Homebrew, open source, repurposed, hacked, software defined, open hardware

Saturday, 21 May 2011

The roller shutter controller

Here are some photos of the roller shutter controllers.
A torx security screwdriver and you're off and away.
Note the 12 x AA x 1.2V NiMH battery pack.
The battery pack is a nominal 14.4V 1500mAH

These battery packs get very hot when charging....
This might explain why so many have died....
I rang the factory, which is local, but they never
called back about replacement battery packs.
Oh well, that has motivated all this. Hopefully this will
help others confronting this problem, like VK5FDGW.
The unregulated 28V DC wall charger did not impress
me either.

I have seen replacement battery packs on e-bay for
about AUD$80
I have seen AA 1.2V NiMH tagged cells for AUD$4 in
quantity, which are 2000mAH.
So, I could buy a new battery pack, for AUD$80, and
wait for it to die again...
Or, I could make a battery pack, not a simple thing,
as it has to shoe horn back into the controller and I'd
need to solder or spot weld 11 pairs of tags, and it
would cost at least AUD$48 for the AA cells, and then
having done this, wait for it to die again.

Five dead controller battery packs starts to add up to
real money.

So, the next step was to try the trusty bench power supply.


The shutters are fairly large, up to 5 m^2 per roller shutter, which
was the maximum allowed at the time we bought them.
I removed the battery pack, and attached the leads to the socket
pins which the battery pack attaches to. The battery leads are
red and black, which should be observed when attaching an
external power supply.
As the roller shutter starts to go up, it seems in excess of 2.5A are
drawn at roughly 14V DC.
The controller refuses to work if the input voltage is in excess of
19V or so.
The next step was to try an SLA battery, plugged into the controller.
The following battery is a 1.3AH 12V battery which sells for
AUD$18 in quantity.

The battery is being topped up here, after playing with the roller
shutters. It needs needs no more than 130mA to top it up, for
constant voltage charging.
The battery is adequate for lifting and lowering the roller shutters.
The roller shutters only need 1.5A or less, when being lowered.

So, five AUD$18 batteries is sounding much more reasonable.

I suspect I could replace the controllers with simple switches
to move the shutters up and down, but I reckon a simple fly
lead to the nearby (currently) blank wall plate for power will
allow me to hide the battery in the wall further down. There is
a spare cat5 cable behind the blank wall plate which will be
used to float charge the battery daily.

SLA charging is easy, and I will knock up a current limited
fixed voltage charging circuit next, delivering about 130mA.
This should keep the batteries happy and healthy for many
years, and should be more reliable than the NiMH batteries.

The roller shutters are available with AC powered controllers
but we are in a bushfire area and want them to work when
AC power is not available. Also, we are on off grid solar and
don't want phantom loads overnight.

Some inspection of the circuit board revealed some power
conditioning for the unregulated charger input, dropping it to
about 15V with a switched mode circuit, which charges the
onboard battery via a diode at around 160mA.
After this diode is the common 15V bus.
The 15V then moves on to a high efficiency switched mode
power supply chip LP2951 which steps it down to around 5V
for the ATMEGA48.
There is space on the circuit board and some antenna
trackwork for 433MHz receiver bits and pieces for the
remotely actuated version which is sold.
There is also the driver circuitry for the roller shutter
motors, and some current sensing for the roller shutter
current, and some test points.








One down side of using an external SLA battery is that
the controller decides that the battery needs charging, and
it seems the ATMEGA48 has some kind of logic for starting
and completing charging, however broken. The effect of
this is an infrequent, regular flash of a red LED, visible from
the front of the controller.

Feeling too lazy to reverse engineer the ATMEGA firmware,
I plan to cut the single track going to the red LED. Ignorance
will be bliss.

The other thing to do will be splicing in some wiring that will
also go down the fly lead which will allow the controller to be told
to open or shut the roller shutter remotely, via the spare cat5
cable conductors. There will be plenty of space in the controller
without a battery pack if another microcontroller is needed.

Then, just add a web interface and an android phone....

Hmm, I wonder how the warranty will be affected....

1 comment:

  1. Hello,

    I am about to get these installed and would like to avoid the controllers completely. Do you have any thoughts on whether this will work with a 12V DC 10amp power supply and a wall mounted switch? Any help would be much appreciated, cheers

    Anton

    ReplyDelete