TRIAC dimmer circuit

Discussion in 'The Projects Forum' started by JimG, Apr 8, 2010.

  1. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    I'd like to control a heater using a TRIAC dimmer-type circuit. Load will be around 8A on 120VAC power.

    The attached circuit is pretty close to matching my needs. Except I need to be able to automatically control Rvar, and therefore the firing angle, using an external 4-20mA signal.

    My first thought is to drive an LED directly from the 4-20mA signal, and use an LDR in place of Rvar. But my first attempts were unsuccessful since the range of resistance I got from the LDR was around 200 ohms to 400 ohms when illuminated by the LED. I need the LDR resistance to be around 1K to 60K.

    So far I have breadboarded an LED / LDR circuit to test the LDR performance. But I have only simulated the rest of the circuit on LTSpice.

    I don't want to consider any circuits that require a DC power source. The 4-20mA signal will come from a commercial PID temperature controller. I also want the control signal to remain isolated from the 120VAC power.

    My questions:

    1. Where to find an LDR that might match the resistance range I need more closely?

    2. Is there a better approach to this problem, within the constraints I have stated? (I know that I can use a commercial proportional control SSR, but these are expensive and I'd like the flexibility offered by a homespun circuit).

    Thanks in advance.

    Jim
     
  2. Bychon

    Member

    Mar 12, 2010
    469
    41
    I can figure out how to use the 4-20 circuit to put a pre-bias on the diac circuit, but not with isolation.

    Next helper, please step up.
     
  3. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    Have you GOT to use the 4-20mA from PID controller?

    It's normal to control heating elements in 'burst fire' mode and most commercial PID controllers support that. The logic signal from the controller is just fed to a basic SSR which only switches at zero crossings.

    Usually the 4-20mA output is for remote monitoring of the temperature sensor, not intended for control. Does yours actually vary immediately if you change the setpoint slightly (which it would if it's a control signal), or is it just reading the temperature sensor?

    Phase control of heavy loads causes interference and spikes, and generally the response time of a heater is so slow that cycling on & off every second or two is fine.
     
  4. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Thanks for your thoughts, Robert.

    You are of course correct about PWM and normal zero-cross SSR's being the default, and very effective. My reasons for using a phase angle controller are primarily because of the "collateral damage" from PWM for my specific application.

    That is, in many kitchens (the heater I want to control is in a kitchen appliance), incandescent overhead lights share a circuit with appliance outlets. Switching the 1000W heater on/off abruptly creates a mild, but sometimes annoying, strobe effect.

    Phase angle control eliminates this problem (as well as another minor problem with pressure pulses from a small pump). And so far, I have not observed any ill effects from harmonics associated with clipping the sine wave in mid-phase.

    The output from the PID controller could also be an analog DC voltage range (0-10V or 0-5V), but the mA output seemed to better fit my original plan to drive an LED.

    So I am still looking for a way to interface analog output from the PID to a TRIAC circuit, and will welcome additional suggestions.

    Jim
     
  5. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Could you use a zero-cross monitor and a timer to fire the triac?
     
  6. Duane P Wetick

    Active Member

    Apr 23, 2009
    408
    19
    First of all, what is the means of input for the PID, RTD? Thermocouple? Fluid Level Sensing? The usual means of controlling temperature ia a on-off-on variable duty-cycle burst type control, not a proportional phase control as the hysteresis (control band) is very wide and response times are slow. I`m not sure why you want to use a 4-20ma signal to control temperature...am I missing something?

    Cheers, DPW [Everything has limitations...and I hate limitations.]
     
  7. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Possibly. Can you tell me a little more about this?

    Thanks.

    Jim
     
  8. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    If you monitor when the ac-phase crosses zero, you can use that moment to start a timer. By using the timer to vary delays you can choose the firing angle of the triac.

    Cross zero at t=0
    start timer
    timer 22ms
    timer runs out,
    fire triac
     
  9. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Thermocouple.

    Yes, I know (see earlier post). I think any PWM option, even burst type, will cause the strobe effect in the overhead kitchen lighting and pulsing of the pump. This is what I'm trying to eliminate. I've used proportional control driven by 4-20mA with a Crydom phase angle SSR, so I know it works. I want to design my own circuit in place of the SSR, though, for flexibility and cost reasons.

    I want to use phase angle control to switch the heater. 4-20mA is a convenient option (could also be 0-10V or 0-5V) for directly driving the LED I was trying to use with an LDR.

    Jim
     
  10. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    That's essentially what my initial circuit tries to do (attached below). The timing piece is done via an RC network. The time required to charge the cap up to 30VAC (which then fires the DIAC) represents the phase delay.

    The problem I can't solve is that the resistance of the LDR ("Rvar" in the schematic) is too low to cover an entire half cycle of the sine wave.

    I estimate I need an LDR with a resistance range from 1K to 60K (approx), instead of the 100 to 400 range I'm getting now as I vary the LED current from 4mA to 20mA.

    For this to be a workable, I need a solution that only uses passive components because the circuit won't have a DC source to power any IC's.

    Jim
     
  11. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    If the LDR has a high enough dark resistance to shut down the power circuit, I'd concentrate on adjusting the LED current to give the correct illumination range to control the firing angle.

    Something like a visible LED with a lowish series resistor, eg. 100 Ohms as the main load for the 4-20mA loop, then the LDR led in series with a pot across that so you can tweak the maximum LDR illumination.

    You could put another resistor or pot across that circuit, something like 470 Ohms or a 1K pot, so you can bypass some current and the LEDs don't start to illuminate until around 4mA to give low end adjustment.
     
  12. badbaud

    Member

    Mar 5, 2009
    26
    0
    It's called a Vactrol similar to the one (p/n VTL5C1) I recently used on a project to control an existing motor speed control by connecting to the speed control pot directly with the Vactrol, 3 MCU port pins switched different resistor values via transistors into the Vactrol input for S,M &F speed control. PWM'ing the MCU output would give you variable control with one Vactrol. Newark electronics can provide you data sheets.
     
  13. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    It a good product, however, the OP has a requirement of passive components only due to a lack of power for the uC..
     
  14. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Good suggestions. I'll give them a try. Made me realize that maybe I need to think in terms of operating the LED at the lowest end of its output range, rather than the highest. With any luck, the system might be more linear there, too.

    Jim
     
  15. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    The Vactrol looks to be exactly what I've been trying to build (didn't know such a thing already existed). The VTL5C1 data sheet shows approx. the resistance range I need as long as I can keep the input current below 1mA. That might take some work, but can probably be done. (The lower current is consistent with rjenkins' thoughts, too).

    Better yet, the Vactrol meets my definition of "passive," at least for these purposes.

    Thanks again to all who answered. Hopefully, I can take things from here. But if not, I'll be back :)

    Jim
     
  16. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Gotta love this forum.
     
  17. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    Happy to report a small amount of progress. Using an online recipe for a homespun Vactrol, I built something that has an LDR and a green LED at opposite ends of an old 35mm film canister.

    The LDR is less sensitive to green wavelengths, so there's a larger working range on the LED current when using green instead of red.

    First mockup gave LDR resistances ranging from 18K to 75K corresponding to LED currents of 20mA and 4mA, respectively.

    Jim
     
  18. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    You can also use a green piece of plastic as a filter to further narrow down the sensitivity of the LDR. That may be just what you need to push it into where you need to be.
     
  19. badbaud

    Member

    Mar 5, 2009
    26
    0
    And much cheaper then a Vactrol.
     
  20. JimG

    Thread Starter Member

    Dec 7, 2009
    53
    0
    A little more progress to report. Homebuilt vactrol now is built using 1.5" long x 3/8" ID polycarbonate tube. Stuff a pair of conical plastic caps in the ends.

    Poke a couple of holes for leads, and mount LED into one of the caps, and the LDR into the cap on the other end.

    A few dabs of epoxy to anchor everything, and cover with 1/2" black heat shrink tubing.

    Resistance range goes from 42K to 200K. LED is fed directly from 4-20mA output from Watlow 96 controller. 1/R is nearly linear with the LED current.

    Jim

    ED -- Using PDV-P9008 LDR from Advanced Photonix
     
    Last edited: Apr 15, 2010
Loading...