What the hell am I doing wrong here?

Why wont that bl00dy BD140 conduct? It wont conduct if I flip it either.

If I detach my multivibrator circuit from the BD140 and attach a signal generator instead then the inverted boost converter works great and I get about -20V out.

Your astable multivibrator alternates between a positive volt level and zero (as seen at R3).

However for your PNP transistor to turn on, it needs a negative volt level at the base, because you are running it on a negative supply.

To get the converter to run, apply zero at the top of the astable, and apply V- at the bottom.

You should keep the PNP flipped around from its orientation in your schematic. The PNP icon is supposed to have the arrow pointed in the direction of current flow.

BradtheRad said:

Your astable multivibrator alternates between a positive volt level and zero (as seen at R3).

However for your PNP transistor to turn on, it needs a negative volt level at the base, because you are running it on a negative supply.

To get the converter to run, apply zero at the top of the astable, and apply V- at the bottom.

You should keep the PNP flipped around from its orientation in your schematic. The PNP icon is supposed to have the arrow pointed in the direction of current flow.

Click to expand...

That last tip is a good one. Makes sense in hind sight but I haven't been paying it much attention until you pointed it out.

- - - Updated - - -

I officially give up with discrete transistor multivibrators!

They might be OK if you a doing a LED flasher or some other novel circuit.

But trying to use them to create specific components such as these is just too much hassle.

Think I will leave this high level design to the experts and just stick with 555s where all the hard work is done!

In multisim I can't even get a transistor multivibrator to relaibly generate negative 18V from positive 18V presumeably due to 18V exceeding the maximum voltage rating on the base of the transitors. In any event I spradically get a circuit errors that can't be resolved. Sticking voltage dividers on the bases then messes up the oscillation.

Where as some 555 can take a supply voltage of 18V.

Even if you got the transistor biassing correct, the 1N4007 is not suitable for high frequency rectification, you need a device with faster switching speed.

At first glance, if all you are trying to do is create -18V, why not use an NPN transistor (or better still a MOSFET) , the same positive rail for the output stage as the multivibrator uses and keep the same output configuration. How much current do you need to draw from the -18V ?

Brian.

It is commendable that you have the motivation to try to do-it-yourself.

Seeing that you were on the right track...

and since you were not terribly far from success when you said you give up...

and since it would be a shame if you really do give up...

I'll post my screenshot of an animated simulation of your schematic with the -12V supply (because I only figured out how to get it to work a few hours ago). It includes my recommended revisions.

Notice that the output stage has -18 V on it.

The leftmost oscilloscope trace shows output at the astable multivibrator.
The middle trace is volts and current in the coil.
The right trace shows final output V.

The red color represents negative polarity. Green represents positive (by preference of the author of the simulator, Paul Falstad).