Motorized Window Blinds – Phase 3

December, 2014

Wow, what a nightmare!  The final stage in motorizing my home’s window blinds has taken far longer than I ever imagined, and it’s still not done!  This phase was focused on adding motorized shutters to the two quarter and one semi-circular arch windows in my Great Room.  This 6-month phase started in June, 2014 with long pauses between activities that were beyond my control.

Before 2

Product Selection

I did a lot of research and found 2 main options for motorized arch window coverings:  Cellular shades and blinds.  Among the front runners were window treatments from Norman Shutters’ Perfect Tilt and Adjust-A-View Movable Arches by Omega. I got pricing for both and found the Norman blinds to be much less expensive (one-quarter the price:  about $2,600 USD) so went with those.  I went with my favorite blind installer since they have provided good customer service in setting up my current and previous house, and they don’t charge sales tax (since they treat the installation as a service), unlike others.

I selected Norman ShuttersPerfect Tilt RF system for a variety of reasons.  I liked the Norman Shutters Woodlore Plus 2 inch blind option because it looks like wood, but is actually made of ABS Polymer that resists warping, and discoloration from extended sunlight exposure.  They run on 8 AA batteries embedded in the window frame, and come with a remote control that can control shutters individually or in groups or all at once.  According to the videos, the motors are silent (and I confirmed this once they were installed – slow, but silent), which is much better than the Somfy or Hunter Douglas window coverings I installed earlier.

At the time I purchased, I was told that Somfy motors were used and that my Somfy remote control system could be used to control them.  In Phase 2, I had deployed Somfy motors throughout the house, so I figured this would be the perfect solution for integration into the existing automated blind control system.

I’ll state clearly here an now – I’m very disappointed with the Norman Shutters for a variety of reasons, and DO NOT RECOMMEND them to anyone.

Installation

Since each shutter is custom made to fit the window, the first step was to take careful measurements and use a template to capture the nuances of the window frame.  My initial reaction was to just measure the radius of the window and use that information for sizing.  However, the installer pointed out that the windows were not perfectly circular so careful measurement was required.

Norman has a factory in China, so the templates were shipped to China on a slow boat.  Weeks later, my installer contacted me and explained that Norman had rejected the templates because they were using older (paper-based) templates and not the newer (plastic-based) templates.  So, they had to re-do the templates and ship those over which contributed to the overall long duration.  Next, Norman returns a CAD drawing to the installer for review and approval.  Once approved, Norman starts to build the shutters.  It was 3+ months before the blinds arrived.  I was told that the custom manufacturing process, plus the long ship time was the cause for the duration.

The installer brought them to my house and had them installed within an hour.  The installation did not meet my expectations, and I was frustrated when I came to inspect.  For maintenance, the shutters are hinged and secured with a magnet.  For the lower blinds, I wanted the hinges to be on the vertical side of the window, but the installer put them on the bottom (horizontal side) so they swing down, instead of sideways.  The main problem with this was that it required them to add a new shelf below the blinds to mount the hinges which blocked my access to the wood blinds that were already hanging below.  For the upper blinds, the installer used only magnets and did not include any hinges.  Maintenance requires climbing a tall ladder, detach them completely, and bringing them down.  Replacing batteries on these shades is not going to be a lot of fun.

After 5

Next, the installer attempted to connect them to the remote control provided.  Norman did a good job of making the pairing process easy – when the batteries are first installed, they automatically enter configuration mode.  You can set them up to work on an channel individually or in groups.  This is where the problems really began.

Two of the blinds worked fine with no adjustment needed.  For 2 of the blinds, when attempting to move the blinds with the motors, we heard a loud repeated clicking noise.  The installers were clueless, so they called Norman for guidance.  The Norman rep attempted to diagnose the problem over the phone, and concluded that the motors and/or remote was bad.  During the diagnosis process, we swapped out some of the bad motors and replaced them with the working motors from the other blinds.  This resulted in loud clicking from the swapped motors.  4 new motors and a new remote were ordered.

About a month later, the motors arrived and the installer attempted to replace the old motors with the new ones.  Unfortunately, the exact same problem happened with the new motors – loud clicking.  Again, the Norman rep was unable to assist over the phone and the installers departed with a promise to return with a solution.

About a week later, the installer called me and informed me that Norman was about to release their next generation motors which should be much better.  I agreed to wait, with the understanding that they would not be available for at least 1 month.  While waiting, I got curious and opened up one of the motors.  During the removal process, I heard the were pieces inside rattling around.  Inside I found a bunch of plastic gears that had been cracked and shredded.

About a 6 weeks later, the installer returned with the next generation motors and a new remote.  During installation ,we did not hear any clicking – yeah, progress!  However, we found a new problem – the angle of the shutters was not consistent across all 4 shutters.  When sent the signal to close fully, two would close fully, but the other 2 would not and allowed small amount of light through.  The Norman rep attempted to walk us through the process of fine tuning the shutter over the phone, but the remote would not accept the command to enter into the tune mode.

After 2

Finally, the installer took some pliers and pulled out the axle of the gear motor, rotated it 10 degrees and re-inserted it.  We re-installed the motor and it worked better.  I can live with the results (above).

Conclusion:  The gear axles were installed incorrectly at the factory.  Attempts to move them with a motor without a feedback mechanism caused them to hyper-extend and strip the gears.  Replacing the motors without fixing the axle problem resulted in stripping more motor gears.  It appears that the new generation motors have a feedback sensor to prevent the hyper-extension which ultimately allowed us to properly diagnose the problem and fix it.

Integration

Once we got past the motor problems, I attempted to pair the motorized shades with the Somfy remote so I could tie them into the home automation system.  Unfortunately, no attempts to get these to work together were successful.  I was very frustrated with my installer at this point since they assured me they were compatible and this compatibility was a primary reason for selecting the product in the first place.  Grrr!  The installer felt badly, so he left me the ‘non-working’ Norman remote to allow me to attempt to hack it.  My thinking was that I would figure out the button wiring and add some relays to allow me to automate the button press sequence.

I noticed that the remote has a Micro USB port on the side.  My thinking was that I could hack the serial interface to allow me to send commands to the shutters and I wouldn’t need to hack the remote buttons after all.  I later learned that this port was used to interface with the Norman eTimer app.  This is a browser-based app that lets you set up automated adjustments to the blinds at certain times during the day.  Ah, maybe I could reverse engineer the protocol between the app and the remote to send my own “Just in time” command – nope!  Turns out that the eTimer app allows you to set a schedule and then upload it to the remote for off-line batch processing.  It’s certainly a nice feature for anyone who does not have an integrated home automation system, but did not suit my needs.

So, back to the original plan of hacking the buttons.  I popped off the case and began the process of tracing the capacitive touch buttons.  I discovered the whole remote runs off of a PIC24FJ64GB106 – a nice 16-bit CPU with the power needed to run the eTimer scheduler.  There was a nice ribbon cable connecting wires from the capacitive touch buttons directly to the IO pins of the PIC microcontroller, so it was pretty easy to figure out what wire was used for each command.  I’m especially proud of my soldering job connecting extension wires to each of the ribbon cable pins – this was my first time soldering such fine wires, and it turned out better than I expected – the liberal use of flux helped.

Remote 2 Remote 4

This lead me to my next challenge.  In order to conserve battery power, the remote has a “feature” that requires you to first shake it to “wake it up”, then press the buttons in sequence to  move the blinds.  It turns itself back off after 15 seconds – very annoying.  This switch turned out to be a small component soldered on the mother board with a tiny ball bearing inside a closed plastic box that closed a circuit between a microcontroller pin and ground (you can hear it rattling around inside when you shake the remote).  I ended up removing the component and soldering leads to the 2 connections in order to force the remote awake via a relay.

Now that I had all the wire leads soldered, I attempted to get it to actually move the blinds with it by touching wires together.  At first,this was not working at all.  So, I followed the steps in the Quick Start guide to remove the batteries in one shutter and then re-install to put it into “learn”mode.  I was able to tie it to the hacked remote and move the blind – great!  Uh oh – in the process of training the blind to work with the hacked remote, I found it no longer responds to the 2nd remote.  Wow, it would be bad if I couldn’t use 2 remotes at the same time (1 for automation, 1 for manual over ride).

My installer was clueless about connecting the 2nd remote, so I reached out directly to Norman via their Customer Care web site.  I filled out their online form for Customer Help and was happy when a response was sent within 24 hours.  Unfortunately, the response was not helpful – it instructed me to review the Quick Start guide to setup up the remote, and did not address my question about connecting 2 remotes at the same time.  I responded with clarification and got this response:

Hello,
Unfortunately, we do not work directly with end users but our retailers are fully equipped and willing to assist!  If you purchased from one of our dealers or retailers, please contact them for further product information.

Wow – really disappointing!  I responded immediately and explained that my installer was unhelpful, but have not received any further communication.   Not only was Norman unhelpful to my installers during the installation process, but they flat-out refuse to help me with (I would expect), a simple question.  I suspect the real answer is that their devices only work with a single remote and they’re just not wanting to upset me with more bad news.  “Ha!  We already have your money, now go away!”

During this testing, I found that the range on the remote is quite poor.  I have the Somfy remotes in the basement on a rack and they are able to easily trigger the blinds throughout the house.  For the Norman blinds to move, I needed to be within 25 feet with no obstructions.  I’ll need to think strategically about where I mount this hacked remote once it is all done – I don’t think I’ll be able to keep it in the basement with the Somfy remotes as I planned.  Perhaps extending the antenna will help (it helped for the Somfy)?  I see a test point on the PCB antenna, so I’ll solder a wire on and play with that to see if it helps.

One other drawback worth mentioning – the quiet motors have a surprising drawback:  it’s hard for me to tell when they’re moving.  I’ve found that sending move commands to groups or all at once often results in a slow response and sometimes what appears to be one-at-a-time movement of the blinds.  Because it is so quiet, I have several times gone to re-send the command only to find that the shutters are actually moving.

Conclusion

Here’s a quick summary of the reasons I don’t recommend Norman Shutters:

  1. Poor quality control in the factory
  2. Cheap parts that break easily
  3. Terrible customer service, and poor/zero training offered to installers
  4. Products are proprietary and don’t offer integration options
  5. Very long lead times for product delivery

I would add “bad design” for the motors, but it appears they’ve corrected that in the latest revision.

Currently, I’m still in the process of designing the circuit to hack the remote.  I’m thinking of using transistors instead of relays this time in order to reduce the footprint of the circuit (14 connections would be a lot of bulky relays).  I’ve got the eTimer working on the “good” (un-hacked) remote, so at least I’m able to do a bit of automation while I finish connecting this into the home automation system.

Advertisements

1 Comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s