Integrated circuit for HID

Hi Seniors,

I am new to this forum and I am designing an electronics circuit for 73 watt HID lamp. The lamp power is controlled with TRIAC. The spice model of my design is shown below in the first attachement. I can already control the current through lamp, but the the problem is the current waveform is not smooth enough. I mean the the current is rising suddenly to the maximum value as you can see in the following attachement 2 (dimmer), but I want that it would increase smoothly just just like a sine wave. My desired result is shown in the third attachement below.

I tried to go through all of the threads related to TRIAC, but I could not fine any solution for my problem. Could you please help me to modify my my circuit to get desired result.

I tried to explain the situation completely, but if you want to know something more about it please ask me, and I sorry if this is very low level question for you and for wasting your time.

Thank you in advance,

Sacha

I don't believe a triac based dimming solutions are feasable for HID lamps, typically the ballast is set up with switches that switch in/out capacitors to allow the current to be controlled, they're much more finicky about dimming that other bulb types.

What is the specific type of your HID bulb?

Hello sceadwian,

Thank you for your very quick response.The type of the lamp is Philips MasterColour 70W CDM-TC Elite Ceramic Metal Halide. for more details: https://www.energybulbs.co.uk/produ...+CDM-TC+Elite+Ceramic+Metal+Halide/3931446778

Yep, definitely an HID lamp, you can't use a triac to dim these, more sophisticated ballasts are required.

You are right that a ballast is necessary for dimming HID lamp. I had already plan to use a ballast and now when I have added ballast I got the required result. thank you for pointing out.

Now I have to control the TRIAC gate current electronically not by using variable resister, for that have to use a triac driver and some circuit that can create pulses for triac driver.

My goal is to measure the current and voltage of lamp and limit the curent if necessary by using TRIAC, to keep the lamp running at constant power, where the lamp voltage is changing between 50-125 volt.

I have found some optocouplers e.g from "fairchild semeconductor" with zero crossing detection. I was wondering if you have some better suggestion.

Hi,


One of the best ways to trigger a triac is to use a zero crossing detector followed by a variable delay circuit that puts out a gate pulse at the end of the delay period. For 60Hz the delay would be from near 0ms to near 8.33ms, for 50Hz the delay would be from near 0ms to near 10ms. Near 0ms the output is fully on, and near the higher end (8.33ms or 10ms) the output is fully off.

Thank you MrAl, for your reply, I have already tested the TRIAC with different delay by using a variable resister, but now I want to control it electronicaly, I mean the lamp is running with constant power i.e 73 watt, so if lamp voltage will increase, it should be measured by the circuit itself and to keep the power constant the TRIAC should decrease the current. I can use a zero crossing detector, but the deley circuit? Could you please give some suggestions about how to control TRIAC delay,

Thanks in advance.

Hi,

The typical way to do this is to simply trigger a one shot with the zero cross detector. The one shot always starts at the zero cross, so if you make the one shot delay time variable it triggers the triac gate after the delay has expired. That turns the triac on of course, then the triac turns off at the next zero cross and the one shot is trigger again. This process repeats over and over. As you make the delay longer and longer the output time 'pulse' gets shorter and shorter so output power goes down lower and lower. As you make the delay shorter the power goes up until the delay is barely any time at all (near zero) and that causes the triac to put out full power.
If you intend to measure the output and use that as feedback you'll have to use a voltage controlled one shot maybe like the 555 timer for example or a full fledged PWM circuit.

So the two parts of the timing circuit are the zero cross detector and the adjustable one shot.

I assume you are dimming the ballast and not the lamp directly, and that you have experimented with that already and it does in fact dim using that method and the ballast accepts the triac waveform without a problem. I've never tried this myself so i cant say if it works or not.
Or, you have tried dimming the lamp directly after the ballast and found that it works ok that way.

There are other issues that come up too here, such as how fast to dim the lamp from full brightness. A slow level decrease is recommended over perhaps a 2 minute period.

Thank you again for such a details reply. I am dimming the lamp directly not the ballast and it works fine, but only the difference is. I used variable resister to to adjust the delay for TRIAC, but now I want to control it electronicaly that I mentioned already. Could you please give me some suggestion, how can I feedback the lamp voltage that can change from 50 volt to 130 volt to the timer.

Hi,

The idea in any regulator circuit is the output is measured and compared with a reference voltage and the difference becomes the error voltage and this is integrated and used as a control signal to PWM the output.
To get the basic idea how this works, lets say you are using a one shot to get the delay. A longer delay means less output and lets say a higher control voltage means a longer delay and lets say the output is positive. This would mean measuring the voltage, comparing it with a set reference (the user control voltage setting), and the output of that integrated so that the output of the integrator ramps UP with higher output voltage. As the output drifts up, the difference becomes greater, which causes the integrator to ramp up, which causes a longer time delay, which causes the output to ramp down slightly which keeps the output constant.
The integrator time period has to be long enough to smooth out all the line frequency variations also.

Is this clear enough?

Hello MrAl,
So far it is quite clear what were you suggesting. I have made Triac trigger control circuit that can convert the pulse signal (current) from PIC or some thing else (I have not decided yet) to get suitable -ve trigger signal (5mA-50mA and 0.9- 1.5V) to trigger the triac. I have made one that is doing this job but it became very complicated, because I used four transister and two DC power supplies. could you please suggest me some simple signal that can do this job.

Thankyou.

Hi,


Show me what you have so far and tell me what part you want to improve. Is it just the trigger drive circuit?

Sorry for bothering you again and again. right now I am just working on trigger drive circuit that can convert the pulse into a suitable signal to trigger the triac that is in this case 5-50 mA and 0.9-1.3 volt. the pulse should be -ve because the triac should work in 2nd and 3rd quadrants.

1. In the attachement 1, I got the required pulse but I want to have simpler solution.

2. The circuit shown in attachement 2 is very simple that creates +ve pulses that means it just work in first quadrants. I checked it by changing the polarity of current source but then the output pulse is so small that it does not trigger the triac. I could not succeed to get a pulse with higher amplitude. It will be very help full it you help me work it out.

3. I also tried to get the output from collector, in the similar circuit that can give -ve pulse but it is very unstable to keep the triac running continously.

in my design i will use a ballast and hid lamp, but here i use just a resister as load for simplicity. may be the triac added in this design will not work for you. I hope you will be able to help me.

View attachment 62035
View attachment 62034

hi sacha,
If you require precise control for each quadrant, you have to detect the zero crossing point of the 50Hz mains supply.

A Triac Gate requires a current input for triggering, say 25mA to 50mA.

Hi Eric,
Thank you for your reply. Sorry if i was not clear enough. I am using a triac BTA06-600TW, for that I need 5mA to trigger, to get that, I used higher resieter in series with supply so that is ok. If I am not wrong I need -ve trigger pulse in triac gate to control the upper half as well as lower half of input AC signal (quadrant 2 and quadrant 3). I dont need to control it in other two quadrants.
I was simulating the circuit that you have sent to me, it works fine for upper half (quadrant 1) but it does not work for lower half, because of positive trigger signal.

To get the negative trigger signal, I changed the polarity of DC supply and it works just fine.

Later on I will have to use a comparator to control and create a delay signal for TRIAC. Do you think, I can get a required trigger signal (5-10mA) from a delay signal generated by a comparator?
I think i will have to use some amplifier for that.

Could you pleasegive me some suggestion here.

hi sacha,
The positive trigger will control the Triac in the upper and lower half cycles.??

If you used an Opto-coupler or low voltage transformer to provide a mains zero crossing pulse [ using a comparator] you could then use a 555 monostable to give a 'rough' Ramp voltage into the 2nd part of the comparator to give a trigger pulse, Varying from approx 0deg thru ~180deg.

I really appreciate your help eric.

You are right that TRIAC can work in four quadrants, but the one that I am using works just in first three quadrants at very low gate curent 5mA. I have attached my design here, I have to use an electromagnetic ballast with hid lamp, because of that the lamp voltage is crossing zero point at different time than the main voltage, that's why it does not trigger properly.

Can you please have look in my design give me some some remarks, whether i am on right direction.
View attachment 62043

hi,
I do not have the LTS model for that Triac.

Hi guys,

Eric:
I think he knows about the zero crossing, dont you Sacha?
Some triacs either have to be triggered in only 3 quadrants or else should be.

Sacha:
You do know about the zero crossing right?
I think you found out already, but if you connect the positive supply voltage rail to the main terminal instead of the negative supply rail then you can easily get a negative pulse to the gate.
Can you post the model information for your triac for LT Spice?

Thanks guys,

I know about zero crossing.

I have already made a firing circuit for my triac model and it is very simple and works just fine, offcourse it was possible with the help of your feedbacks and suggestion. I have tested it using a 1V pulse from function generator and i used an optocoupler for isolation and it works well.

In the next step I want to measure the voltage and current of lamp and calculate the lamp power and compare it with reference power (73watt). I am trying to find out which solution will be better analogue or digital (using some microcontroller) and what will I need for either of these.

Any suggestion will be appreciated,
Sacha