I have designed an Astable multivibrator to produce an output square wave of 125 kHz, however when a load is applied the output fails. Does anyone have any advice for a driver circuit or solution?

Thanks!

 

Post your schematic please so that forum can help you. What is the load
you are using ?

Regards, Dana.

 

Psu not enough current supplied, output transistors not powerful enough ,,,,post your circuit please ..

 

The load was applied after a resonating charging circuit which works if a function generator supplies the source but not the oscillating multivibrator.

Attached is the circuit of the multivibrator, any load would naturally be attached to either collector. I have looked into other topologies but get either the desired frequency or current output.

 

any load would naturally be attached to either collector.

What is the nature of the load?

 

Wireless charging coils which need to resonate at above 100 kHz, after which AC rectification so I can power a supply.

 

Source power supply*

 

Wireless charging coils which need to resonate at above 100 kHz

OK, let me try again...

What is the impedance of the load?

 

I tried it with a 1kOhm load and no output was produced.

 

Your circuit pulses almost 0.5 A through each transistor (assuming the other collector resistor is 19 ohms and called R2 behind the obscuring block, making a 50% duty cycle). That is a lot, even at 50%. The base current is over 16 mA. This is both too high for long term reliability and too low to saturate the transistors with 0.5 A collector current. I can see why/how the circuit is functioning, but it looks delicately balanced and not good for long term reliability.

Separate from that, the immediate problem is that the width of each half-cycle is determined in part by a collector resistor. Anything attached to a collector (going to either Vcc or GND) in addition to the resistor will change the current flowing through the collector AND change both the current available to drive the other transistor's base low through its timing capacitor, and change the impedance pulling the timing capacitor high to reset it.

I strongly recommend abandoning this circuit and changing to a more straightforward low power squarewave oscillator followed by a power driver that is appropriate for whatever the load is. A 250 kHz crystal oscillator is 93 cents at Digi-Key - in ones. If you want adjustability and don't need crystal-accuracy, a CMOS 555 will work. The datasheet has a 50% duty cycle app schematic. A bonus is that it uses only one resistor and one capacitor to set the timing for both half cycles, so you con't have to match components to get 50/50.

Edit: What is the expected coil current, peak-to-peak?

ak

 

The loads are inductive. They will make current flow long after transistor switch-off.

 

Some good points made there I have clearly overlooked. I don't require crystal accuracy so maybe a low power 555 timer with a power driver will work. However the load will still be inductive, could you recommend any power drivers that don't include either BJT's or FET's if transistors are to be a problem?

 

What is the inductance of the coils?
How fast do you want the inductance to be discharged?
Is the intent to transfer energy to the secondary during charge or discharge or both?

[EDIT] It is hard to tell from the original circuit and words if the intent is to actually run the coils at their natural resonant frequency. If that is the case, the circuit needs to be quite different from what would be used for forced behavior.

 

At 250 kHz, I think relays are out, so you're stuck with solid state devices. The world produces billions of switching power supplies per month (week?), so driving inductors with transistors has pretty much been solved.

1. What is the voltage source for the load coil(s)?

2. What is the full load coil current?

ak

 

They are identical at 7uH. I'm not sure what you mean by "Is the intent to transfer energy to the secondary during charge or discharge or both?" because it is the AC presence that enables the power transfer between the coils.

It's not 250 kHz but just around 100 kHz if that helps any further.

The voltage source for the load and subsequent circuitry will be purely from the primary ie. the transmitting coil.

Full load I assume will be the maximum the coils are rated for which is around 2A, however, I can't imagine a current neat that value will be reached.

 

Not sure why this is so difficult. The intent of your circuit is to switch current through an inductor that is the primary winding of an air-core transformer. That means one end of the inductor is connected to a switching circuit.

1. What is the other end of the inductor connected to:

a) Ground
b) a non-zero voltage source
i) what is the DC value of that voltage?
c) another switch of the opposite phase

2. What is/are the value(s) of the voltage source(s) powering the primary coil driver circuit?

ak

 

Yes the intent is to switch current through the primary in order for it to induce in the secondary. It shouldn't necessarily mean that one end of the inductor is connected to a switch though unless you are referring to a switched mode power supply? I intend to resonate the primary coil using a tank or series RLC circuit. The input frequency to tank or RLC circuit needs to be matched to the frequency at which it resonates or there will be a mismatch, this is the need for an oscillating circuit of around 100 kHz.

So to answer your question more directly yes there is a ground connection on one end of the inductors.

10V should be an adequate DC voltage source.

 

Below is the LTspice simulation of a self-resonating push-pull driver that drives a series LC circuit at it's resonant frequency. It avoids having to match the oscillator frequency to the LC resonant frequency.
It also has the advantage of always oscillating at the resonant frequency, even if the presence of the receiving coil affects this frequency.

The peak inductor current is about 5A at ≈105kHz for the components shown.
You can reduce that current, if needed, by lowering the supply voltage or adding some resistance in series with the output.
The values of R2 and R3 may need some tweaking of values to get the real circuit to oscillate.

 

A slight snag: shoot-through .

 

Try this circuit.
Select value of capacitor C4 for maximal efficiency.