IGBT Dimmer problem..

[Bazza175]

I am trying to replicate this circuit here:



I have it built on a circuit board, and it sorta works.

If i apply a signal to the optoisolator. the output initially gets 240v output. After about 0.5 seconds, the output voltage drops to about 160v and just stays like it until i remove and reapply the signal to the optoisolator.
Aswell as this the IGBT is getting really hot, there is no flicker in the bulb used on the output and the load is currently 60Watts.

Does anyone have any idea whats happening ?

 

[chemelec]

Do you have a HEAT SINK on the IGBT?
It NEEDS ONE!

I am trying to replicate this circuit here:

View attachment 69307

I have it built on a circuit board, and it sorta works.

If i apply a signal to the optoisolator. the output initially gets 240v output. After about 0.5 seconds, the output voltage drops to about 160v and just stays like it until i remove and reapply the signal to the optoisolator.
Aswell as this the IGBT is getting really hot, there is no flicker in the bulb used on the output and the load is currently 60Watts.

Does anyone have any idea whats happening ?

 

[Bazza175]

yeah, i have a heatsink on it but thats not really the main problem. Although the heatsink is not really doing a lot the IGBT is getting so hot the paint on the heatsink is bubbling.

 

[chemelec]

Probably Need a BIGGER Heatsink.

Or AC RIPPLE could be your Problem!
Possibly ADD a FILTER CAP on the Supply.
Across the +/-of the Bridge Rectifier.

yeah, i have a heatsink on it but thats not really the main problem. Although the heatsink is not really doing a lot the IGBT is getting so hot the paint on the heatsink is bubbling.

 

[Bazza175]

Any idea on why i get a voltage drop after about half a second?

 

[chemelec]

See My EDIT Above.

 

[Bazza175]

Any particular size ?

Also what voltage rating will i need?

 

[chemelec]

What Supply Voltage?

If it's 110, Than probably 200 Volts.
If it's 220 Volts, than 400 Volts.

Capacitance, I would Recommend 100uF Minimum, But The More the Better.

Any particular size ?

Also what voltage rating will i need?

 

[Bazza175]

Ahh dont have any that big. I wonder tho how would ripple cause this problem, i initially get 240v then it drops to 160v, ripple wouldnt cause that surely how can i just loose 80v ?

 

[tcmtech]

So is there a reason you are using this type of circuit instead of a triac or back to back SCR based one for AC power control?

Seems overly complicated for such a simple process.

 

[chemelec]

My Thought is: Once the load comes on, It causes HIGH RIPPLE Currents.
These will now go into The Gate.

Another thing you could try is a 10 or 47uf, 25 Volt Cap across D8.
This might help without the other cap.
(It Might Produce a Slight Delay, (as Capacitor is Charging) When Power isInitally Turned on.)

Good Luck, Sorry I'm Going OFF LINE NOW.
And SEASONS GREETINGS!

Ahh dont have any that big. I wonder tho how would ripple cause this problem, i initially get 240v then it drops to 160v, ripple wouldnt cause that surely how can i just loose 80v ?

 

[Bazza175]

Im using it over a triac as they have the problem of dying under in a short, if the bulb connected to it blows this causes a temporary short that usually kills a triac. I also want to impliment reverse phase dimming. Back to back SCR sounds much like a triac.

 

[Bazza175]

Thanks for the help Chemelec and seasons greetings to you too.

 

[Bazza175]

I just looked again at the schematic and there is already a 47uF cap across D8

My Thought is: Once the load comes on, It causes HIGH RIPPLE Currents.
These will now go into The Gate.

Another thing you could try is a 10 or 47uf, 25 Volt Cap across D8.
This might help without the other cap.
(It Might Produce a Slight Delay, (as Capacitor is Charging) When Power isInitally Turned on.)

Good Luck, Sorry I'm Going OFF LINE NOW.
And SEASONS GREETINGS!

 

[chemelec]

Back on for a Minute.

Another Possibility, that 100K resistor to the Zener can only Deliver Less than 3 Ma to you opto Isolator.
You Might Reduce that resistor to 22K for better Drive Current.

Thanks for the help Chemelec and seasons greetings to you too.

 

[Bazza175]

Thanks for the pointer there chemelec, that appears to have done the trick.

I have noticed something and was hoping you might be able to educate me a little. I have noticed now, that when the lamp is dimmed low, the 22K resistor gets hot. When the lamp is not dimmed and is at full power then the IGBT gets hot.

Is this normal for this sort of circuit or is there something wrong with it ??

 

[chemelec]

Not Sure What would cause that.
I would thing Current through the resistor would be higher the other way around.

Thanks for the pointer there chemelec, that appears to have done the trick.

I have noticed something and was hoping you might be able to educate me a little. I have noticed now, that when the lamp is dimmed low, the 22K resistor gets hot. When the lamp is not dimmed and is at full power then the IGBT gets hot.

Is this normal for this sort of circuit or is there something wrong with it ??

 

[kwilbern]

We have a similar circuit and I had the same issue initially. I believe your problem is that when the circuit is fully on, the gate voltage to the IGBT drops. This causes the IGBT to get hot because it isn't driven hard enough. The problem with this circuit is that you are creating 15V with the D1,R7,D8 combination. However, when the IGBT is fully on, the leads are shorted and the voltage to the D1,R7,D8 is zero. Your C1 cap then has to supply the opto with the 15V. So if you have the IGBT fully on, your gate voltage sags, thus heating up your IGBT. You can verify this by measuring the gate voltage. We fixed this issue by breaking the D1 connection to the IGBT and tying the anode of D1 to neutral. That way you'll get a constant AC signal to D1 regardless of the state of the IGBT. Your gate voltage should be a steady voltage now.