Firstly, this is mostly a learning exorcise in analog design. No commercial or practical application as yet. I am trying to create a mild HV supply without using IC's or micro-controllers.

What I am attempting to do is to use a blocking oscillator to feed a more or less stable 40kHZ signal from the first stage, the BJT blocking oscillator, to a mosfet gate to switch an inductor for a higher current, unregulated (as yet) HV supply.

My problem: The positive gate signal swings too high, +28v (the neg. signal is okay at -7v). I'm concerned about damaging the gate oxide layer. I have added three zener diodes in an attempt to stop the gate from rising too high, placed before and after the gate resistor (18v+18v+4.7v). They are helping pull down the overshoot from ~44v to 28v, but I'm still still 8 volts too high. I'm stumped.

Additional notes: I tried a speed-up capacitor across the gate resistor, but this made the gate voltage swing way too high and uncontrollable. The Schottky diode on the base of the BJT squares off the waveform, making it close to 50-50%, which I like, but I don't have an understanding as why this is so. Unloaded, the circuit draws 158mA. I expect some fluctuation in the signal frequency under load, but experimentally, it ranges between ~30kHZ to 68kHZ, well outside of the audio range, which is my desire.

Attached are hand-drawn schematic, and a picture of my circuit.
Updated schematic to correct reversed source & drain.

 

Exactly how are you measuring this gate voltage?
Are you using a properly compensated 10X probe with its ground clip connected to the MOSFET source terminal?

 

I'm using a 10x probe on my scope. No, not using ground clip.

Correction: Yes, using a ground clip. Source terminal? The grounded lead, yes.

 

I'm using a 10x probe on my scope. No, not using ground clip.

Correction: Yes, using a ground clip. Source terminal? The grounded lead, yes.

Is the Zener connected directly to the gate and source terminals?
If so, then your should see the Zener clamping the gate voltage to near its voltage rating.
Can you post a clear picture of the voltage spike, with the vertical and horizontal sensitivity settings.

 

I was NOT using 10x on my probe after all... Following are 10X:

 

I'm new to oscilloscopes... And I don't know why I sacrificed for so long... I love it!

At 10x on the probe, am I, or am I not dividing the signal by 10?

Because at the 10X setting, all my problems go away

 

My suggestionto protect against the excess Vgs is not zener diodes but forward biased fast silicon diodes.
AND, for scope probes, now you see that there certainly IS VALUE for those more expensive probes.

 

At 10x on the probe, am I, or am I not dividing the signal by 10?

You are reducing the signal voltage by a factor of 10.

A 1X probe has a much larger capacitance and lower resistance which can alter high frequency/fast rise-time signals.

 

My suggestionto protect against the excess Vgs is not zener diodes but forward biased fast silicon diodes.

What's the basis for that recommendation?
Zeners respond in picoseconds or nanoseconds, so not likely significantly slower than a silicon diode.

 

So, multiply 2.56v times ten= 25v on the gate. The zeners aren't working. I've used low and high value gate resistors, and apart from different waveforms, it didn't help. I lowered the 5.5k gate resistor to 1k for example, thinking i wasn't getting enough current to activate the zeners, but no luck...

 

The problem may be in the physical arrangement of the circuit. THAT is often a problem with beginner PCB layouts. Or when a board house make a change without asking. Or just makes a mistake.

 

The zener is on top of the mosfet. I don't know how I can improve the layout, but I'm open to any ideas.

 

What is the observed voltage at the output of the 10nF capacitor?

 

Thanks for asking! It's +1.60v. (16v). Interesting. Does that help us?
Let me add, there's an 18v and a 4.7v zener from gate to souce.

 

Thanks for asking! It's +1.60v. (16v). Interesting. Does that help us?
Let me add, there's an 18v and a 4.7v zener from gate to souce.

Something is amiss.
How does 16V get amplified to 25V at the gate?
Don't see how that's possible with the circuit you posted.

 

Have you got the Probe Ratio set correctly in the scope hardware now that you have the probes themselves set to 10x? See page 2-3 in the User Manual on how to adjust for that. Once it's done the readout will automatically correct for the probe and you won't need to do any multiplication of the readings in your head!