PWMSlope Behavior

[icecool2]

I'm having some issues with the PWMSlope function. If you enter into the function inside of the time window for making a change, the function doesn't seem to work. For example, if I restart my controller at 6:00p and it is set to fade out from 5:00p to 7:00p, the PWMSlope function only returns the top value rather than the current value it should be at based on the time.

Maybe I'm having a strange problem... Here's my code:

Code: Select all

ReefAngel.PWM.SetActinic(PWMSlope(11,0,20,0,100,70,120,ReefAngel.PWM.GetActinicValue()));

[rimai]

The PWMSlope function does not work with inverted values.
The parameters you are using is for a slope from 11:00 to 20:00 going from 100% to 70% within 120 minutes.
Is this what you had in mind? Inverted slope?
If this was not what you wanted, here is the correct way of using the slope function:

Code: Select all

ReefAngel.PWM.SetActinic(PWMSlope(11,0,20,0,70,100,120,ReefAngel.PWM.GetActinicValue()));

[icecool2]

That could explain the problem. My LEDs run inverted to the default values. 100 = off, 0 = on.

[rimai]

Try this:

Code: Select all

ReefAngel.PWM.SetActinic(100-PWMSlope(11,0,20,0,0,30,120,ReefAngel.PWM.GetActinicValue()));

Let me know if this works for you.

[icecool2]

I need to spend some more time looking at this, but I'm seeing a strange behavior with that. One iteration of the loop returns the correct value, a second iteration of the loop returns only the slope value and then that repeats.

For example:
Actinic: 97
Actinic: 3
Actinic: 97
Actinic: 3

I'm pretty sure I wiped out everything that could be overriding the SetActinic call, but I will need to double check that again.

[rimai]

Hum...
I remember now.
You are correct.
You have to create a new variable to hold value, otherwise the GetActinicValue will keep doing that.
This is what you need:
Above the setup() section:

Code: Select all

byte BuckPuckActinic=0;

In the loop() section:

Code: Select all

      BuckPuckActinic=PWMSlope(11,0,20,0,0,30,120,BuckPuckActinic); 
      ReefAngel.PWM.SetActinic(100-BuckPuckActinic);

Buckpucks use the same PWM values... 100=off and 0=on
I made this code for nanotrevor, who uses buckpucks and needed inverted values too:

Code: Select all

// Autogenerated file by RAGen (v1.0.4.92), (06/03/2011 21:13)
// RA_060311_2113.pde
//
// This version designed for v0.8.5 Beta 12 or later

/* The following features are enabled for this PDE File: 
#define DisplayImages
#define DateTimeSetup
#define VersionMenu
#define ATOSetup
#define DirectTempSensor
#define DisplayLEDPWM
#define SingleATOSetup
#define StandardLightSetup
*/

#include <ReefAngel_Colors.h>
#include <ReefAngel_CustomColors.h>
#include <ReefAngel_Features.h>
#include <ReefAngel_Globals.h>
#include <ReefAngel_Wifi.h>
#include <Wire.h>
#include <OneWire.h>
#include <Time.h>
#include <DS1307RTC.h>
#include <ReefAngel_EEPROM.h>
#include <ReefAngel_NokiaLCD.h>
#include <ReefAngel_ATO.h>
#include <ReefAngel_Joystick.h>
#include <ReefAngel_LED.h>
#include <ReefAngel_TempSensor.h>
#include <ReefAngel_Relay.h>
#include <ReefAngel_PWM.h>
#include <ReefAngel_Timer.h>
#include <ReefAngel_Memory.h>
#include <ReefAngel.h>

byte BuckPuckActinic=0;
byte BuckPuckDaylight=0;

void setup()
{
    ReefAngel.Init();  //Initialize controller

    ReefAngel.FeedingModePorts = B00000001; //turns off return pump
    ReefAngel.WaterChangePorts = B01110001; //turns off return pump, heater, fan/chiller and ATO
    ReefAngel.OverheatShutoffPorts = B00101110; //turns off lights and heater
    ReefAngel.LightsOnPorts = B00001110; //turns on lights

    // Ports that are always on
    ReefAngel.Relay.On(Port1);
    ReefAngel.Relay.On(Port8);   
}

void loop()
{
    ReefAngel.ShowInterface();
    if (hour(now())<12)
      BuckPuckActinic=PWMSlope(7,0,9,0,0,60,60,BuckPuckActinic); //actinic leds ramp up from 0% at 7:30am to 60% at 8:30am and ramp down from 6:00pm to 7:00pm back to 0%
    else
      BuckPuckActinic=PWMSlope(19,0,20,0,0,60,60,BuckPuckActinic); //actinic leds ramp up from 0% at 7:30am to 60% at 8:30am and ramp down from 6:00pm to 7:00pm back to 0%
        

    BuckPuckDaylight=PWMSlope(8,30,18,0,0,60,60,BuckPuckDaylight); //daylight leds ramp up from 0% at 8:30am to 50% at 9:30am and ramp down from 5:00pm to 6:0pm back to 0%
    ReefAngel.PWM.SetActinic(100-BuckPuckActinic); // invert value for buckpucks
    ReefAngel.PWM.SetDaylight(100-BuckPuckDaylight); // invert value for buckpucks
   
       // Specific functions
    ReefAngel.StandardLights(Port2); 
    ReefAngel.StandardLights(Port3);
    ReefAngel.StandardLights(Port4);
    ReefAngel.StandardFan(Port5);
    ReefAngel.StandardHeater(Port6);
    ReefAngel.SingleATOHigh(Port7);
  
}

[icecool2]

I'll give that a shot this evening and report back.

[NanoTrevor]

hey ice, that should set you up, i run pucks also and roberto gave me the inverted code and they work perfect. only issue is that the values displayed will be inverted, if you have code at like 70% the display will read 30%, what binder said i could do is go into the reefangel_nokia cpp and towards the bottom put the 100- in front on the display pwm line stuff
looks like this;
DrawText(DPColor,DefaultBGColor,x+60,y+10,"DP:");
DrawSingleMonitor(100-DaylightPWMValue, DPColor, x+78, y+10,1);
DrawText(APColor,DefaultBGColor,x+60,y+20,"AP:");
DrawSingleMonitor(100-ActinicPWMValue, APColor, x+78, y+20,1);

[icecool2]

Thanks. That code did work.

Trevor, thanks for the tip! I actually didn't realize those values were on the front panel

[NirvanaandTool]

Hmm. Bookmarking this when I actually get brave enough to wire my Buckpucks into my RA.
Did you guys just run Ref into + on the RA and - off the LED to - on the RA like in the Wiring guide for the Puck?

[rimai]

nanotrevor is having a hard time with pucks.
In my opinion, the way the pucks were designed is wrong, but that's just me. They may think the same thing of my controller design.
The issue he's having is exactly because of the - from RA connecting to LED string directly. The LED string is a high side of the driver and creates disturbunces and noise that is carried to the RA, causing it to lock-up or present weird behaviors.
I'm working with him on trying to isolate the 2 sides with opto-isolators, which should take care of his noise problem.

[icecool2]

Well, you can mark Recom drivers as compatible with your design. I do have some isolation built into my PCB, but nothing unusual that you wouldn't expect any novice to design in.

[NirvanaandTool]

nanotrevor is having a hard time with pucks.
In my opinion, the way the pucks were designed is wrong, but that's just me. They may think the same thing of my controller design.
The issue he's having is exactly because of the - from RA connecting to LED string directly. The LED string is a high side of the driver and creates disturbunces and noise that is carried to the RA, causing it to lock-up or present weird behaviors.
I'm working with him on trying to isolate the 2 sides with opto-isolators, which should take care of his noise problem.

Hmm gotcha. Would it be better to build that circuit they describe on the wiring diagram in order to have them run in a non-inverted state?

[rimai]

I don't think that would solve the issue. The biggest problem is connecting the LED- to RA ground.

[NanoTrevor]

I don't think that would solve the issue. The biggest problem is connecting the LED- to RA ground.

if its wired with a transistor there is no - led connection. i tried it, i have the parts but didn't have any luck, i wired the other way cause it was quick and worked with the code inversion...but this would prob solve all probs to wire this way

[rimai]

The picutre above would never work the way it is.
The signal LO/HI needs a ground reference, which isn't represented in the design.

[NanoTrevor]

The picutre above would never work the way it is.
The signal LO/HI needs a ground reference, which isn't represented in the design.

hmm, prob why it didn't work for me, i thought i wired it wrong somehow.

[rimai]

Well, you can mark Recom drivers as compatible with your design. I do have some isolation built into my PCB, but nothing unusual that you wouldn't expect any novice to design in.

Can you post your schematics?
I'm interested to know how you connected the RA GND to your drivers.
I'm starting to think I originally led nanotrevor into the wrong path

[NanoTrevor]

Well, you can mark Recom drivers as compatible with your design. I do have some isolation built into my PCB, but nothing unusual that you wouldn't expect any novice to design in.

Can you post your schematics?
I'm interested to know how you connected the RA GND to your drivers.
I'm starting to think I originally led nanotrevor into the wrong path

[rimai]

Sorry Trevor...
I've been reading lots of places that people just have these problems with the buckpucks. Not just you.
The solution always seems to be optoisolation, like we've been working on.
But, there was a couple of places where they mention connecting the Vin- to ground, instead of LED- to ground.
I don't know if that would make any difference on the whole issue, but it would be worth trying.

[NanoTrevor]

Sorry Trevor...
I've been reading lots of places that people just have these problems with the buckpucks. Not just you.
The solution always seems to be optoisolation, like we've been working on.
But, there was a couple of places where they mention connecting the Vin- to ground, instead of LED- to ground.
I don't know if that would make any difference on the whole issue, but it would be worth trying.

i actually tried that on another arduino based controller i had tried and it didn't work connecting to vin - instead of led -

[icecool2]

Sorry, I don't have a schematic handy. I have a current limiting resistor on the signal line from the RA to the driver. On the driver side, I have the digital ground separated from the analog ground and joined with a small bridge. There is also a weak pull up on the driver as well. I have the GND from the RA run to the digital ground of the drivers. The analog grounds are connected on the AC side between the RA and the power supplies for my drivers.

[NirvanaandTool]

Yikes. Sounds like problems all over.
Maybe I'll skip trying to get the Buckpucks to run off the PWM and get some RA moonlights instead.