Post by thread:Suggestions for AC chopper?

Hello all, This question is only marginally related to pic's since I am going to use a 16C84 to control the chopper, but there seem to be quite a few bright people out there who may be able to help... I am trying to make an incandescent lamp dimmer from a 16C84 and planned on doing a simple waveform chop using a triac. I realized, however, that since the pic is going to be doing a bunch of other time sensitive stuff that I don't have the processor time to manually sense the zero and chop the wave using the pic. So, is there any easy method to passively (from the pic's point of view) chop the AC wave? i.e. use 8 I/O pins to a D-A to give a reference to the chopper circuit and have it do the rest? That way I can just output the light level to the pins and it would stay put leaving the rest of the processor available? Thanks for any help, aaron ------------------------------------------------------------------------------- - Aaron J. Miller Email: spam_OUTmilleraTakeThisOuTstanford.edu - Physics Graduate Student Web : http://www.stanford.edu/~millera - Stanford University Phone: (415) 372-5575 (until 4/12/97) -------------------------------------------------------------------------------

part 0 2131 bytes Like you, I wanted to use the timer to keep track of other things. One thing I found was an I2C real time clock - Phillips PCF8593. You could set the RTC to give you an alarm interrupt, freeing up the timer. This was optimal for me because some of the delay times were very long. The application might want to wait an hour and then dim the light for 5 minutes. I drove the gate with a MOC3033 triac optoisolator. You can built a zero crossing detector from a comparitor to synchronize the the gate transitions with the zero crossings. If you want the chopping to be pasive to your PIC, why don't you use two PICs? Regards, Chris Atkins ---------- From: Aaron J. Miller Sent: Thursday, March 27, 1997 2:15 PM To: .....PICLISTKILLspam@spam@MITVMA.MIT.EDU Subject: Suggestions for AC chopper? Things are changing at NetForward! http://www.netforward.com/changes.shtml ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Hello all, This question is only marginally related to pic's since I am going to use a 16C84 to control the chopper, but there seem to be quite a few bright people out there who may be able to help... I am trying to make an incandescent lamp dimmer from a 16C84 and planned on doing a simple waveform chop using a triac. I realized, however, that since the pic is going to be doing a bunch of other time sensitive stuff that I don't have the processor time to manually sense the zero and chop the wave using the pic. So, is there any easy method to passively (from the pic's point of view) chop the AC wave? i.e. use 8 I/O pins to a D-A to give a reference to the chopper circuit and have it do the rest? That way I can just output the light level to the pins and it would stay put leaving the rest of the processor available? Thanks for any help, aaron ------------------------------------------------------------------------------- - Aaron J. Miller Email: milleraKILLspamstanford.edu - Physics Graduate Student Web : http://www.stanford.edu/~millera - Stanford University Phone: (415) 372-5575 (until 4/12/97) -------------------------------------------------------------------------------

Thanks for the response! > If you want the chopping to be pasive to your PIC, why don't you > use two PICs? I am thinking about it... I just took apart one those dimming halogen lamps and noticed that they do an elegant chopping thing with a RLC circuit somehow (I don't understand the circuit). I figured if there were an easy way to simulate the variable resistor in the circuit it would be nice. Any clues there? -aaron ------------------------------------------------------------------------------- - Aaron J. Miller Email: .....milleraKILLspam.....stanford.edu - Physics Graduate Student Web : http://www.stanford.edu/~millera - Stanford University Phone: (415) 372-5575 (until 4/12/97) -------------------------------------------------------------------------------

On Thu, 27 Mar 1997 13:15:44 -0800 "Aaron J. Miller" <EraseMEmilleraspam_OUTTakeThisOuTLELAND.STANFORD.EDU> writes: {Quote hidden}>Hello all, > >This question is only marginally related to pic's since I am going to >use >a 16C84 to control the chopper, but there seem to be quite a few >bright >people out there who may be able to help... > >I am trying to make an incandescent lamp dimmer from a 16C84 and >planned >on doing a simple waveform chop using a triac. I realized, however, >that >since the pic is going to be doing a bunch of other time sensitive >stuff >that I don't have the processor time to manually sense the zero and >chop >the wave using the pic. > >So, is there any easy method to passively (from the pic's point of >view) >chop the AC wave? i.e. use 8 I/O pins to a D-A to give a reference to >the >chopper circuit and have it do the rest? That way I can just output >the >light level to the pins and it would stay put leaving the rest of the >processor available? > We are using the 16c74 to phase control four triacs. We detect zero cross by doing a negative edge interrupt on the external interrupt pin. Positive zero cross is determined based on a timer. We determine when to turn on a triac by using the compare mode of a capture/compare timer register. At zerocross we determine when the first turn-on will be, then put it in the compare register. On receiving the interrupt from that compare, we turn on that triac and load the next. The external interrupt port pin is driven directly from the power supply transformer secondary with a series current limit resistor. The clamp diodes in the 16c74 square it nicely and the hysteresis on the input prevents noise problems. Note that we actually get the interrupt a little BEFORE the zero-cross, but that's a lot better than a lot after (which would be the case on a positive edge trigger). A 16c74 running at 16 MHz handles this quite nicely. While doing the phase control it is also receiving a 250 Kbps datastream and scanning 4 A/D inputs. If you want to generate triac firing externally, you can do it with an op-amp ramp generator (that is synced with the AC line) driving a comparator for each triac. You can get analog output of many of the PICs by using pwm into a simple low pass filter. We're doing something similar in a product where we take the 250 Kbps datastream and pull out 48 bytes, converting them to analog signals. Now... speaking of AC choppers... Anyone aware of a bunch of PWM generators (like those in PIC chips) but all in one chip (maybe a dozen of them or so that could be driven by SPI or I2C)? Harold

Aaron J. Miller wrote: {Quote hidden}> > Hello all, > > This question is only marginally related to pic's since I am going to use > a 16C84 to control the chopper, but there seem to be quite a few bright > people out there who may be able to help... > > I am trying to make an incandescent lamp dimmer from a 16C84 and planned > on doing a simple waveform chop using a triac. I realized, however, that > since the pic is going to be doing a bunch of other time sensitive stuff > that I don't have the processor time to manually sense the zero and chop > the wave using the pic. > > So, is there any easy method to passively (from the pic's point of view) > chop the AC wave? i.e. use 8 I/O pins to a D-A to give a reference to the > chopper circuit and have it do the rest? That way I can just output the > light level to the pins and it would stay put leaving the rest of the > processor available? > > Thanks for any help, > aaron > > ------------------------------------------------------------------------------- > - Aaron J. Miller Email: milleraspam_OUTstanford.edu > - Physics Graduate Student Web : http://www.stanford.edu/~millera > - Stanford University Phone: (415) 372-5575 (until 4/12/97) > ------------------------------------------------------------------------------- A diac and an optoisolator with cds out and a single pic pin PWM with a resistor /capacitor integrator infinitely variable and you need the same parts anyway.my $.02 Tony M.

I looked at doing this for a project. What really matters is that once the tirac is on, the gate goes low before the next zero crossing if you want it to go off. This can be done by setting up the timer to interrupt you whenever the output needs to change state. In the ISR to handle the timer interrupt, you would toggle the output and immediately schedual the next event. Like you, I wanted to use the timer to keep track of other things. One thing I found was an I2C real time clock - Phillips PCF8593. You could set the RTC to give you an alarm interrupt, freeing up the timer. This was optimal for me because some of the delay times were very long. The application might want to wait an hour and then dim the light for 5 minutes. I drove the gate with a MOC3033 triac optoisolator. You can built a zero crossing detector from a comparitor to synchronize the the gate transitions with the zero crossings. If you want the chopping to be pasive to your PIC, why don't you use two PICs? Regards, Chris Atkins ---------- From: Aaron J. Miller Sent: Thursday, March 27, 1997 2:15 PM To: @spam@PICLISTKILLspamMITVMA.MIT.EDU Subject: Suggestions for AC chopper? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Hello all, This question is only marginally related to pic's since I am going to use a 16C84 to control the chopper, but there seem to be quite a few bright people out there who may be able to help... I am trying to make an incandescent lamp dimmer from a 16C84 and planned on doing a simple waveform chop using a triac. I realized, however, that since the pic is going to be doing a bunch of other time sensitive stuff that I don't have the processor time to manually sense the zero and chop the wave using the pic. So, is there any easy method to passively (from the pic's point of view) chop the AC wave? i.e. use 8 I/O pins to a D-A to give a reference to the chopper circuit and have it do the rest? That way I can just output the light level to the pins and it would stay put leaving the rest of the processor available? Thanks for any help, aaron ------------------------------------------------------------------------------- - Aaron J. Miller Email: KILLspammilleraKILLspamstanford.edu - Physics Graduate Student Web : http://www.stanford.edu/~millera - Stanford University Phone: (415) 372-5575 (until 4/12/97) -------------------------------------------------------------------------------

part 0 587 bytes content-type:image/x-MS-bmp; name="POWER.BMP"Not exactly an ac chopper but it controls AC loads very well Take a look at this circuit it's opto Isolated and you drive the opto led with PWM at whatever frequence you program allows (within reason, 50Hz to 10KHz) It requires no synchronisation so only one PIC pin is needed It's in bmp format If you have any questions email me direct if you like Peter Cousens RemoveMEpeterTakeThisOuTcousens.her.forthnet.gr Content-Type: image/x-MS-bmp; name="POWER.BMP" Content-Disposition: inline; filename="POWER.BMP" Attachment converted: wonderlandfive:POWER.BMP (????/----) (0000D6C4)

On Fri, 28 Mar 1997, Michael F. Hagen wrote: > Hi Aaron, > Don't chop! Use zero crossing detection. Decide to turn On or Off the > lamp, and leave it alone for a half cycle (1/120s). Chopping it causes > LOTS of grief. Radiated noise and excessive switching currents, will be > greatly reduced if you switch when the voltage is zero. Depending on > the inductance of the load, the current might even be near zero. Also > use opto-isolation to protect the PIC and other circuits. The PIC > should have time to do a calculation every 1/120S? > Good Luck > Mike > Mike, Do you mean to just play with a full half-cycle at a time? Meaning, do I leave the lamp off for 5/120ths of a second and on for one, off for 5, etc.. would this cause a flicker in the lamp at low intensities or is the incadescent filament slow enough to take care of it? I am just wondering about a nearly-off lamp where I pause for, say, 12/120ths of a second and on for 1/120. This is effectively flickering at 10 times per second which is noticable, is it not? Thanks, aaron ------------------------------------------------------------------------------- - Aaron J. Miller Email: spamBeGonemilleraspamBeGonestanford.edu - Physics Graduate Student Web : http://www.stanford.edu/~millera - Stanford University Phone: (415) 372-5575 (until 4/12/97) -------------------------------------------------------------------------------