Shutter Relay Fail Safe

Hi CISers,

Just installed a shutter relay, and am trying to set the fail safe value equal to the run time of the blind for full travel. The idea is to enable reliable percentage open settings. For example, when blind is fully closed, open to 50% should land at almost exactly the same spot as when moving to 50% from fully open.

Having limited success with this, and I believe the issue is timer resolution of one second.

Can anyone think of a better solution than setting the top limit setting of the blind to slightly lower (physically) to match the time of travel set in the fail safe field?

Cheers!

Steve.

 

I found that trying to get accurate percentage openings was pretty much impossible because the blind speed varies with direction of travel and takes time to accelerate and decelerate so it rapidly loses position. Until the shutter relay takes these into account and has current sensors to detect end of travel and realign itself it's not really up to proportional control. I am currentl only using proportional control to set a projector screen between 4x3 and 16x9 and I always fully close it and then open to the desired percentage. Using this I still only get about a 30mm repeatability.

 

When the shutter relay is given commands to fully open or to fully close, the internal model of the blind position IS reset to match reality. That's all that can be done, because the relay can be used for hardware which has varying speeds depending upon direction or temperature or other factors. If you want reliable opening to a defined position, as you suggest it would be best to always start from the same position. If you always start from fully closed for example, you are more likely to get consistency.
The only way to improve on this is to get some positional feedback from the blind. I don't know how common this is (I haven't seen it). If there is some standard signalling method used, it would be easy to get exact positioning control.
That would be cool!

 

As always, love your work, gents.

Thought as much. Even if the timer resolution were greater, there would likely be wild variation (as much as a dad trying to catch a kid swinging wildly in a playground... eventually someone's ghoulies are gonna get hit!). Great idea of always starting from a known position, and will have a crack at it. My PAC should make short work of that...

One thought about positional accuracy, Don... A hall effect sensor on the blind feeding pulses to a GIM, monitored by a PAC. Could get a quite accurate speed rate. Would need a slght proportianal adjustment for blind winding radius... Only a little bit hard core, given the variance of starting position. Reckon I can blow at least $2K and six weeks on that thought... and for the longer term, also can really see Somfy building a hall sensor pulse line in, too... not. Drop that in a suggestions box, eh? Wrought iron hang gliders spring to mind.

LOVE automation! Love it.

Cheers,
S.

 

Actually, before the shutter relay came out I was using my own home developed one. This included a current sensor so the controller knew when the blind stopped. Whenever the blind was fully opened or closed it timed the duration and internally calibarted itself for each direction. It also made approximate allowances for multiple starts and stops and ended up tracking surprisingly reliably. My main reason for changing to the shutter relay was future support, particularly for any new owner

What would be nice is a small cbus rotary encoder that could be placed in blinds, movie screens, garage doors etc.

 

Hi there,
some more options for you guys:
elero has a motor (radio) which can memorize (built in encoder = highly accurate) 2 intermediate positions via remote. Interface is available so commands could be sent via C-bus ("double click") to access these 2 positions.
If you REALLY want to go into the deep end, have a look at www.smi-group.com
This is a standard motor communication protocol which allows motor position feedback, motor status (overheated, blocked, etc) and much more. Of course you can drive to (Accurate!) positions (technically up to 65536 but mostly limited to "only" 255).
This protocol even allows connection of up to 16 motors in parellel via 1 5-core, standard 230V cable and yet still indivdually control all motors.
Interestingly: Schneider (Clipsals owner) is a member of this group but doesn't have a controller....
elero is having motors for this system, but no interface for C-bus either. However, there is a RS232 interface available which is able to control up to 16 motors. However it is originally designed for motor setup, not for running. Therefore programming has to be done very carefully not to effect motor positions or parameters.

 

@Ashley

Just some fruit for thought:
Why not use a reed sensor with some magnets. (On tube and idler support)
You probably easily can count whenever it passes the sensor. If you always stop from one direction (from top down) the application should be stopping always at the same spot.
Problem with Encoder is: You would almost always need to re-design the support end to allow for enough space and proper fixing for an encoder.

 

As a second option to the reed sensor. A hall sensor array and some magnets in the bottom of the blind inside the lip. Hall sensors are part of every car built today so there are literally thousands of them in scrap yards all over the country. The advantage of the hall sensors is that they can be tuned in pairs to ignore blinds going up and operate according to logic when they are going down or up. Any PIC chip (the same ones used in satellite hacking cards) is edquate to control the whole thing. Arduino is the lastest catch phrase in the marketplace. All of these tiny little chips are the reborn versions of the Apple IIe. And so on.

A person who is able to code a PAC unit should have the ability to build sensor logic using some CBUS input modules and a bistable series of sensors. Using two logical states at a point allows you to determine the direction of travel.