Author Topic: Driving (AC) mosfet switch directly from MPU using GDT  (Read 6129 times)

0 Members and 1 Guest are viewing this topic.

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Driving (AC) mosfet switch directly from MPU using GDT
« on: October 22, 2019, 08:19:11 pm »
Hello,
I've already written a few lines of code
Code: [Select]
 127                    .L6:
 128 003a C29A                  sbi 0x18,2
 129 003c C098                  cbi 0x18,0
 130 003e 00C0                  rjmp .
 131 0040 0000                  nop
 132 0042 C09A                  sbi 0x18,0
 133 0044 C298                  cbi 0x18,2
 134 0046 0000                  nop
 135 0048 00C0                  rjmp .L6
to output on two Attiny85 8MHz MPU pins ~600kHz signals like below, thanks to optimizing _delay_us with taking into account delay in RJMP .L6 those output seams to be very symetrical, while I'd like to use them as inputs via resistors from MPU to small gate drive transformer primary, where GDT secondary turns will be connected directly to mosfet gate.



For the moment I've MPU connected to 3.7V Li-ion baterry without any voltage regulator, so we have 3.7Vcc and mosfet gate voltages in the range +/-20V in my case.
Do I need eg. zener diodes to protect mosfet gate against spikes above lets say 15V, when we chose GDT turn ratio to something like 1:3.75, so 5 turns on GDT primary vs 19 turns on GDT secondary connected to mosfet gate?

Idea behind this GDT is control power of resistive load (80W 230VAC red light bulb)  by managing MPU pulses frequency or maybe better pulses width, but at such low amount of turns on GDT primary probably better to keep this frequency high, however it will depend of course on GDT core material used.

I had some success of driving this way AC mosfets switches, but in on/off mode where GDT was used as transformer to step up voltage from 5V to 12V clamped by zener diode on mosfet gate with additional rectifier diodes and gate resistor around 100k.

Didn't tried this directly without rectifier diodes, so is it possible to  :-BROKE mosfet since its gate is itself short circuit with GDT secondary windings?  ::)

What do you think?

 

Offline Simon

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #1 on: October 22, 2019, 08:43:00 pm »
I'm not an expert in pulse transformers but the idea of a zener sounds like a good one. It would mostly short circuit any negative outpu or you can put two back to back, or just rectify the pulse and then pun a zener in parallel. You should also put a resistor to discharge the gate quikly enough so that it turns off again particularly if you rectify the pulse.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #2 on: October 22, 2019, 08:57:52 pm »
0. Why use an MCU at all?  Especially bit-banged, where a CPU crash fries the driver (and probably the switches too)?
1. Why bit-bang instead of use the timer-counter?
2a. Why does the lamp/switch need to be driven at 600kHz?
2b. Or is this more like an isolated switch, why not use PV drivers instead?  Do you need bigger transistors than those are suitable for (because of leakage current and switching speed), so are considering a transformer coupled driver?  In that case, what kind of circuitry are you proposing to place around the transformer?

You didn't provide a schematic so it's hard to give any commentary on things like zener diodes or transformers.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Yansi

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #3 on: October 22, 2019, 10:09:03 pm »
You should also put a resistor to discharge the gate quikly enough so that it turns off again particularly if you rectify the pulse.
Resistor not needed since as we can see below GDT secondary shorts mosfet gate  ;)



Those two zener diodes (12V) are really needed here since we have transformer secondary connected directly to mosfet gate, so when MPU do not sens pulses to GDT primary low resistance wire shorts mosfet gate?  ::)

Another thing is maybe we need also protect MPU pins against GDT primary few turns? We have 2k in series resistance with GDT primary,
but didn't tested if this will work.

I've already made such circuit with GDT 5 turn primary : 19 turns secondary, but without zenners on mosfest gate  :-/O
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #4 on: October 22, 2019, 10:20:35 pm »
0. Why use an MCU at all?
Light bulb power should be adjusted among others based on temperature sensor or simply controled by radio communication etc, so MPU as control logic is a must.

2a. Why does the lamp/switch need to be driven at 600kHz?
Because of we have only a few turns on GDT primary, so the higher switching frequency the better I think, since it works like DC-DC converter to step up voltage from VCC to something like 12V mosfet gate low RDSon gate voltage pulses-it will not be visible.
Another story is whether mosfets I've used in AC mosfet switch will turn on/off at such switching frequency by using this MPU driven GDT,
since so far I've used rectifier diodes in on/off switch, but I'd like to try control power without 100Hz mains blinking, so maybe AC mains will be rectified to 325DC with additional capacitor, so it desn't matter if it will be 10kHz or 600kHz.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #5 on: October 22, 2019, 10:25:10 pm »
Ain't going to happen and don't even dare to try. :-BROKE

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #6 on: October 22, 2019, 11:27:36 pm »
You'll be much better off with a phase controller.  Instead of a transformer, use an SSR.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #7 on: October 22, 2019, 11:41:03 pm »
You'll be much better off with a phase controller.
I've zero cross detection hardware and MPU software, but I wanted to try do it in a diffrent way  :-/O

I've AC mosfet switch as shown below based on two IRF840.


VGS is rated as +/-20V and 12V zener diode is on this prototype PCB with additional 100k resistor which of course in this experiment we should remove  8)
I think 12Vz zener diode should be fine even without second one, because of gate voltage will be clamped to 12V at worst case during GDT pulse or to -1.2V something like this at negative pulse, so no way we  :-BROKE those mosfets with this GDT, but I didn't made any calculations based on IRF840 datasheet how much charge we need to fully turn those two mosfets ON since gate capacitance is in parallel in AC mosfet switch and probably those two 1k resistors on GDT primary which limits input current to ~2mA might be too low, but we can easy limit to 20mA on those two MPU pins and we'll see what happends.

No magic smoke so far, so good  >:D

 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #8 on: October 23, 2019, 03:17:12 am »
I think you will discover something interesting if you dig into the gate charge.

Recall that:
- Charge is the time integral of current, i.e., it takes a pulse of current (of some magnitude and length) to charge the gate from/to some voltage.  I = dQ/dt.
- We can calculate an equivalent gate capacitance by using the gate charge divided by the rated on voltage (Vgs(on), usually 10V).  C = Q/V.
- If we have an RLC series circuit, we can calculate the impedance and quality factor from Zo = sqrt(L/C) and Q = Zo/R.  We typically want Q low for a gate drive circuit.  (Note that, for a transformer coupled gate drive, this is the leakage inductance.  We can however apply the same reasoning to the magnetizing inductance and the primary side drive circuit, which usually has a coupling capacitor to block DC bias.)
- If we have an RLC circuit on the drain, we can also calculate the peak voltage there, and see if it may damage the transistors.  We'd have to make some assumption about the inductance of the load (probably not much, as a light?) and the mains supply (ballpark 200uH).
- In the event the drain voltage is excessive, what might we do about that?  RC damper?  TVS diode?
- Also, MCU pins are standard LVCMOS, i.e. about 30 ohms output resistance.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #9 on: October 23, 2019, 03:49:41 am »
You should also put a resistor to discharge the gate quikly enough so that it turns off again particularly if you rectify the pulse.
Resistor not needed since as we can see below GDT secondary shorts mosfet gate  ;)



Those two zener diodes (12V) are really needed here since we have transformer secondary connected directly to mosfet gate, so when MPU do not sens pulses to GDT primary low resistance wire shorts mosfet gate?  ::)

Another thing is maybe we need also protect MPU pins against GDT primary few turns? We have 2k in series resistance with GDT primary,
but didn't tested if this will work.

I've already made such circuit with GDT 5 turn primary : 19 turns secondary, but without zenners on mosfest gate  :-/O

Almost... Try this:

This will ensure that always 1 mosfet on each side are always on at the same time during the 600Khz square wave.
Note that you should run this at 10-20Khz square wave too as the gate capacitance of the mosfets and weak IO output drive of the MCU wont cope good with 600Khz square wave.

Only use this circuit for resistive loads like incandescent light bulbs.  Anything else might go haywire like inductive loads which may send huge voltage spikes and currents above the mosfets capabilities while switching.

« Last Edit: October 23, 2019, 04:50:46 am by BrianHG »
 

Offline Simon

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #10 on: October 23, 2019, 06:46:10 am »
Oh, all you wanted to do was dim a light? so just use a conventional dimmer setup. For 50Hz an opto coupler can give you any isolation desired.
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #11 on: October 23, 2019, 07:17:56 am »
Oh, all you wanted to do was dim a light?
Nope, I want control power of this light bulb using GDT (gate driver transformer) directly from MPU and additionally eliminate 50Hz flickering.
However, according do IRF840 datasheet its gate capacitance 832pF in prallel double of that, realized now now chance to get 600kHz switching speeds at such limited MPU GDT primary current ~2mA.
I've also no specs of random E core used for testing, so it require more sophisticated calculations and more tests.
 

Offline Yansi

  • Super Contributor
  • ***
  • Posts: 3930
  • Country: 00
  • STM32, STM8, AVR, 8051
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #12 on: October 23, 2019, 09:53:51 am »
Jeeeez... And how do you eliminate that flicker with this, as the bulb still operates at AC?
 

Offline Simon

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #13 on: October 23, 2019, 10:36:11 am »
If it's incandescent the time constant for the filament cooling is longer than the 100 Hz pulse period. You may get some flicker at very low brightness. I found i had to run LED's at a few hundred Hz to get rid of flicker. You don't need 100's of kHz, even 1 kHz is ample which again is well within the limits of an opto and other more expensive isolators exist if you can power both sides.

If you are paranoid about flicker why not rectify the supply and use a very small smoothing capacitor to not ruin power factor and then just use 1 MOSFET in a classic setup?
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #14 on: October 23, 2019, 11:05:25 am »
Brian: would be easier to just FWB rectify the gate signal. :P

There's no 50Hz flicker, it's 100Hz, and you can't get rid of that without energy storage.  You could make a VFD where AC comes in, gets rectified to DC, then is output as either infrequently changing DC (i.e., swapping polarities to reduce sputtering of the filament; low frequency AC, full duty cycle square wave), or high frequency AC (a few kHz will not be visible to the eye under any condition).  The DC supply could be varied (with a buck converter, or since efficiency really doesn't matter here given the incandescent light and apparently thermal application, a linear regulator), varying output.  In the higher frequency AC case, duty cycle could also be varied.

I once did this for strings of LEDs:
https://www.seventransistorlabs.com/tmoranwms/Elec_LEDs.gif

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #15 on: October 23, 2019, 06:53:34 pm »
There's no 50Hz flicker, it's 100Hz, and you can't get rid of that without energy storage.
Yep, of course it is 100Hz flicker - I've light sensor to test quality of the light around me ;)

Anyway, flickering is not a problem, because of we could rectify mains and use step down converter with inductor and yep AC mosfet switch instead of single mosfet, so (AC) marked this way, because of I want drive those mosfets using GDT.

I think, it  is time to make some real experiments with 100k resistor removed in my AC mosfet switch shown a few posts above  :-/O



 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #16 on: October 23, 2019, 09:22:45 pm »
I think you will discover something interesting if you dig into the gate charge.
Yep, probably now it is time to look closer into this, since I've quite nice shiny leds in antiparallel configuration with 1k resistor on my GDT secondary  >:D


GDT is quite "bulky" but it does the job and @ 600kHz with 4Vcc Attiny85 MPU 8MHz current on GDT primary is now limited for pin safety by two 100 Ohm resistors  8)


As we can see test LEDs are powered directly from MPU with galvanic insulation by GDT  :popcorn:

Plenty of charge to turn on/off AC mosfet switch, but now question is what switching frequency of this AC mosfet switch based on two IRF840 we can get?

Since, I do not know turns ratio in GDT (randomly chosen from old project) lets rectify its secondary and see maximum voltage we can charge capacitor to have some idea what turns ratio could be?  ???
« Last Edit: October 23, 2019, 09:25:53 pm by beduino »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #17 on: October 23, 2019, 10:06:17 pm »
Plenty of charge

Citation needed. ;)

Quote
to turn on/off AC mosfet switch, but now question is what switching frequency of this AC mosfet switch based on two IRF840 we can get?

And how low can the transformer go, say if in some strange reality it can't supply enough charge after all?

Say, did you have a go with any of those equations?  Like, C = Qg / Vgs(on) and t = RC?  How does t compare to 1/(600kHz)?


Quote
Since, I do not know turns ratio in GDT (randomly chosen from old project) lets rectify its secondary and see maximum voltage we can charge capacitor to have some idea what turns ratio could be?  ???

Mind that, if the secondary waveform exhibits ringing, you'll be measuring that, in addition to the turns ratio.

I suspect with ~280 ohms in series with the primary, it's going to be underdamped so this won't be an issue after all, but with that windup, you may have the opposite problem that you're running out of bandwidth, i.e., the output will be very soft and very not square at all.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #18 on: October 23, 2019, 10:38:49 pm »
Say, did you have a go with any of those equations?  Like, C = Qg / Vgs(on) and t = RC?  How does t compare to 1/(600kHz)?
IRF840 datasheet says Ciss=832pF Input Capacitance @ VDS25V f =1MHz VGS=0  , 1664pF capacitance for two mosfets in AC switch.
It should be at least +/-10Vmax/min on GDT secondary according to quick test I've made right now,
so we have something like
Q:  1.664e-8

However, there is no current limiting resistor on GDT secondary, only turns of copper wire, so its resistance should be quite low - cheap YATO multimeter says: R: 1.3 Ohm  ::)

This gives: t: R*C: 2.1632e-9   
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #19 on: October 23, 2019, 10:54:56 pm »
Now calulated something  like average current needed to provide charge for 10Vgs of those IRF840 @ 600kHz and it looks liek this if I do not missed something-I've assumed half of the time charging, second half discharging gate by GDT secondary:
Code: [Select]
(%i11) F: 600000.0;
(%o11)                             600000.0
(%i12) T: 1.0/F;
(%o12)                       1.666666666666667e-6
(%i13) T2: T/2;
(%o13)                       8.333333333333334e-7
(%i14) t: R*C;
(%o14)                             2.1632e-9
(%i15) T2/t;
(%o15)                         385.2317554240631
(%i16) Q/T2;
(%o16)                             0.019968
Looks like something about ~20mA average current.

Probably by lowering down this switching frequency a few times will help get enougth current for full open/close AC mosfet switch? :-\
« Last Edit: October 23, 2019, 10:57:38 pm by beduino »
 

Offline BrianHG

  • Super Contributor
  • ***
  • Posts: 8275
  • Country: ca
    • LinkedIn
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #20 on: October 23, 2019, 11:18:37 pm »
Brian: would be easier to just FWB rectify the gate signal. :P

There's no 50Hz flicker, it's 100Hz, and you can't get rid of that without energy storage.  You could make a VFD where AC comes in, gets rectified to DC, then is output as either infrequently changing DC (i.e., swapping polarities to reduce sputtering of the filament; low frequency AC, full duty cycle square wave), or high frequency AC (a few kHz will not be visible to the eye under any condition).  The DC supply could be varied (with a buck converter, or since efficiency really doesn't matter here given the incandescent light and apparently thermal application, a linear regulator), varying output.  In the higher frequency AC case, duty cycle could also be varied.

I once did this for strings of LEDs:
https://www.seventransistorlabs.com/tmoranwms/Elec_LEDs.gif

Tim

What I think the OP is trying to do, and not clearly explaining to us all, without a power transformer, drive a 12v lamp from 240v mains, pwm-ing the middle of the source sine wave so that it appears to look on average like  12v AC square wave 'power wise' to the filament.  If the op sticks with the 120v lamp, he can remove the flicker only if he wants the lamp to operate around 10% brightness...

     This means the bulb will still shine full brightness since when the source AC waveform is between that tiny period where it's within +/-12v, the mosfet is full on.  While the AC waveform is larger than +/-12v, the MCU chops up that power making it on average to appear to be a regulated + or - 12v over time.

     With such a short off time, the bulb will appear to be truly DC powered.  This project will also have the advantage of maintaining a true high power factor.

     The op needs the high speed on-off mosfet drive to really make tiny pulses during the almost 300v peaks.  He will also need a soft power-up cycle as a 12v halogen bulb drives massive current when the filament is cold.

     The 4x mosfet circuit I posted protects the one side off mosfet since a transformer drive cant keep both mosfets on at one time unless the OP want to loose half of his power.

     If they were fast enough, (I know they are for 1-2Khz, they turn off slow, it's been awhile since I used them.)  I would recommend using 2 photovoltaic optocouplers to drive the mosfet gates instead of a gate transformer.  Also, 2-4ma from the MCU will run the photovoltaic fine.

     As for our eyes, filament flicker isn't too bad, as well as for many cameras, but when used for high quality photographing/filming or special sensitive situations, this project will make the bulbs effectively appear DC powered.

The final cheapest solution means a bridge rectifier and 1 mosfet and the MCU driving the gate of 1 mosfet, though, the OP will be touching the mains with his MCU directly and I think that what's her trying to avoid.
« Last Edit: October 23, 2019, 11:49:29 pm by BrianHG »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #21 on: October 24, 2019, 01:44:42 am »
IRF840 datasheet says Ciss=832pF Input Capacitance @ VDS25V f =1MHz VGS=0  , 1664pF capacitance for two mosfets in AC switch.

I pulled out Qg specifically, because Ciss underestimates it by typically about a factor of 4.

Which... hm, you've already got to be careful here, because there's the original, created by International Rectifier:
https://www.datasheets360.com/pdf/-653557743118012551
which says 63nC max Qg and 1300pF typ Ciss.  The Fairchild (dated 2002) and the Vishay/Siliconix (2016) datasheets have very similar spec.  The ST (2002) version however shows significant improvement with 39nC max Qg and 832pF typ Ciss:
http://www.alphacron.de/download/hardware/IRF840.pdf

So it matters which one you're buying.  Not all manufacturers make exact copies of a multi-sourced part!

(FWIW, Vishay is the current owner of the original IR product line, including this part.)

If we use the average capacitance calculation, we get Ceff = (63nC) / (10V) = 6.3nF.  Note this is max, so it's not directly comparable to typ Ciss.  The Fairchild datasheet gives 42nC typ., or 4.2nF equivalent, 3.5 times the small-signal Ciss figure.

What's going on?

The gate capacitance charges as normal, and is in fact around 1.2nF, and doesn't vary much with voltage by itself.  The stinker is the G-D capacitance, which because the drain voltage has a huge swing (up to 500V), this capacitance is hugely amplified (Miller effect).  The only reasonable resolution is to look at the charge required to move the gate from 0 to 10V, and divide by that voltage swing to turn it into an average capacitance.

That's the justification for the Qg calculation. :-+

So if the gate looks like 4.2nF (or up to 6.3nF), and it needs to charge in a tiny fraction of 600kHz (1.6us), what resistance is required?

If we say 1/20th of a cycle maximum, or 80ns, in two time constants (t = 2RC), we need a resistance of about 10Ω maximum.

We likewise need to deliver a peak current of (10V) / (10Ω) = 1A, preferably even more to drive it faster.  Per transistor, so two in parallel also.

With a ~2:1 current ratio on your drive transformer, that's a demand of over four amperes from the poor ATtiny!

But it cannot deliver anywhere near that current, because the primary circuit has about 280 total ohms.  It can deliver about 14mA short circuit.

What could we achieve, then?

This in turn suggests that Fsw should be about as many times slower as the gate drive is.  So, if we can't do 4A but we can do 14mA, we should run at a frequency 4/(0.014) times lower, or about 21kHz.  And again, lower would be preferable, to save on switching losses.

Or we can add a gate driver IC, but we need to power it at some point.  If it's wired directly to the transistors, it needs an isolated supply of 9 to 15V.

Quote
However, there is no current limiting resistor on GDT secondary, only turns of copper wire, so its resistance should be quite low - cheap YATO multimeter says: R: 1.3 Ohm  ::)

Yikes!  Realize what kind of an assumption this is -- if the secondary had a source resistance of 1.3Ω period, then it must be that, no matter what resistance the primary circuit has, you can draw many times more power from the secondary.  You've made a huge power amplifier!  That's a very fishy conclusion, so the premise probably was in error. :D

The equivalent circuit for a transformer may look odd.  It's a chunk of wire, right?  Well, that chunk of wire is supposed to look like a large AC impedance (that's what the turns around the core does).  So it doesn't factor into our circuit.  Rather, the primary and secondary sides are connected by the ratio.  This looks more obvious if we draw it with no transformer at all -- but to do that, we have to scale all the values from one side, by the ratio, so that they are equivalent to what's on the other side.

Say we scale up the primary side.  Assuming 1:2 turns ratio, the primary voltage shows up on the secondary side as double, and the primary current shows up as half.  2 * 1/2 = 1, because energy is conserved of course.  This means that the primary resistances appear to go up by a factor of 2*2 or 4.  So the MCU and two 100Ω resistors looks like 8V logic (instead of 4V), with ~160Ω equivalent pin resistances, and 400Ω series resistors.  A total of almost 1kΩ!

This is why the gate drive will be, needless to say, a bit disappointing. :(

*Technically, any nonideal transformer can be replaced by an ideal transformer (if needed for 1:1 isolation) and three inductors, if you don't mind that one of them may be negative.  Obviously, negative inductances are hard to come by, and this is just an equivalent circuit.  But it's nice to know that equivalent circuits don't stop working just because a number comes out unrealistically. :)


Now calulated something  like average current needed to provide charge for 10Vgs of those IRF840 @ 600kHz and it looks liek this if I do not missed something-I've assumed half of the time charging, second half discharging gate by GDT secondary:
Code: [Select]
(%i11) F: 600000.0;
(%o11)                             600000.0
(%i12) T: 1.0/F;
(%o12)                       1.666666666666667e-6
(%i13) T2: T/2;
(%o13)                       8.333333333333334e-7
(%i14) t: R*C;
(%o14)                             2.1632e-9
(%i15) T2/t;
(%o15)                         385.2317554240631
(%i16) Q/T2;
(%o16)                             0.019968
Looks like something about ~20mA average current.

Probably by lowering down this switching frequency a few times will help get enougth current for full open/close AC mosfet switch? :-\

I'm not sure what numbers you used here.  Qg(tot) * Fsw does equal supply current for a gate driver IC.  The peak current is circa T2/t times higher though.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #22 on: October 24, 2019, 06:39:33 pm »
This is why the gate drive will be, needless to say, a bit disappointing. :(
Looks like your calculations are right for 600kHz, since I've just tried to connect to GDT secondary with only one zener 12V on AC mosfet switch gates and no resistance drop - cheap YATO multimeter says infinity resistance on all ranges, while when connected this AC mosfet switch gate to 4VDC we have ~3.8 Ohm resistance on mosfets drains, so ~1.9 Ohm per each IRF840 in series, so AC mosfet switch itself is fine after GDT test  8)

So, for the moment this is the only good news - we have working AC mosfet switch with DC Vgs >:D
 

Offline beduinoTopic starter

  • Regular Contributor
  • *
  • Posts: 137
  • Country: 00
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #23 on: October 24, 2019, 07:00:47 pm »
     The op needs the high speed on-off mosfet drive to really make tiny pulses during the almost 300v peaks.  He will also need a soft power-up cycle as a 12v halogen bulb drives massive current when the filament is cold.

I've infared light bulb 230VAC 80W-100W like shown below, and it is used in winter days to warm up my custom made sleeping capsule where warm air is prowided in aluminium pipes, so temperature controll needed and light flickering even at 100Hz can disturb my dreams, not mention that I'd like to experiment with different freqencies to study how it affect sleeping quality or maybe it also affects water molecules in air, so that is why I might be interested in high frequency switching, but now I'm shocked it doesn't work with pure MPU pins output @ 600kHz switching frequency, so maybe external transistors needed in totem pole configuration I've already used in other project  to be able  use lower resistors on GDT primary  ;)
« Last Edit: October 24, 2019, 07:04:16 pm by beduino »
 

Offline Simon

  • Global Moderator
  • *****
  • Posts: 18118
  • Country: gb
  • Did that just blow up? No? might work after all !!
    • Simon's Electronics
Re: Driving (AC) mosfet switch directly from MPU using GDT
« Reply #24 on: October 24, 2019, 07:35:38 pm »
light flickering at 100Hz disturbs your dreams? the light does not flicker.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf