MOSFET Headache

[Speakerguy]

There are CMOS 555 timers that claim astable operation to 3MHz. I've used one up to 1MHz successfully and the signal looked good, I don't doubt the 3MHz claim. The part # was LMC555.

ETA: The cap size for this frequency range is teeeny. I was a dummy and put the traces to my 22pf cap over a ground plane. The added capacitance threw off all the normal 555 equations and I had to tweak capacitance and resistances in the 555 ckt by hand.

 

[turbobill]

Ok, I'm taking another run at this thing.

Is your inductor really just three turns of #4 wire wrapped around the 3" bobbin?

I would guess that's about 100 nanohenries, and the current from 12V would increase at the rate of about 120 amps/microsecond. Another way of saying, if you turn it on for more than a microsecond, bang!

Where's your fuse? (I mean other than the MOSFETs). Exploding batteries aren't pretty. Acid burns are forever.

The inductor is actually a 250:1 air-core transformer with 4 turns of #4 wire and 1000 turns of #22 magnet wire. I've got the modulation down to about 2MHz which is a pulse of about .26 microseconds by using a TS555 timer that is rated at 2.7MHz.

The timer will drive a MCP1406 MOSFET driver with output of 6Amps to the gate bus with 20hm: resistors for each gate of the ON brand NTB125N02R MOSFET.

So what do you guys think? Are we getting closer? Will this thing work?

The object is to produce pulsed DC from the transformer's secondary.

Thanks,
William

 

[Roff]

Have you considered the capacitance between windings on your transformer secondary? I suspect that the resonant frequency of the secondary will be less than 2MHz.

 

[turbobill]

I suspect that the resonant frequency of the secondary will be less than 2MHz.

Have not considered that. How do you find resonant frequency of the secondary?

 

[Roff]

Have not considered that. How do you find resonant frequency of the secondary?

Do you have an oscilloscope? How about a signal generator (square and/or sine waves)?

 

[turbobill]

I have a scope but not a signal generator.

 

[Roff]

I have a scope but not a signal generator.

Set up a 555 timer to run at about 100kHz, and 40-50% duty cycle (not critical). Leave the primary of your coil floating, and connect one end of the secondary to the ground of your 555 oscillator. Connect the 555 output (pin 3) through a 10k resistor to the other end of the secondary. Put your 10X scope probe on the junction of the secondary and the 10k resistor (probe GND to circuit GND). If the Q of the secondary is decent, you should observe ringing (damped sinusoid) on the scope. The frequency of the ringing is the resonant frequency. Be sure you use a 10X probe, because a 1X probe will put a lot of additional capacitance across the coil, lowering the resonant frequency.
You may have to adjust the frequency lower to see the ringing.

 

[turbobill]

The frequency of the ringing is the resonant frequency.

So then we should run the primary at the resonant frequency of the secondary?( I'm a total noob at this )

And what about this:

I would guess that's about 100 nanohenries, and the current from 12V would increase at the rate of about 120 amps/microsecond.

If the pulse is too long won't we get too much current?

 

[Roff]

So then we should run the primary at the resonant frequency of the secondary?( I'm a total noob at this )

And what about this:


If the pulse is too long won't we get too much current?

You're getting too far ahead. You need to check the resonant frequency first. Then we can talk about whether you can use this scheme. As I said, for this test, you leave the primary floating.

 

[turbobill]

Ok. Thanks a bunch. I'll do that ASAP.

 

[turbobill]

Is this what we're looking for? See attached pic of scope:

The Time/CM is set on 10:mu:s, the Volts/CM is set on .5

 

[Roff]

Is this what we're looking for? See attached pic of scope:

The Time/CM is set on 10:mu:s, the Volts/CM is set on .5

Yes, but slow down your oscillator by a factor of about 10 (use 10X the capacitance). Then take another picture.
What supply voltage are you running on the 555? 0.5V peak seems a little low. Is your scope set on 0.5V/div, with a 10X probe, yielding 5V/div actually?

 

[turbobill]

What supply voltage are you running on the 555? 0.5V peak seems a little low. Is your scope set on 0.5V/div, with a 10X probe, yielding 5V/div actually?

Supply voltage is 12v. Scope is set on 0.5V/div. Yes, using the 10X probe.

 

[turbobill]

Changed C1 from 0.01:mu:f to 0.10:mu:f, this is what we get. Time/div is set at 0.1ms. What is it suppose to look like?

 

[Roff]

Changed C1 from 0.01:mu:f to 0.10:mu:f, this is what we get. Time/div is set at 0.1ms. What is it suppose to look like?

Can you post a picture at 10usec/div?

 

[turbobill]

Ok, these pics are both at 10:mu:s, TransFreq3 is at 0.2 Volts/div and TransFreq4 is at 0.5 Volts/div.

 

[Roff]

I don't understand what that waveform means. Show us the waveform on pin 3 of the 555, at 10usec/div and at 100usec/div.

 

[turbobill]

I don't understand what that waveform means. Show us the waveform on pin 3 of the 555, at 10usec/div and at 100usec/div.

Ok, those two pics were with C1 = 1.0:mu:f. I'm going to drop back to the 0.1:mu:f cap and give you the pin 3 output with out the transformer connected, at 10usec/div and at 100usec/div. Then a series of pics at those same values with the transformer connected.

 

[turbobill]

The first pic is at 10usec/div, the second at 0.1msec/div( I think that's 100usec ), my scope doesn't have 100usec. This is just pin 3 output with out transformer.

 

[turbobill]

Ok, these pics correspond to the ones above but WITH the transformer secondary connected.