWEVER this doesn't stand up in practice
Switcing Function Sticking to just the solinoid modes of ...
- 16 = Max. switching : “Period“ indicates the max. switching time
- 17 = Min. switching "“Period“ indicates the min. switching time
- 18 = "Min. switching with end switch Switching time is “period“ or until the end position is reached
... we need to understand two separate aspects.
1) The turnouts with end stop micro switches will (should) cutout the power to the solenoid when the end is reached. On the face of it, it is logically that if the controller m83 can detect this then in should stop wasting any more energy driving the port.
This is what
Mode 18 offers.
This is most noticeable in testing, if you have nothing connected to the port - no mater how long you extend the period for, the observed pulse on the LED associated with the port will be seen for a very short time. (it turns on, determines no load, turns off)
NOTE: as per comment above aboult real world "bounce back" - in cases where the bounce back causes the end switch to make again then it WOULD be good to have the drive energy still available to correct -BUT this is not actually how the 74490 works
2) Controller defined period (and its override).
Please understand here that a CS2 "currently" will send a "turn on" and "turn off" pair of commands regardless. e.g. "Address 83 Red, please energise" and then a "Address 83 Red, please de-energise". While this may offer more variation in what the CS2 can offer, it does mean that, for a majority of cases, it is wasting valuable processing time managing the timing and transmission of a followup secondary command.
To offload this function the delay handling may be delegated to the m83...
PLEASE NOTE THIS IS MY INTERPRETATION OF WHAT I BELIEVE THE M83 IS TRYING TO ACHIEVE.
BECAUSE OF THE WAY A CS2 GFP WORKS THIS JUST IS NOT ABLE TO WORK FULLY
BUT EVEN TESTING WITH A 602x CONTROL WITH SPECIFIC ON OFF COMMANDS MY CURRENT M83 DOES NOT PERFORM FULLY AS DOCUMENTED
THE MOST NOTABLE EXAMPLE IS THE INABILITY TO HOLD THE UNCOUPLER BUTTON DOWN WITHOUT IT PULSING
Mode 17In some cases it is just convenient to say to the m83 "please pulse" and leave the m83 to do the work. This is where the mode 17 "Should" come in. A quick energise/deenergise will cause the m83 to start the pulse to the addressed port (e.g. 83Red) and use its own timer to define how long the pulse is actually on for.
BUT note the "Min Switching" reference.
What this "Might" mean is that IF the CS2 delays its "deenergise command" such that it is after the m83s period setting the port will STAY energised anyway.
One example of possible usage here would be an uncoupler, where you hold you finger on the button for an extended period, while the train moves into position.
The only flaw in this ideal, is that the CS2/m83 does not actually work this way (I actually believe it is BOTH a CS2 and a M83 issue (bug) ).
From my observation it is probably the CS2 is not playing the game, as it insists on sending a double energise/deenergise comand anyway AND if your finger is still on the button it just sends repeated PAIRS of commands so you end up with slow pulsing.
I do note that the software programming reference for the CS2 does provide commands to modify the GFP behaviour in the area of accessory usage.
But as far as I am concerned if you cannot achieve this with the UI it is impractical to pursue.
Mode 16(Should be) Similar to 17 but with the emphasis on the "Max switching" rather than the min switching.
As above, because the CS2 insists on sending its pair of pulse, it is not possible to see this in operation as the CS2s command (pair) is always shorter than the Period setting (unless your eyes can detect the difference of a pulse down to 5ms).
That said, Even the 602x environment can't achieve/use this.
If anyone has a test case which does show these functions working as documented, I would love to hear them.