Two mosfet oscillator

[MileHigh]

GL:

If you have a scope and some resistors and a 555 timer or a signal generator you can measure the inductance of your coil.

MileHigh

[Groundloop]

GL:

If you have a scope and some resistors and a 555 timer or a signal generator you can measure the inductance of your coil.

MileHigh

MileHigh,

I have an excellent function generator and a not so excellent o-scope.
Function generator goes up to 20MHz. O-Scope up to 30 MHz.

GL.

[MileHigh]

GL:

You would be entering uncharted territory if you made the measurement.  It's just a basic measurement that people usually do in their first few labs when they study electronics.

Just set up a typical NPN transistor energizing a coil.  You know the one that everybody knows, think tech porn a la xxxxxx.

Instead of the charging battery just put a resistor in that position.  (Also, you don't need the diode and you don't want to use it.)

Look at the waveform on the scope and measure how much time it takes for the waveform to decrease in amplitude by 63%.

That's the L/R time constant.  So just solve for L.

MileHigh

An editorial comment:  This relates back to the flack from Aaron and Mr. Tech Porn and what 'free energy research' is all about.  In five years of following the forums on and off, I have never seen a single person do the test mentioned or even discuss this test.  I read a lot of comments saying, "I wish I had an inductance meter" and stuff like that.  Yet the test is the first thing that you are supposed to do on the bench as part of the learning process about how inductors work.  We are talking the second lab in Electronics 101 here.  It's just a reminder of how wide the gulf is between the real world of tech and 'free energy research.'  Sometimes it's worth reminding yourself that there is always room to learn and grow and to not pretend you are something that you are not.  No offense to anybody intended, just a polite reminder about what is really going down.

[BEP]

GL,

Try this: http://ironbark.bendigo.latrobe.edu.au/~rice/lc/

It worked very well for me and it is more accurate than my high-end Fluke meter 

I'm looking for the 'Cascode' example.

Your circuit is visually closer to a cathode follower with the odd parallel anode setup. I was speaking about the way I believe your circuit functions.

i.e. Similar to the turn-off of an SCR, your Q's turn off when the voltage drop decreases across the other Q. So, it is a voltage oscillator.

[Groundloop]

MileHigh, WW,

Thanks for the information and links.

GL.

[poynt99]

@All,

So what to test next?

GL.

I guess that depends on what you would like to accomplish with this circuit. Do you want to perform similar measurements to what Rosemary did? Did you insert a load resistor to measure? Did you consider replacing the coil with an inductive resistor of 10 Ohms?

Just a note: Since my version of Rosemary's circuit oscillates with only one MOSFET, and produces essentially the same wave form as those the RATS posted, I am not convinced that your circuit is the same as hers. I'm not sure she was able to get hers to work down to 12V either; I know in my sim, 24V is about the lowest it can go.

.99

[poynt99]

GL.

Your circuit simulated.  

Shown running on 1.5V supply. Runs also on 12V supply. The scope is inverted, but otherwise same wave form.

.99

[poynt99]

I have more to show you.

Modify your circuit slightly as shown in this next schematic. We are replacing M2 with a diode (I used MUR160, but you can use any, even a 1N4148 or 1N914) and a 1nF capacitor.

The resulting wave form is more sinusoidal, (i.e. no non-linearity on the one excursion as before) but about the same frequency and amplitude.

See if this works for you.

.99

[Groundloop]

I have more to show you.

Modify your circuit slightly as shown in this next schematic. We are replacing M2 with a diode (I used MUR160, but you can use any, even a 1N4148 or 1N914) and a 1nF capacitor.

The resulting wave form is more sinusoidal, (i.e. no non-linearity on the one excursion as before) but about the same frequency and amplitude.

See if this works for you.

.99

.99

Did try with a BY255 diode instead of my Q2 mosfet.
I got the attached oscillation that did not run for more than a second.
And I only got that by fast connecting  and disconnecting the bias battery.

GL.

[poynt99]

What supply voltage was that? 1.5V?

Did you try with 6V, and 12V?

Sounds like it is close to wanting to oscillate. Try a 2.2n instead of a 1n cap as well.

.99

[Groundloop]

What supply voltage was that? 1.5V?

Did you try with 6V, and 12V?

Sounds like it is close to wanting to oscillate. Try a 2.2n instead of a 1n cap as well.

.99

.99

Supply voltage was 5,42 volt.

When I connected a 3,3nF ceramic capacitor as shown in the attached drawing, then I got a stable oscillation.

GL.

[poynt99]

.99

Supply voltage was 5,42 volt.

When I connected a 3,3nF ceramic capacitor as shown in the attached drawing, then I got a stable oscillation.

GL.

That's good news!  

How does the wave form look now?

[Groundloop]

That's good news!  

How does the wave form look now?

See above.

[Groundloop]

.99

Now I have removed the diode and the oscillation is stable.

GL.

[poynt99]

OK, that's cool. Not too surprising, but cool anyway.

So what do you make of these last two circuits in comparison to the two-MOSFET version? What is your present operating analysis on the single MOSFET? How much has the circuit really changed?

[Groundloop]

OK, that's cool. Not too surprising, but cool anyway.

So what do you make of these last two circuits in comparison to the two-MOSFET version? What is your present operating analysis on the single MOSFET? How much has the circuit really changed?

.99

The newest circuit I posted above is a Colpits oscillator.

But instead of using a normal NPN transistor biased to half the Ub at the base
and ac coupled to ground with a capacitor, we use a mosfet that is negative biased at
the source pin. Then we have the capacitor ac pulse the same pin so that
the oscillator will run.

GL.

[poynt99]

OK.  

And the circuit hasn't really changed that much. M2 never was active in your original configuration; only its D-S capacitance was being utilized. The body diode has an affect as well, but more so at higher voltages it seems.

Now, you say that the M1 has a negative bias on the source. That's somewhat correct, but the more correct phrasing is that the M1 MOSFET has a positive V_GS bias, which of course is what an N channel MOSFET requires to operate...as I've been saying all along.

So sorry to break it to you, but technically speaking, there never was anything operating with a negative bias (V_GS); the active MOSFET always is the one with a positive V_GS bias. And the bias on a MOSFET is always determined by the Gate voltage relative to the Source voltage, not vice versa. 

There is nothing "unique" about this oscillator, but it is neat nonetheless.

btw, I removed the diode from my sim, and it still oscillates as well. But now the clean sinusoid is looking more like it did before, slightly nonlinear on one excursion.

[Groundloop]

OK.  

And the circuit hasn't really changed that much. M2 never was active in your original configuration; only its D-S capacitance was being utilized. The body diode has an affect as well, but more so at higher voltages it seems.

Now, you say that the M1 has a negative bias on the source. That's somewhat correct, but the more correct phrasing is that the M1 MOSFET has a positive V_GS bias, which of course is what an N channel MOSFET requires to operate...as I've been saying all along.

So sorry to break it to you, but technically speaking, there never was anything operating with a negative bias (V_GS); the active MOSFET always is the one with a positive V_GS bias. And the bias on a MOSFET is always determined by the Gate voltage relative to the Source voltage, not vice versa. 

There is nothing "unique" about this oscillator, but it is neat nonetheless.

btw, I removed the diode from my sim, and it still oscillates as well. But now the clean sinusoid is looking more like it did before, slightly nonlinear on one excursion.

.99

You probably meant to say "D-G capacitance was being utilized".

I assumed that I did not have to go into detail about the bias and that it was quite clear that the n-type mosfet needed a more positive level
on the gate with regards to the source. I was referring to the ground level in my circuit as a reference point.

But all in all, I agree that we have fully explained why this oscillator works as it does, and as long as there is a very high resistance
from the source to ground, then there will be very little chance of transferring any serious current through the mosfet.

GL.

[poynt99]

.99

You probably meant to say "D-G capacitance was being utilized".

There are several interlead capacitances, but I meant total Thevenin D-S capacitance.

I assumed that I did not have to go into detail about the bias and that it was quite clear that the n-type mosfet needed a more positive level on the gate with regards to the source. I was referring to the ground level in my circuit as a reference point.

It is necessary actually. I interpreted your statements of "negative bias" as meaning negative VGS. As ION says, the devil IS in the details.

[muDped]

...
But all in all, I agree that we have fully explained why this oscillator works
as it does, and as long as there is a very high resistance from the source
to ground, then there will be very little chance of transferring any serious
current through the mosfet.

GL.

Aye, the "resistance from source to ground" provides
some of the feedback which sustains the oscillations.

This circuit configuration (Grounded Grid or Grounded
Gate) is quite popular as a stable RF Power Amplifier
in certain applications.  The Signal Input is at the
Cathode or Source and the Signal Output is at the
Plate or Drain.  It is a low impedance input circuit.

In order to realize any substantial Drain Current in
this configuration the Input Source Current (Signal
Current) must be equally high.  The input current
and the output current are essentially the same,
however, there can be a large voltage difference
(amplification) in the two signals.