PIC18F4550 PWM Issue

[aaronik19]

Dear All,

I am designing a lighting controller 0-10V and I made the setup (HF Ballast and T5 lamp). I also interfaced the PIC18F4550 and the LM358 op-amp circuit to convert the PWM signals to 0-10V. The circuit can be found here:

http://forums.parallax.com/showthread.php/123572-0-10v-lighting-control

when I powered the circuit the voltage on the DMM started to go up slowly. I suspected that something is wrong. I remove the green cable (between the PIC Micro and 10K input resistor) and the voltage still continues to rise until reaches 10V! I started to test the code on the demo compiler of CCS 5.05 recently released. I made the code that on start-up the signal will immediately goes to ~6V (dims the ballast to 60%). But still no luck! Strictly speaking this is my first time that I am using PWM and I am very looking forwad to work with this circuit.

The photos can be found: http://share.pho.to/2jMRx

One idea/suggestion, to dim 0-10V over long distance, can I use Digital Potentiometers or I2C DACs? Some forums say it is not ideal and some says that PWM configuration is not a good idea. Really thanks for your reply.

My Main Program is:

[temtronic]

stuff to consider...
1) without a 'bleed' resistor across the 1uf input cap of the opamp, and PIC not connected the op amp will eventually 'saturate',going 'full on'.

2) what is a 'long distance'? inches, feet, a mile or two? If more than say 2 feet, consider using a second PIC( master-slave) similar to DMX512 protocol(which was designed for this operation !)

hth
jay

[aaronik19]

I can't figure out where is the problem. Do someone has a circuit which I can try?

[PCM programmer]

[aaronik19]

Dear Temtronic, the distance could be 100-200 meters long. How come you mixed DMX with 0-10V dimming techniques? DMX is based on RS485 hardware and from my knowledge that was the reason that it is more capable for longer distances? Do you have any circuit which you can explain better?

[aaronik19]

PCM Programmer, thanks. It worked fine. I noticed that I had the clock parameters not good. When I replaced that line, it worked perfect as wanted. The last issue is the hardware because I needed for long distance. Is it possible to solve the problem by using Mosfet technology?

[temtronic]

The 'slave' PIC would receive a command from the 'master' PIC and convert it into the 0-10V signal your lamp/ballast unit requires.
I said 'DMX' as it's a well known standard for remote lighting control.
Any serial communications 'protocol' would work for 200m. Current loop, RS232,RS485,DMX,etc. though you have to work out the details of speed vs distance.
As for the 10V control, depending on the resolution(# of bits/step), you could use a PIC port and make your own DAC with R-2R ladder network.

How you proceed depends upon your skill level, design requirements,etc.

hth
jay

[aaronik19]

Thanks for your reply. But when pm is better than dac and vice versa regarding 0-10v dimming?

[Mike Walne]

[aaronik19]

thanks to all of you guys. at the moment i am going to stick with the op-amp technique. i would like to add the feature of "Fade-In Time". for example i want to set_pwm_percent(400) witin 5 seconds.

First I tried to vary the frequency using the forumlae T=1/f; set_pwm_frequency(). Make a routine to keep this frequency for 5 seconds, but the still did not work. Do someone has any idea how I can implement this?

[oxo]

You need to fade in by changing the duty cycle, not the frequency.

The RC filter you have in the opamp solution is a fixed value, and will not work properly with changing frequency.

[aaronik19]

thanks. do you mean that I need to use the set_duty_cycle() before set_pwm_percent()? At the moment I can't recognize it...sorry

[oxo]

[aaronik19]

Ok now we are on the same line...

Yes i am using duty cycle to dim and bright. So far so good. Now my problem is that i need to make a sort of ramp-up or ramp-down. For example bright from 58% to 80% in 5seconds.

[oxo]