Go Down

Topic: rectified AC phase control (Read 11741 times) previous topic - next topic

wvmarle

That sounds bad! And like a typical risk of working with high voltages.

Did you try it without the coil in place? If so, did you get the correct signal on the MOSFET? That would at least confirm that the software and the switching circuits work.

My best guess is that you've got hit by the reverse voltage of the switching coil. Did your flyback diode survive the ordeal? If not, then that's a very likely candidate for the cause of this catastrophe.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

dgueorguiev

I use for GRD lead 10K resistor to minus of rectifier bridge and for probe 10K resistor connected to "+ " input to the coil  L2 . I  check voltage to Mosfet gate it was weird curves  mostly with "-" values  (probe on gate , GRD lead to (- )of rectifier bridge  . I have only one probe for the oscilloscope.

dgueorguiev

schematic view is attached

DrDiettrich

If your scope has a USB connection, you most probably short the Ardino Gnd to the scope Gnd. You cannot have both the Arduino and the scope connected to the PC, at the same time.

wvmarle



I don't really understand what that scope image shows, you have to give a bit more detailed description - and/or post an annotated image.

For the setup itself: I see you're using a solderless breadboard, but I recall you mentioned currents of a few Amps. That's too much for such a breadboard. I don't think it's the cause of your problems but it may very well be part of it. This will have to be soldered on perfboard or similar. The MOSFET you may need to place in a socket for easy replacement, or leave some space to solder in another one. Be prepared to break stuff, it's part of the game, I've seen my fair share of smoke and am not using your high voltage level.

You have a scope which is a great tool.

What I'd suggest you is to start off without the coil (as coils, being inductors, can cause problems) and first make sure everything else works as expected.

1) make sure the software works: get the zero crossing pulse working, and make sure your Arduino outputs the correct signal for the optocouplers (use your scope).

2) add the optocoupler part: make sure the signal as presented to the gate of the MOSFET is as you expect. You now know the software is good, so any issues are in the hardware.

3) optional extra step: add a resistor (1k would be a good start - make sure it can handle the heat, 100V = 100 mA so 10W of power to dissipate!). No inductive problems to deal with now. Then you can make sure that the resistor (later the coil) is switched the way it should be.

4) add the coil, with appropriate flyback diode.
You may use a Schottky as it is much faster than a regular diode (around 50ns reverse recovery, against 500ns to a few ms for regular diodes, which is plenty of time to fry other components at these currents and voltages). I don't know how to calculate back current, but as you put some 4A through that coil I'd go for a diode that can handle a >40A peak. Also make sure the reverse breakdown voltage is at least 200V, maybe even 300-400V to make sure.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

dgueorguiev

Thank you for all suggestions wvmarle. I will follow all of them and will give you update.It will take some time.

 answer to DrDiettrich :     NO, the oscilloscope is RIGOL and is powered by 115V  AC. That is why I use 10K  ohm resistors to my probe and probe GRD to prevent from damages, but look like is not enough.  I have only my ARDUINO board connected to my laptop.

wvmarle

I just see your added schematic.

Your voltage has gone up a lot! Now 200V to the rectifier, which means a peak of 282V! A lot more than the 150V I calculated with.

The symbol for the MOSFET is wrong, that's a PNP transistor (which is a totally wrong part, an NPN transistor could work but a PNP definitely not). The MOSFET type you indicate is suitable: 600V, 60A.

Another thing to test, before connecting the coil: make sure the 470k resistor can keep the 50 uF cap well charged, as you continuously charge the MOSFET gate. Especially at lower voltages that may be an issue.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

dgueorguiev

#127
Dec 30, 2017, 06:54 pm Last Edit: Dec 30, 2017, 08:38 pm by dgueorguiev
Hi everyone and Peaceful NEW YEAR!!!
I look again the original schematic and I think that we have switching circuit parallel to L2 and current trough D5 and D4 can make trouble. If we connect the switching circuit after L2 It will be better by mine opinion, but I am not sure. (see attached new version of the Schematic) .Plus Voltage drop inL2 will help to lower the voltage in switching circuit.
Q2 is a PNP transistor and I connect +5V to pin 3 (emitter) and (collector) pin2-  to pin1 of 4N25 . Is this wrong?
 about Q1 (IXFQ50N60P3) Gate is pin1 ,Drain is pin2 and source is pin3 - Is this is correct connection?
Original schematic R6 =220K  see the attached new schematic (R10) I do not see what is the function, because I think draining to"0"V the gate of Mosfet is trough pin 5-to pin 4 of U2(4N25) when U1 is shut off and we have current trough pin1 and pin2 of U2. Is this R6(R10) for protection, or It could be removed ?

wvmarle

Hi everyone and Peaceful NEW YEAR!!!
I look again the original schematic and I think that we have switching circuit parallel to L2 and current trough D5 and D4 can make trouble.
Where is D5?

Quote
Q2 is a PNP transistor and I connect +5V to pin 3 (emitter) and (collector) pin2-  to pin1 of 4N25 . Is this wrong?
 about Q1 (IXFQ50N60P3) Gate is pin1 ,Drain is pin2 and source is pin3 - Is this is correct connection?
Check the data sheet. That's where you can find the pinouts.

Quote
Original schematic R6 =220K  see the attached new schematic (R10) I do not see what is the function,
It's to keep the MOSFET switched off when both optocouplers are off. Never keep a gate floating.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

dgueorguiev

  See attachments:
"photo jpg" is signal to pin1 of U1  Like suppose to be 2ms wide pulse.
"photo(1) jpg" signal to pin1 of U2  close to expected.
"Photo (3) signal to Gate of MOSFET   Negative value ,no idea why?
"photo(2) voltage curve across 10K resistor I put on place of Coil L2, L1 is not connected needer.

I change C1 to 10uF,50V  no changes.
I checked all components could not fine the problem. I am not very good  with troubleshooting .I will appreciate any suggestions.

dgueorguiev

I do not know why photo(3) was not posted .I try again -no result ,but it is the same like in previous post when I use the load coil L2 and L1. Look like some component is failing ,but I can not catch this. I will order new diodes and Mosfet.

DrDiettrich

Photo(2) suggests that the FET is never fully off.. OTOH a 10k load is far away from the intended coil impedance. What if your try with 1k or less?

wvmarle

Indeed, it looks like the MOSFET didn't switch off for some reason. That image should be very close to the image you see in photo.jpg, basically mirroring the signal placed on gate - the second opto is just to help it switch off faster - it pulls the gate very fast to about +0.6V, the forward voltage of the phototransistor in the opto, which should be low enough to fully switch off the MOSFET.

MOSFET gate voltage is best measured with the source pin as reference, and as close to the pins as possible, preferably right at the pins. Of course it should never be negative, unless your ground reference is off somehow.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

dgueorguiev

I change all parts of the schematic strictly as required by  #3 ( of wvmarle) schematics with the exception of Mosfet which is IXFQ50N60P3 instead of IRF730 and part # for Q2 is 2N4403. I include the coil L1 and L2 in the circuit  too. I tested and it was showing 4.6 A  up to 120V input to the bridge rectifier. Vgs at the Mosfet was 5.6V . But when i analyze the  attached picture of the oscilloscope seamed to me that I do not have 2ms pulse and this high picks are 180 degree apart .When I increase the voltage above 120V AC ,my Mosfet stop working and I have to replace with new one. I am confused and do not know what to check next.

cattledog

Did you follow the recommended sequence of post #124? Where did things begin to look wrong?

Go Up