Curtain Control with Cortex

[Andrew Millne]

I'm about to set up a single silent gliss curtain track as a test but was wondering the best way to set up cortex.

The unit comes with an exteranl switch that I was intending to replace with a DRB module relay. The operation of the switch is;

Press switch » curtain moves
Press switch during moving » curtain stops
Press switch again » curtain moves back

Any tips?

[Andrew Millne]

I'm looking at the motor control object for the pulse relay 1 open, pulse relay 1 close option

[Gumby]

Cortex had an extra motor mode added specifically to work with AutoGlide tracks (which I think became SilentGliss) by replacing the supplied override button with relays in a DRB or similar. This is the fourth mode, pulse 1 to open, pulse 2 to close, as you have noted.

You can either cut the button off the cable or make your own cable (eg from a modem/telephone cable with RJ11 connector). See my blog (search on "curtains") to see the wiring.

You can set up the open and close times on the "Motor Control" tab, which then allows Cortex to approximate proportional control (ie command open 50%). You can then use the Curtain object connections to trigger open/close however you want ...

I have closing triggered from time and also outside light level. I have opening done by a macro that starts from the scheduler and then watches for several presence conditions to detect movements likely to be getting up (rather than just a nighttime bathroom trip), plus ensuring bedrooms are empty to avoid shocking the neighbours

The proportional control will accumulate errors, but you'll probably do full open and/or full close at least once a day, so not too much of a problem.

Occasionally my curtains (quite heavy) "stick" and this stops the motor before they have fully opened or closed. In theory I could add limit switches to detect completion of the motion, but it's not a significant issue. Not a Cortex/Idratek issue, just a consequence of the SilentGliss controller.

[Karam]

I suppose you could just wire the two relay outputs in parallel and then use the pulse relay 1 to open, pulse relay 2 to close option . But the problem with your setup, as I understand it, is you cannot determine whether a 'press' will cause an open or close. You could start off by synchronising manually but relying on dead reckoning over a period of time is bound to fail at some point.

[Gumby]

Oh, hang on, I misread your post.

Have SilentGliss changed the button? Mine had a definite but small rocking action, internally it had two switch contacts and the wiring had three cores: common, "open" and "close".

[Gumby]

If the connector is still and RJ11 you could buzz it out with a multimeter and see if there is still an "open" and a "close" contact. If there is then it's same as mine. If not then it's going to be more tricky.

[Andrew Millne]

I suppose you could just wire the two relay outputs in parallel and then use the pulse relay 1 to open, pulse relay 2 to close option . But the problem with your setup, as I understand it, is you cannot determine whether a 'press' will cause an open or close. You could start off by synchronising manually but relying on dead reckoning over a period of time is bound to fail at some point.

I originally had this thought and came to the same conclusion.

Oh, hang on, I misread your post.

Have SilentGliss changed the button? Mine had a definite but small rocking action, internally it had two switch contacts and the wiring had three cores: common, "open" and "close".

I've done some more investigating and although the button is the same rocker style switch it is now only using two of the cores.

Luckily I found this handy PDF (http://www.letsautomate.com/pdf/silentglisswiring.pdf)

It seems as though the unit is now measuring the voltage drop across the same two cores rather than having distinct cores for open and close. So I guess all I need to do is dig out some resistors.

[Gumby]

Wow, the price of copper must have jumped to make cutting out one core worthwhile.

Still, glad you've found a solution.