💬 Roller Shutter Node

Takero

It is possible that you dont use the asc for the autocalibration yet? I cant find it in your Code .... I found only the read process.

Corect me if i am wrong. :-)

scalz

acs read can set an endstop bool = "event" which trig the shutter state machine. you should see it ;)
in case you have some trouble, then you can set few default define in sketch to set your travel time. perhaps a cleareeprom needed, then shutter should be ready.

Takero

@scalz

Yes, i work on it . . . still my ACS wont work well. At the moment i init the node with the Endstop Bool, that works fine.

I work also on an delay between the UP/DOWN States, i don't like it if the shutter change the direktion directly. But . . . not find the right way yet :-)

EDIT: ACS works right now!!! :+1: I think my IC was defective :-1:

Takero

Dont know why but while autocali. they stop messurement of current . . .

If i start the process the first is go UP, my messurment of power is: 30W
if i Power of the attached Bulb the value stays on 30W, so it is normal he cant detect anythink

In your loop:

 // Read current sensor, endstop
  uint16_t acsread = 0;  
  acsread = readAdcWithFilter(ACS712_SENSOR, PS_64); 
  if (acsread < ACS712_LEVELDETECT) {
      // Endstop detected  
      if (getCalibrationState() > 0) setEndStopState(true); 
      else {};
  }

I modified for debug like this:

// Read current sensor, endstop
  uint16_t acsread = 0;  
  acsread = readAdcWithFilter(ACS712_SENSOR, PS_64); 


  Serial.print(F("ACS712: READ DEBUG: ")); 
  Serial.println(acsread);


  if (acsread < ACS712_LEVELDETECT) {
      // Endstop detected  
      if (getCalibrationState() > 0) setEndStopState(true); 
      else {};
  }

It never change the value of acsread after the init process is started. . .

LOG:

TSP:MSG:READ 0-0-50 s=2,c=1,t=2,pt=0,l=1,sg=1:1
CMD: Autocalibration
ACS712: READ DEBUG: 7
ACS712: READ DEBUG: 29
ACS712: READ: 29
TSP:MSG:SEND 50-50-0-0 s=6,c=1,t=17,pt=3,l=2,sg=0,ft=0,st=ok:29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29
ACS712: READ DEBUG: 29

29 will stay until timout stop. . . between this i power of the attached test bulb (20W)

Takero

Level detection work right now . . .

my modifications:

// send % progression for controller widgets NOT TESTED  
  static uint8_t oldpos = getPosition();
  
  if (oldpos != getPosition() && getCalibrationState()==0) {
    send(msgShutterPosition.set(getPosition()));
    send(msgPercent.set(getPosition()));
    oldpos = getPosition();
  }

Takero

I am finished and everythink is working well right now.

I dit some modifications on PCB:

add an 470uF near to NRF24
add two Diodes recovery diode (or free-wheeling diode) dunno how it called in english :-) at the relays

And some changes in your example code :+1:

Thanks for the Layout, thanks for the code, thanks for help!

I design a 3D Case, but only the bottom side. . . but not tested right now . .

alt text

0_1478263225360_RollerShutter_back.stl
0_1478263256093_RollerShutter_Back.123dx

Fabien

Did you try ACS measurement for up AND down ? On my 2 boards, i have problem with relay just over ACS and errors in current measurment (magnetic field of relay just over Hall sensor)
@scalz : I think we could try to turn 90° acs712, it could be fine.

Takero

With a Little Bit Trouble yes. The Problem is the Dimension of the amplifier. Below 60W the messured voltage is to Low.

I think it is netter the Diode must be in the Feedback cycle of the amplifier. But in my case the Free running voltage is below 6 and with running Roller above 20. So it is possible to detect the End stops. Just modify the variables.

I will Test it in the"real" System tomorow. If it Works i will post the Sketch.

scalz

@Fabien
i will take a look asap.. i'm actually playing with my nano32, and some design ;)
perhaps for the atsam version when more time to release..

@Takero
thx for feedback :+1:

i'll look to your suggestions too. for the wheeling diode if i remember it's integrated in the relay driver. depends what you mean.
For sketch, hehe, i warned it was not complete :) but the most important there. I'll look at your change and update it, sure !i have actually no time to play with it, still using an old sketch.

too much coding to do, and hardware to play! but this is so fun :)

see you soon

Takero

@scalz said:

for the wheeling diode if i remember it's integrated in the relay driver. depends what you mean

It isn't :-)

scalz

well, it is a zener, not wheeling diode, which should be enough..
http://www.onsemi.com/pub_link/Collateral/MDC3105-D.PDF
perhaps not for nrf! cool if it improves.

sidenote: i won't change nrf version in future. as i only have rfm69 (doing nrf was bonus), i don't want to waste my time improving both! and i have enough pcb ahah

Takero

Dont Care :-) its cool. Zenner its not enoth but dont worry. I need it .. the nrf is verry "female" :-D (sensible). But i think for both version, you have to modify the current messurment. No Roller shutter has >200w so u are on the Limit below 100w of adc scale.

But hey it Works and your Layout is verry cool and Safe.

Fabien

Very stange. Zener in forward side is equivalent to classical diode.
@Takero, you change the amplifier ? Wich mods did you make ?

Takero

I was a little bit bussy the last days.
My Nodes are completed installed right now, it works well. BUT, i had to change some thinks.

I had the Probleme of automatic reset of the device if i set the shutter to 50% or anythink else 0% or 100%.
The Diode alone dont solve these Probleme. I think the Problem is to stop the shutter while running. If i set it to 100% or 0% the shutter shut of before the relay falls down.

I solve the Problem with a 220uF Capacitor at the 3.3V Out and deactivate the BOD.

My Fuses right now: L: CF // H:DA // E:07

Now it works very well!!

@Fabien: I dont change the amp . . i just tested some thinks but it wont work . . for i dont use the current messurment and setup the Node with the manual endstop. It works realy good!

Finaly again, Thanks @scalz

The Temp. of the devices is below 30°C ;-)

In the picture the green temp.

fabiancrg

Hi
I was wondering if the ACS712 can detect the current if the roller shutter power is 85W?

kduino

Hi, there is something similar available although without an ac712. So you can't check the percentage o the shutter.
See https://github.com/JanGoe/esp8266-wifi-relay
I had one at home and it is working well. Because of the esp8266 it is directly connected to any controller using udp or MQTT or other protocols.
The only problem is the missing box. I have to admit I wouldn't put this in my wall without a proper box using AC 240 V!!!
The project you proposed is of course the most advanced. Do you think there is a smaller version available perhaps less parts on the BOM and easy to build?
Thank you very much, kduino

Fabien

Cloudy/rainning day ... Good for MySensors

Fabien

There is a mistake on the above photo. Radios are on the wrong way (180°)

Fabien

@scalz, little mistake on BOM, C4 is 100n AND 10n ..

Sergio Rodrigues

Hi, I make this project, pretty cool.

How I can upload the firmware? What I need to do?

💬 Roller Shutter Node

24

11.7k

11.2k

113.1k