PFC maybe? -

[fuseshut]

I have something I don't quite get. I have an AC circuit that controls an incandescent light. 220VAC as on/off switch - like an AC solid state relay. Works fine. Once I connect a switching power supply the AC output of my circuit drops the voltage down to nothing. I have no load on the switching power supply. I install an incandescent 75W light, my circuit works peachy. nothing is burning, there is no current draw.

 

[MacGyverS2000]

Most switchers want to see a minimum current draw. If your circuit isn't pulling that, it may shut down.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[itsmoked]

I agree with Dan.
Hook your 'unloaded' switcher to your AC source directly to confirm this.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[OperaHouse]

I'm not sure if I understand what he is saying, but it might be this. The solid state relay is in series with the power to the switching power supply. The power supply doesn't get any AC input unless a lamp is also placed at the input.

If that is the case, I believe it is because the supply's input caps fully charge at no load. Current can only flow at the very peak of the cycle. As soon as the current latches the solid state relay on, the peak voltage lowers and the relay turns off till the next cycle. Result is an apparant very low AC voltage at the supply input. This will need some resistive load at the input in order to work.

 

[fuseshut]

Yes, like I said I hook up an AC light to my circuit. I can control the light with no problem, there's 220VAC at the input of the light - no decrease in voltage. I hook up a AC/DC power supply (that has 220VAC input) thinking my circuit can drive it and the voltage decreases to almost half. The AC/DC power supply is unloaded so I shall at least get the 220VAC right at the input terminals of the AC/DC power supply - right?

 

[OperaHouse]

The power supply may still operate properly even though the voltage you measure is lower. A PFC power supply looks more like a resistance load. Other power supplies only draw current at the peaks when the sine wave voltage exceeds the charged capacitor voltage. It may not seem obvious, but an input voltage reading is not a valid measurment and doesn't tell you anything. This could be very dangerous to someone testing the unit who thinks a voltage doesn't exist.

 

[fuseshut]

Yeah the AC/DC power supply I am using does not have PFC. I am using a triac to drive the AC output with a RC snubber across the triac terminals. I am wondering if it is a charging/discharging issue. I am going to do some more probing with a scope. The input the DC power supply shall not read 1/2 of the mains AC voltage. SUre it might be 215VAC or even 205 but not 125.

 

[cbarn24050]

Your triac needs current to keep it conducting. A smps draws current for a small part of the cycle and so goes off before the voltage crossing point. If you have zero crossing fireing and are using pulses to fire it then it may not even conduct at all as the smps doesnt draw current at this time either.

 

[fuseshut]

Cbarn that makes sense. The triac must see current in both directions to stay latched. I am using an opto-triac (MOC3052) to fire the 3 terminal Teccor triac. I would think there's a work around to use a triac to power up a SMPS. Could implement so some of filter on the front end o the SMPS so that the triac operates?

 

[fuseshut]

Didn't know this but now I do. SMPS require a minimal input current to operate? What I did was place a power resistor in parallel and the SMPS is working fine.

 

[OperaHouse]

Again, is the problem it doesn't work or you just don't like the input voltage you measure? I would have suspected
that you would measure an even lower voltage than 125V. The triac will not conduct till the sine wave voltage is higher than that of the power supply cap and will turn off when the voltage is lower than the supply cap. Current draw is exactly the same as if you didn't have the triac. From what you have told me this is not a problem at all. What you see sounds perfectly normal!

 

[itsmoked]

fusehut; I am confused to-no-end..

1) Are you saying that you are trying to run an unloaded SMP with an opto triac?

2)And it does not work until you put a load resistor across the line in parallel with the SMP's input? (i.e. no discussion or tinkering about the SMP's output)

Just trying to understand.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[cbarn24050]

You could allways use 2 mosfets (or igbts) instead. There is an application note on IR's website.

 

[fuseshut]

Itsmoked,
I am connecting an SMPS (AC-DC)to my triac circuit. There is no DC load connected. I measure the AC voltage out of my triac circuit (the AC input to the SMPS) and it reads ~120-135VAC. Now, the input to my triac circuit is 220VAC so when I digitally control my triac (not as a dimmer but as a SSR) I shall get 220VAC output from the triac circuit, not 120VAC. When I connect a simple power resistor that will draw say 2Amps I measure the voltage across the power resistor and read 220VAC, like I should. When I connect the SMPS I measure the half voltage. If I connect the SMPS and power resistor in parallel, everything is happy, the voltage is ~220VAC, the DC output of the SMPS is finally on.

 

[OperaHouse]

I would suspect that 2W line load would be sufficient to indicate full voltage at the input of the SMPS. I still see this as a non problem. Won't you allways have a load on the SMPS anyway? It seems obvious that you should measure 220V at the input of the SMPS, but this is not correct thinking. If you looked at the waveform you would see something closely looking like a modified square wave. Read my posts again. If you power supply has input voltage monitoring, you might think of looking at an alternate supplier. A RC network (.47uf 100 ohm) might be sufficient to provide quicker turn on without generating a lot of heat. There is also the old standby - the relay. Why have all that circuitry when you can buy a SMPS with a turn on pin.

Are you really saying that with no load the SMPS does not turn on?

 

[itsmoked]

Thanks fuseshut for the clarification.

I would agree with Opera 1) get an SMPS with a turn-on input. Or 2) Use a relay not a triac.



Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[LionelHutz]

I'd like to see the circuit you're using to trigger the triac to fully understand what you're doing.

I think you're operating a triac into a bridge rectifier and you're using the voltage drop across the triac to trigger it. It so, you may not have the voltage available to trigger the triac that you think you have. Then, if you are not triggering the triac at the right time you won't trigger it.

Instead of worrying about the AC input voltage worry about turning the triac on at the right times to actually charge the DC capacitor in the power supply.

Maybe replace two of the diodes in the bridge of the power supply with SCR's then use opto-SCR's to trigger these instead.

 

[BlownUp]

I am thinking that what you are seeing is that the triac is switching on and then is switching off before then end of the half cycle. Once a triac is kicked on it will stay on only while current is flowing through it and must be switched on every half cycle. Once the current goes to zero the triac will switch off. There are a couple of PS input designs that might cause the current to go to zero before the end of a complete half cycle. This may result in the triac switching off until your trigger circuit kicks it on again in the next half cycle. Most SMPS take a few cycles being on before the switching circuit will start. I am guessing that the SMPS internals are not starting because the effective RMS voltage is too low. If you look at the input to the SMPS with an oscilloscope you will likely see that the triac is only on for a little more than 1/2 of each half cycle. I suspect this is what is happening. Adding the load resistor to the output of the triac keeps the current from dipping low enough for the triac to switch off.

It would be interesting to see how your trigger circuit is setup. It might be possible to modify it to hold the triac on.

If however you can live with the power lost in and resulting head from the resistor then is sounds like the problem is solved.

 

[itsmoked]

Welcome to the fora BlownUp. Nicely summed up.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com