Go Down

Topic: rectified AC phase control (Read 7019 times) previous topic - next topic

wvmarle

#30
Dec 08, 2017, 04:37 pm Last Edit: Dec 08, 2017, 04:37 pm by wvmarle
OK, and now I should stop playing with this :-) It's a good exercise.

MOSFET I chose the IRF730 for sufficient voltage (500V) and 6A max current. Hope that's enough.

Based on that recalculated the resistors, keeping in mind that we'd want a 10µs switching time for the MOSFET. This results in a 20k, 5W resistor for R4.

Checking the data sheet I also noticed the 4N25 optocoupler is not suitable for this application as the breakdown voltage is 70V, and the peak wave here is more like 210-220V. So this needs a different part, one that can handle 300V or so.


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.

TomGeorge

OK, and now I should stop playing with this :-) It's a good exercise.

Checking the data sheet I also noticed the 4N25 optocoupler is not suitable for this application as the breakdown voltage is 70V, and the peak wave here is more like 210-220V. So this needs a different part, one that can handle 300V or so.
Keep going. :)


Split R4 into two resistors in series and connect a zener to the midpoint and MOSFET source, that way you have a zener regulator.
To keep  Vmax to the collector below 70V of the 4N25, add a 10uF or so to give you some smooth gate voltage.
Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

wvmarle

Good idea. That could also help solve the massive 5W heat dissipation. The on switching is the real issue, I was thinking already of adding a cap somewhere to help giving that MOSFET gate a good push to get on.
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.

wvmarle

#33
Dec 08, 2017, 07:30 pm Last Edit: Dec 08, 2017, 07:30 pm by wvmarle
So like this.

Zener D7 limits the voltage over the cap to 25V. Every time the voltage goes over 25V on the rectifier (or over 150V if the optocoupler is on) the cap C1 will be topped up to 25V. Then when the optocoupler switches on, there's this 25V supply from the cap to help switch on the MOSFET faster. D4 prevents back flow from the capacitor when the wave goes below 25V.

The zener also manages the CE voltage of the optopcoupler.
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

OK ,you spend to much time for me.Thank you so much , but  I have some more details:
 1. What about if I lower R7 from 10K to 2.2K? Do you think it will work better for "0" crossing signal to Arduino?
 2. The idea for TRIAC is to open 3ms after"0" crossing and close at next "0" crossing, by itself. This will cut off first1/3 of half AC wave. ( I order TRIAC T-835H-6T) and (optoisolator triac driver 782-IL410 ). Modified AC after full wave rectifier bridge "+DC " go trough Mosfet ( I order logic mosfet  IPP200N15N3G) ,which will open at "0"cross and close 5ms. after that. the overlap of open circuit will give me 2ms pulses I need. May be there is easy way to achieve this goal, but I am not aware.
 3. I order zener   IN5253 .Is this correct? If not can you help me with new part number and the place to order

  Look like We are close to final view of the schematic. Can you help me with programing cod for the ARDUINO board? I have ARDUINO UNO board. And I have no idea how to write the program.
        Thank you ones again. You are amazing !!!
                                               Dimitre

TomGeorge

#35
Dec 09, 2017, 09:49 am Last Edit: Dec 09, 2017, 09:49 am by TomGeorge
Hi,
We were puzzled as to why you needed the Triac to start the waveform.

You don't need it as the MOSFET can do the switch ON and OFF, as the MOSFET is switching a DC supply, even though it is pulsing.
All your timing is referenced to the zero crossing detector.

Tom.. :)


Everything runs on smoke, let the smoke out, it stops running....

wvmarle

After posting I was thinking about my schematic again, and realised that as R3 and R4 are forming a voltage divider again, gate voltage will be limited to <4V with my suggested values. That doesn't work (plus the rise time of the gate is much too low).

The solution I was thinking of was to add a second optocoupler, in parallel with R3. Then R3 value can be increased drastically (to 220k or so) as it's main purpose is to keep the MOSFET off upon startup. R4 value can be lowered then, zener can also go down a bit, 15V would be fine.

The purpose of this is to make the switching as fast as reasonably possible - the fastest is to short the gate to either the higher voltage or the lower voltage: that's basically what those two optocouplers do. Well, with a current limiting resistor to keep things under control.

The reason of going to these lengths: you want the MOSFET to switch at specific points of the waveform, for a period of 2 ms. A delay of 20 us is 1% of that time, and that will start to influence any measurement you want to do. You must try and minimise this delay as much as possible.

A serious complication is the high voltages you have. You're delaying with about 210V peak, while the MOSFET can have no more than 30V on it's gate. So I'm looking to find a way to "kick" the MOSFET into on and off states respectively, providing the required 20+ mA to charge/discharge the gate which is needed to ensure fast switching.
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.

wvmarle

IPP200N15N3G

V(DS) = 150V.

Wrong part - your RMS is 150V so that means peak is more like 210V. Also running it at the limit is not a good idea, you may have voltage spikes. That's why I came up with a part that allows up to 500V V(DS).

Furthermore, it feels not very "logic level" when I read the data sheet. R(DS,ON) values are given for 8V and more. The graphs later also show very limited conductance at 5V V(GS). Now for this application you don't need a logic level anyway, as you have plenty of voltage available.

The maximum V(GS) for this part is 20V so that's another mismatch. If after fixing the schematic I posted you manage to get the full 25V of the zener to the gate, you'd destroy the part even before you blow it with the too high V(DS).
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.

wvmarle

#38
Dec 09, 2017, 03:26 pm Last Edit: Dec 09, 2017, 03:27 pm by wvmarle
This should allow for ~2 µs switch times, using a single digital pin from the Arduino (which makes programming it just a bit easier).

Q1 on: gate >8V.
Q2 off: gate <3V.

D11 HIGH:
- U1 on, gate of Q1 pulled to 12V (with the help of the charge from C1) through R3, Q1 on, L2 on.
- At the same time: base of Q2 pulled high, Q2 blocks, U2 off.
- Current through R6 is irrelevant.

D11 LOW:
- Base of Q2 pulled low, current (1 mA) flows, transistor conducts, U2 on, gate of Q1 pulled to 0V through R4, Q1 off, L2 off.
- U2 off.
- Current through R6 is irrelevant.

D11 INPUT (upon startup):
- U1 off.
- Base of Q2 disconnected, Q2 off, U2 off.
- R6 pulls gate of Q1 low, Q1 off, L2 off.

D11 is to source/sink about 13 mA when active.

All you experts, please comment :-) Hope I got it all correct now.
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.

DrDiettrich

I don't understand the coil between the transformers. Should it denote a filter of some kind?

Your circuit may work, except the two LED in series (opto coupler and external) may not work as expected.

wvmarle

I don't understand the coil between the transformers. Should it denote a filter of some kind?
OP explained that one before in this thread.

Quote
Your circuit may work, except the two LED in series (opto coupler and external) may not work as expected.
That's again original OP's doing - as it's in series why wouldn't it work? Resistor value is based on the total voltage drop expected (may have to be adjusted if OP's LEDs have different forward voltage).

Not that I really see the point of that LED, it'll simply flash very fast so look kinda dim.
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

Reply to Dr Diettrich: I explain in previous posts that L1 is my vertical coil(load) and L2  is second coil at 45 degree next to first. Reaction between the two coils is what I am interested.

Reply to "wvmarle ": Thank you very much .I think it is much simple now and I am prepared to build the circuit .I order zener diode 1N5349B(12V,,5W) and MOSFET _IRF730 (5.5A,400V),but I am not sure for PNP transistor .I have one BC636 . Do you think I can use it or I have to order something different ?

wvmarle

Any small signal transistor is fine there, nothing special about it. All it has to do is switch 5V,  and about 12 mA. Yours will do just fine.
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.

TomGeorge

I don't understand the coil between the transformers. Should it denote a filter of some kind?

Your circuit may work, except the two LED in series (opto coupler and external) may not work as expected.
L1 is another coil that is used with L2, L1 is constant flux, L2 variable, from what I understand.
We may be designing an antigravity device out of terrestrial parts from alien schematics. :o :o :o


Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

DrDiettrich

L1 is another coil that is used with L2, L1 is constant flux, L2 variable, from what I understand.
We may be designing an antigravity device out of terrestrial parts from alien schematics. :o :o :o
Then the zero-transition circuit does not reflect the proper phase. Every transformer adds another phase shift.

Go Up