Delayed AC Line turn on

Thread Starter

nekojita

Joined Nov 19, 2010
170
Hi,

From the Induction PWM controller thread, I need to be able to switch on AC line voltage after the +5V is already up. I used a MOSFET to drive a opto-triac driver chip.

This chip in turn drives the gate of the main triac to switch the AC line to the PWM Line input.

In the LTSPICE sim, the circuit works exactly as planned giving me a 100ms delay from 5V up to AC Line available to the PWM drive circuit.

When I built the circuit, the results were not the same: with R4 at 1k, the triac never turned on. As I lowered R4 value from 1k, I found that the triac was always on.

If I understand correctly, the triac needs a minimum load current to conduct fully. Since I don't have any way of knowing the load current, what AC switching circuit can provide a Line voltage switch regardless of the load current?

Does anyone in the Forum have a suggestion?

Thanks,

Neko
 

Attachments

eetech00

Joined Jun 8, 2013
3,856
Hi,

From the Induction PWM controller thread, I need to be able to switch on AC line voltage after the +5V is already up. I used a MOSFET to drive a opto-triac driver chip.

This chip in turn drives the gate of the main triac to switch the AC line to the PWM Line input.

In the LTSPICE sim, the circuit works exactly as planned giving me a 100ms delay from 5V up to AC Line available to the PWM drive circuit.

When I built the circuit, the results were not the same: with R4 at 1k, the triac never turned on. As I lowered R4 value from 1k, I found that the triac was always on.

If I understand correctly, the triac needs a minimum load current to conduct fully. Since I don't have any way of knowing the load current, what AC switching circuit can provide a Line voltage switch regardless of the load current?

Does anyone in the Forum have a suggestion?

Thanks,

Neko
The opto needs a minimum 15ma to guarantee turn on.
Change R2 to 220 or 240 ohm resistor.

What is the line voltage (VAC)?
 
Last edited:

Thread Starter

nekojita

Joined Nov 19, 2010
170
The opto needs a minimum 15ma to guarantee turn on.
Change R2 to 220 or 240 ohm resistor.

What is the line voltage (VAC)?
Hi eetech00,

Thanks for your response and comments. I'll try your suggestion of reducing the value of R2. Though the voltage is 370V p-p (120V rms), I don't know how this plays into a PWM chopped (10kHz) AC waveform in this triac switching scheme.

I haven't worked with triacs and don't know what to expect especially considering that the p-p voltage is constant but the on time per each AC cycle varies with the PWM setting.

Do you have any thoughts on this?

Thanks,

Neko
 

eetech00

Joined Jun 8, 2013
3,856
Hi eetech00,

Thanks for your response and comments. I'll try your suggestion of reducing the value of R2. Though the voltage is 370V p-p (120V rms), I don't know how this plays into a PWM chopped (10kHz) AC waveform in this triac switching scheme.

I haven't worked with triacs and don't know what to expect especially considering that the p-p voltage is constant but the on time per each AC cycle varies with the PWM setting.

Do you have any thoughts on this?

Thanks,

Neko
Take a look at this app note.

https://www.onsemi.com/pub/Collateral/AN-3006.pdf.pdf
 

Thread Starter

nekojita

Joined Nov 19, 2010
170
The opto needs a minimum 15ma to guarantee turn on.
Change R2 to 220 or 240 ohm resistor.

What is the line voltage (VAC)?
EETECH00,

i
The opto needs a minimum 15ma to guarantee turn on.
Change R2 to 220 or 240 ohm resistor.

What is the line voltage (VAC)?
Hi eetech00,

I changed the
The opto needs a minimum 15ma to guarantee turn on.
Change R2 to 220 or 240 ohm resistor.

What is the line voltage (VAC)?
Hi eetech00,

I changed R2 from 470Ω to to 240Ω. The triac turns on with R4 at 1k, however. If I leave out R4, the triac does not turn on. I read the app not and possibly, I don't meet the minimum load requirements for the device.

Any thoughts?

Thanks,

Neko
 

Thread Starter

nekojita

Joined Nov 19, 2010
170
Why not use a relay with a 5V coil to switch the mains?
Hi Hymie,

I looked at standard and solid-state relays for this application. I am space limited and the standard relays come in at about 20x10x10 mm, a bit larger than I'd like. The solid-state versions are smaller but cost quite a bit more. Certainly the standard relay is a reliable way to go.

Thoughts?

Neko
 

eetech00

Joined Jun 8, 2013
3,856
EETECH00,

i

Hi eetech00,

I changed the

Hi eetech00,

I changed R2 from 470Ω to to 240Ω. The triac turns on with R4 at 1k, however. If I leave out R4, the triac does not turn on. I read the app not and possibly, I don't meet the minimum load requirements for the device.

Any thoughts?

Thanks,

Neko
Why do you want to leave out R4? It's supposed to be there.

What is the load? inductive? resistive?
 
Last edited:

eetech00

Joined Jun 8, 2013
3,856
Hi eetech00,

Thanks for your response and comments. I'll try your suggestion of reducing the value of R2. Though the voltage is 370V p-p (120V rms), I don't know how this plays into a PWM chopped (10kHz) AC waveform in this triac switching scheme.

I haven't worked with triacs and don't know what to expect especially considering that the p-p voltage is constant but the on time per each AC cycle varies with the PWM setting.

Do you have any thoughts on this?

Thanks,

Neko
Are you talking about two different subjects here?
 

Thread Starter

nekojita

Joined Nov 19, 2010
170
Why do you want to leave out R4? It's supposed to be there.

What is the load? inductive? resistive?
In the original design that I found, there was no R4 and the sim showed it working fine. I added R4 from the app notes that I found for the device.

The circuit sim below does not match the actual circuit performance. If I go too high on R4, the triac does not turn on; too low on R4 and it's always on. h

On the bench and in the sim, I am testing with a resistive load. The actual load is a fan motor. The sim results with a resistive load:

1604788626988.png

Thoughts?
 

eetech00

Joined Jun 8, 2013
3,856
In the original design that I found, there was no R4 and the sim showed it working fine. I added R4 from the app notes that I found for the device.

The circuit sim below does not match the actual circuit performance. If I go too high on R4, the triac does not turn on; too low on R4 and it's always on. h

On the bench and in the sim, I am testing with a resistive load. The actual load is a fan motor. The sim results with a resistive load:

View attachment 221718

Thoughts?
The fan motor is an inductive load, not a resistive load. A resistive load is like an incandescent bulb.

Where is the PWM controller schematic?
 

Thread Starter

nekojita

Joined Nov 19, 2010
170
Hi Neko,
Your circuit already has good line switch M1, M2.
See how timer U7, R22 and Q1 provide delay 100ms.
View attachment 221722
Hi Danko,

Thanks for jumping in here! So following your suggestion, I added the FET driver for U3 cathode with the R22, C11 time constant at the gate from the +5V turn on. Saved me a lot of hassles using a relay. :)
BTW, the isolated supplies from your previous thread post worked great using high u ferrite toroids.

Is there an easy way to get the turn on at the AC zero crossing to minimize RFI?

Attached is the schematic and sim results.

Thanks a bunch :)

-Neko

1604804818243.png
 

Attachments

When I built the circuit, the results were not the same: with R4 at 1k, the triac never turned on. As I lowered R4 value from 1k, I found that the triac was always on.
This statement is misleading.

Without a load, the triac will appear to be on all of the time if only the voltage is measured.

Now, I won't suggest doing this, but shorting the output (where the load is supposed to be) should not result in magic smoke.
In order to "see" what the triac is doing you need a load of at least the minimum size,

Excessive dV/dt is another way the triac can turn on. That's why you see snubbers, or RC series combinations across the triac.

The other gotcha is that the gate current has quadrants and the minimum gate current required can vary depending on the quadrant of the AC line your controlling. Quadrant is a math concept. Quadrants I through IV and is dependent on the polarity of the voltage and current.

The triac turns off at zero voltage or zero current. In a resistive load, they occur at the same time. In an inductive load they don't.

If your making a "dimmer" from scratch that can handle an inductive load, you should incorporate current limiting and you need to "constantly pulse" the triac gate.

It's imperative you get a decent zero crossing signal and you need to control both "half cycles". If you take the equation of p(t) = (sin(wt)^2)/R; Just assume R=1. Your table of times to turn on in a half cycle is based on a power chunk. lets just say you have a table.

Time to turn on vs Power from1 to 100% and you control a half-cycle. I don't think you have to control each half cyccle independenty for finer resolution of power.

Incadesent loads should use the R(T) values where T is temperature.

I haven;t figured out how one would implement current limit, but I think you can do it with back to back SCR's rather than a triac. If current is exceeded, the gate drive can be removed and both SCR's might be able to fire at the same time. Not sure if it will work though.

In any event, power is proportional to V^2 and the resistance is generally constant for a resistive heater.
 

Danko

Joined Nov 22, 2017
1,829
The timer in your schematic pic is chopped off. Could you post your schematic?
Hi Neko,
I used "time controlled switch" from @Bordodynov's library.
That library contains useful models of many many many components.
Just download it from http://bordodynov.ltwiki.org/.
You also can use voltage source with delay as timer:
1604878436312.png
In post #60 I already told you:
I recommend you download and install full @Bordodynov's additional library for LTspice
from http://bordodynov.ltwiki.org/.
 
Last edited:

Thread Starter

nekojita

Joined Nov 19, 2010
170
Hi Neko,
I used "time controlled switch" from @Bordodynov's library.
That library contains useful models of many many many components.
Just download it from http://bordodynov.ltwiki.org/.
You also can use voltage source with delay as timer:
View attachment 221804
In post #60 I already told you:
Thanks, Danko for explaining. A few weeks ago, I downloaded Bordodynov's library though I didn't install all the files- just what I needed at the time. Lots of good models in the master file. :)
 

Thread Starter

nekojita

Joined Nov 19, 2010
170
This statement is misleading.

Without a load, the triac will appear to be on all of the time if only the voltage is measured.

Now, I won't suggest doing this, but shorting the output (where the load is supposed to be) should not result in magic smoke.
In order to "see" what the triac is doing you need a load of at least the minimum size,

Excessive dV/dt is another way the triac can turn on. That's why you see snubbers, or RC series combinations across the triac.

The other gotcha is that the gate current has quadrants and the minimum gate current required can vary depending on the quadrant of the AC line your controlling. Quadrant is a math concept. Quadrants I through IV and is dependent on the polarity of the voltage and current.

The triac turns off at zero voltage or zero current. In a resistive load, they occur at the same time. In an inductive load they don't.

If your making a "dimmer" from scratch that can handle an inductive load, you should incorporate current limiting and you need to "constantly pulse" the triac gate.

It's imperative you get a decent zero crossing signal and you need to control both "half cycles". If you take the equation of p(t) = (sin(wt)^2)/R; Just assume R=1. Your table of times to turn on in a half cycle is based on a power chunk. lets just say you have a table.

Time to turn on vs Power from1 to 100% and you control a half-cycle. I don't think you have to control each half cyccle independenty for finer resolution of power.

Incadesent loads should use the R(T) values where T is temperature.

I haven;t figured out how one would implement current limit, but I think you can do it with back to back SCR's rather than a triac. If current is exceeded, the gate drive can be removed and both SCR's might be able to fire at the same time. Not sure if it will work though.

In any event, power is proportional to V^2 and the resistance is generally constant for a resistive heater.
Thanks for your comments, many good points to consider! -Neko
 
Top