Thinking of a new light ..

skunkd0c

Well-Known Member
sorry if this is a silly question
but i was wondering can these led chips
use the same kind of liquid cooling that would be used for CPU's
never seen anyone do this wondered why lol

peace
 

stardustsailor

Well-Known Member
sorry if this is a silly question
but i was wondering can these led chips
use the same kind of liquid cooling that would be used for CPU's
never seen anyone do this wondered why lol

peace
I do not see why not ....
Although...
Personally ,I do not like Water/liquids and electricity close together..
But that's more of a 'religious' thing of mine ..(Amongst other,have to admit ....Like the blue leds ! LOL! )
 

Positivity

Well-Known Member
Just kidding sds..much respect..you got skills no doubt. I'm always looking at your stuff and wondering what I'd like to try. The calculations, small parts, breadboards...pretty much foreign to me. But I will for sure look into how these go together out of curiosity
 

skunkd0c

Well-Known Member
i was thinking that in the computer world cpu liquid cooling is seen as better than
fan + heatsink

i guess it would work out more expensive on larger light arrays more than one cooling unit might be needed or maybe larger cooling units designed for larger servers

keeping the leds as cool as possible would allow you to run them at their full rated power without worrying so much
about shorting their lifespan

when i look at these modular led lights they look some much like plug n play modular computer components anyway
pretty sure the cooling designed for cpu's would work with led chips , they are similar size
although cpu's will put out much more heat in that small size

the water electricity thing would be less of a worry since led is low voltage low heat to start with, folk be using these to cool their cpus for many years
i do take your point though water and electricity does put some folk off at the mere thought of it
i think cooling hps with water would be a worrying thing for me that would be more complex

peace
 

stardustsailor

Well-Known Member
DIY king..lol. Seriously...cool project. You know you want a remote control..
I'm after the ultimate led light ...
The ultimate grow -light,as I 've imagined it ...
And never going to get it !
But for what's worth ,I've to,at least,try to build it or something enough close to it....



With these arrays /COBS ,things drift away from the 'multiple panel/diodes ' scheme ....
_Which has shown,by now, how efficient way to utilise leds it is ...

Now ,'steping backwards' to single point /source light fixtures ....
But ,now the 'set' is different ....


I want a led light ,that will last for quite long ,and be efficient enough in growing plants ....
And then I can sit back and see the leds evolving ...
Until ,the next ...design ....
 
Last edited:

stardustsailor

Well-Known Member
i was thinking that in the computer world cpu liquid cooling is seen as better than
fan + heatsink

i guess it would work out more expensive on larger light arrays more than one cooling unit might be needed or maybe larger cooling units designed for larger servers

keeping the leds as cool as possible would allow you to run them at their full rated power without worrying so much
about shorting their lifespan

when i look at these modular led lights they look some much like plug n play modular computer components anyway
pretty sure the cooling designed for cpu's would work with led chips , they are similar size
although cpu's will put out much more heat in that small size

the water electricity thing would be less of a worry since led is low voltage low heat to start with, folk be using these to cool their cpus for many years
i do take your point though water and electricity does put some folk off at the mere thought of it
i think cooling hps with water would be a worrying thing for me that would be more complex

peace
Well,for sure you ain't telling something wrong ...
And I agree with most of them ...
Maybe for more powerful diodes to come ,I should think about water cooling ..(or even liquid Nitrogen ...:-P...)

For now ,the good ol' fans and heatsink ,do the job fine ....
( Still... at least ...)
 

skunkd0c

Well-Known Member
i agree mate it maybe a bit overkill
in the nerdy pc world water cpu cooling its pretty standard

these cooling things for cpus are bought as modular devices not home built or anything
so perhaps in 10 years or so they will be offered by some of the led companies
on their turbo range lol

peace
 

stardustsailor

Well-Known Member
i agree mate it maybe a bit overkill
in the nerdy pc world water cpu cooling its pretty standard

these cooling things for cpus are bought as modular devices not home built or anything
so perhaps in 10 years or so they will be offered by some of the led companies
on their turbo range lol

peace
From what I 'm aware of ,already in some scientific experimental led lights ,water cooling was used sand in some other cases also....
And I think there are already on the market some greenhouse led lights that utilise water/liquid cooling ...
The liquid-cooling tech is already involved in the led world ..
Just not on home/small-scale grow lights...
 

skunkd0c

Well-Known Member
not surprised mate, thanks for the info ill look into that
folk will try anything to push their devices to the limits
for some reason i keep thinking of the dude that put the guts of a pc
in a glass tank full of cooking oil to cool it lol

peace
 

stardustsailor

Well-Known Member
Well ...
At least ,I 've a really fast and presice thermocouple thermometer for the panel ....
P4131008.JPG
( With a dreamy figure of Case temperature way under 25C :blsmoke:..... You wish,Stardust ....)


Since the Max chip uses already most of the analog inputs ,it would be 'risky' to place an amperemeter hall sensor board on one of the remaining analog inputs ....
It might introduce lots of signal noise..
And it will,definately, impact sampling frequency of both amperemeter and thermometer ...( 10Hz for now)
Since I can have an analog or digital amperemeter as a separate device ..
While I can't have an analog or digital temperature contol/ circuit breaker ..(not that easy or cheap,atleast ....)

Arduino handles only the thermal checking and protecting (via power cut-off relay) the CXAs from overheating ..
(Still,I've not programmed that far .... )
At least for present ...
Maybe another idea will come up ,that is


The arduino sketch :
//*****************************Led Grow light Cree CXA3070 Diagnostic Circuit Breaker *********************************
//
//
//
//LIB
#include <LiquidCrystal.h>
#include <MAX31855.h>
//
//DEFs
#define SENSOR_SAMPLING_TIME 100


// DEG SYMB FOR LCD *****
unsigned char degree[8] = {
140,146,146,140,128,128,128,128};


//PINs
int ssrPin = 5;
int thermocoupleSOPin = A3;
int thermocoupleCSPin = A2;
int thermocoupleCLKPin = A1;
int lcdRsPin = 7;
int lcdEPin = 8;
int lcdD4Pin = 9;
int lcdD5Pin = 10;
int lcdD6Pin = 11;
int lcdD7Pin = 12;
int ledRedPin = 4;
int buzzerPin = 6;
int switchPin = A0;

//VAR
double input;
unsigned long nextCheck;
unsigned long nextRead;

// Specify LCD interface
LiquidCrystal lcd(lcdRsPin, lcdEPin, lcdD4Pin, lcdD5Pin, lcdD6Pin, lcdD7Pin);
//TH.Couple Interface
MAX31855 thermocouple(thermocoupleSOPin, thermocoupleCSPin,thermocoupleCLKPin);

//SET-UP
void setup()
{
// SSR pin
digitalWrite(ssrPin, LOW);
pinMode(ssrPin, OUTPUT);

// Buzzer pin init / off
digitalWrite(buzzerPin, LOW);
pinMode(buzzerPin, OUTPUT);

// LED pins init/ on (active low)
digitalWrite(ledRedPin, LOW);
pinMode(ledRedPin, OUTPUT);

// Start-up splash
digitalWrite(buzzerPin, HIGH);
lcd.begin(8, 2);
lcd.createChar(0, degree);
lcd.clear();
lcd.print("CXA Diag");
lcd.setCursor(0, 1);
lcd.print("Loading");
digitalWrite(buzzerPin, LOW);
delay(2500);
lcd.clear();
// Turn off LED (active low)
digitalWrite(ledRedPin, HIGH);
// Initialize time keeping variable
nextCheck = millis();
// Initialize thermocouple reading variable
nextRead = millis();
}

//LOOP
void loop()
{
// Current time
unsigned long now;

// Time to read thermocouple
if (millis() > nextRead)
{
// Read thermocouple next sampling period
nextRead += SENSOR_SAMPLING_TIME;
input = thermocouple.readThermocouple(CELSIUS);
// If thermocouple problem detected

if((input == FAULT_OPEN) || (input == FAULT_SHORT_GND) ||
(input == FAULT_SHORT_VCC))
{ lcd.setCursor(1, 0);
lcd.print("! Sensor");
lcd.setCursor(0, 1);
lcd.print("Error !");
digitalWrite(buzzerPin, HIGH);
delay(500);
digitalWrite(buzzerPin, LOW);
}
else
if (millis() > nextCheck)
{
// Check input in the next 100 millisecs
nextCheck += 100;
{
// Print current temperature
digitalWrite(buzzerPin, LOW);
lcd.clear();
lcd.setCursor(1, 0);
lcd.print("CXA Tc:");
lcd.setCursor(0, 1);
lcd.print(input);

#if ARDUINO >= 100
// Print degree Celsius symbol
lcd.write((uint8_t)0);
#else
// Print degree Celsius symbol
lcd.print(0, BYTE);
#endif
lcd.print("C ");
}
}
}
}
So,I need a simple ,good ol' fashioned 2A amperemeter ,(analog,my favorite ! ),
for displaying the driving current at the CXAs..
 
Last edited:

stardustsailor

Well-Known Member
Finished a simple programming ...
Once Tc goes over 95°C ,arduino will discontinue to power the relay ,that keeps the drivers switched on ...
(or it can be that the arduino enables a 'closed' relay ,and opens it ..
Thus cutting off power from drivers ...Either way...Small mod to software ...)
Once the Tc drops below 95 C ,then drivers are powered up again,switching on the CXAs ...

Now ..I need a 5 VDC / 240 VAC 6 A relay ...
//*****************************Led Grow light Cree CXA3070 Diagnostic Circuit Breaker *********************************
//
//
//
//LIB
#include <LiquidCrystal.h>
#include <MAX31855.h>
//
//DEFs
#define SENSOR_SAMPLING_TIME 500 // 2 samples per sec
// DEG SYMB FOR LCD *****
unsigned char degree[8] = {
140,146,146,140,128,128,128,128};
//PINs
int relayPin = 5;
int thermocoupleSOPin = A3;
int thermocoupleCSPin = A2;
int thermocoupleCLKPin = A1;
int lcdRsPin = 7;
int lcdEPin = 8;
int lcdD4Pin = 9;
int lcdD5Pin = 10;
int lcdD6Pin = 11;
int lcdD7Pin = 12;
int ledRedPin = 4;
int buzzerPin = 6;
int switchPin = A0;
//VAR
double input;
unsigned long nextCheck;
unsigned long nextRead;
// Specify LCD interface
LiquidCrystal lcd(lcdRsPin, lcdEPin, lcdD4Pin, lcdD5Pin, lcdD6Pin, lcdD7Pin);
//TH.Couple Interface
MAX31855 thermocouple(thermocoupleSOPin, thermocoupleCSPin,thermocoupleCLKPin);
//SET-UP
void setup()
{
// Buzzer pin init / off
digitalWrite(buzzerPin, LOW);
pinMode(buzzerPin, OUTPUT);
// LED pins init/ on (active low)
digitalWrite(ledRedPin, LOW);
pinMode(ledRedPin, OUTPUT);
//relay pin
digitalWrite(relayPin,HIGH);
pinMode(relayPin, OUTPUT);
// Start-up splash
digitalWrite(buzzerPin, HIGH);
lcd.begin(8, 2);
lcd.createChar(0, degree);
lcd.clear();
lcd.print("CXA Diag");
lcd.setCursor(0, 1);
lcd.print("Loading");
digitalWrite(buzzerPin, LOW);
delay(2500);
lcd.clear();
// Turn off LED (active low)
digitalWrite(ledRedPin, HIGH);
// Initialize time keeping variable
nextCheck = millis();
// Initialize thermocouple reading variable
nextRead = millis();
}

//LOOP
void loop()
{
// Current time
unsigned long now;

// Time to read thermocouple
if (millis() > nextRead)
{
// Read thermocouple next sampling period
nextRead += SENSOR_SAMPLING_TIME;
input = thermocouple.readThermocouple(CELSIUS);
// If thermocouple problem detected

if((input == FAULT_OPEN) || (input == FAULT_SHORT_GND) ||
(input == FAULT_SHORT_VCC))
{ lcd.setCursor(1, 0);
lcd.print("! Sensor");
lcd.setCursor(0, 1);
lcd.print("Error !");
digitalWrite(buzzerPin, HIGH);
delay(500);
digitalWrite(buzzerPin, LOW);
}
else
if (millis() > nextCheck)
{
// Check input in the next500milsec
nextCheck += 500;
if ((input) <= 85 )
{
digitalWrite(relayPin,HIGH);
// Turn off LED (active low)
digitalWrite(ledRedPin,HIGH);
// Print current temperature
digitalWrite(buzzerPin, LOW);
lcd.clear();
lcd.setCursor(1, 0);
lcd.print("CXA Tc:");
lcd.setCursor(0, 1);
lcd.print(input);
// Print degree Celsius symbol

lcd.print(0,(uint8_t)0);
lcd.print("C ");
}
else
if ((input) >= 95 )
{
digitalWrite(relayPin, LOW);
lcd.clear();
lcd.setCursor(1, 0);
lcd.print("CAUTION:");
lcd.setCursor(0, 1);
lcd.print("CXA HOT!");
delay (1000);
lcd.clear();
lcd.setCursor(1, 0);
lcd.print("PWR OFF");
lcd.setCursor(0, 1);
lcd.print("PROTECT!");
delay (1000);
// Turn on LED (active low)
digitalWrite(ledRedPin,LOW);
digitalWrite(buzzerPin, HIGH);
delay(250);
digitalWrite(buzzerPin, LOW);
delay(2500);
}

}
}
}
P4131013.JPG

P4131018.JPG

P4131020.JPG
 
Last edited:

CannaBare

Well-Known Member
Liking what is going on here! I've recently bought an arduino to fiddle with. I want to automate my watering system for 16 hempy buckets in a 4x4. The DIY with arduino could not be easier!
 

SNEAKYp

Well-Known Member
I like the complexity of your builds. I can't wait to move into a place with a garage and have a real workspace to do DIY projects. I saw a thread recently that was using water cooling systems somewhere I'm too stoned to find the website at the moment. The guy had a couple 3 or so cobs on one system and I remember the water temp was sitting in the the 90's.
 

stardustsailor

Well-Known Member
Ok...
Arduino will be a 'dedicated thermal active protection'....
It will control a Solid State Relay ,which will switch on/off the AC power to the four drivers ..
Of course it is going to be installed a 'protection by-pass switch " at the AC line before the SSR ,also ...

Once T case gets above 95C ,CXAs will switch off ...

I have a SSR that is normally open ( NO ) ..
That means ,that it needs a DC signal ,to let AC pass ....

Either Arduino has to constantly output the 'on ' signal ..
(like it is programmed at present ) and switch off the relay ,if Tc=>95C ...
(Simpler solution ,but needs the 'safety protection by-pass switch,in case of Arduino failing,as the leds will switch off ...)

Or Arduino outputs no signal -at normal operation - to an optocoupler inverter circuit ,which keeps the SSR constantly on ..
If Tc gets above 95 ,then arduino outputs high ,which the inverter ,inverts to low at the SSR ...So the Relay opens and drivers switch off ...
(At this case if arduino fails ,thermal protection fails but leds still will be operating ....
The SSR is going to be powered on ,by the Fan DC PSU-throught the inverter ..
.Drivers will switch off ,only if that Fan PSU fails ....)

Third case(simplest and most expensive ) : A normally closed SSR ,which opens from Arduino signal only ...
(At this case the drivers will stay on ,even if arduino fails ,or the Fan PSU ,it doesn't matter..)



Apologies for the really bad quality of the vid...
Crappy mob camera..


....it went up to 984°C !!!
CXA has kissed it's @$$ 'goodbye' ....LOL!!! :fire:

P4131026.JPG
 
Last edited:

stardustsailor

Well-Known Member
I think I will go for the first option ....

AC line will enter through an EMI Filter ,to a resettable circuit breaker ,then to the
"Power Switch " ...(all of them ,placed at the backside of the led panel )

From there it will split into three ways ...
(Inside the panel ,now...)

1) AC line to the 15 volt Fan PSU ( without on/off switch )=>14V DC to PWM controller=> PWM 13.5 VDC to fans.

2) AC line to small 12 VDC psu( without on/off switch ) => Arduino => Sensing of CXA /control of SSR.

3) AC line to Single pole / Double Throw switch :
-One line goes directly to led drivers (thermal protection By-pass )
-The other to the SSR input .(Thermal protection enabled ) .From SSR output ,to the led drivers.

I can program the shield's Switch button ,so to manually switch on/off the SSR ,thus the CXAs,from the front of the panel ....

So in the front of the panel will be placed :
- 3 digit 7-seg digital voltmeter (blue light ) ,fan pwm voltage indication
- fan speed/ led power single one-turn adjust pot ,(adjusts both at same time ,linear )
- analog amperemeter 2A
-2x8 blue lcd display with alerts and Tcase indication
- toggle switch for manually switching of the led drivers/ CXAs ,only
-arduino reset button (maybe)
-red/or blue indication led (maybe)
-'Spectrum Shift' toggle switch .
 

stardustsailor

Well-Known Member
Well ..I've made a quick (and way risky -dangerous-deadly !!!! ) 'rig' on the floor ,
to check the microcontroller thermal protection of the CXA's ....
I've tried some resistive loads ( a toaster,a hand-coctail mixer,a led desk lamp ) and it worked fine !!!

I've tried a fan also,as an inductive load ...Absoloutely flawless operation ...

And I've asked around once more ...
'who wants to try next ? '

"Me-me-me ,please sir ,Me -please,much please ' said a small asian cc driver ...

'Ok,hop on board ,I replied ...'

As soon as i plugged the driver ,the noise of it ,really distort the temperature reading ...
From ~18 C it went to -2 C !!! (for a moment ....) then to 14C ,back to 8C ..Whatever ...
Too much noise ...

Guod,any thoughts here ?


Then ,I tested the system ...
See for yourselves ..

Flashing-blinking when powered off ......:eyesmoke:
Without even any PWM signal ...:-?
Free energy ...!!!! :dunce:
From nowhere !!
Automatic blinking feature in faulty conditions!!! :eyesmoke:


Those asians ...They 're light years ahead in tech,from the rest of us ...:dunce:

:wall::wall::wall:......



 
Top