Switch realization for this idea

[rtarbell]

I am posting a similar circuit to what I read in an online thesis (I wish I could take credit for this idea, but I can't). Although the author does not go into much online detail, I was wondering how to make such a switch to realize this function (I have elementary knowledge in switch realization - I don't know much beyond the basics and my couple of text books):

In the schematic, the transistor is supposed to be the output device of a class AB audio amp (I didn't draw the other half, but it would be symmetrical). Under normal conditions, the bottom of the capacitor is connected to ground, and thus the cap is charged up to ~ +50V via the supply through the diode.

When extra voltage is required, the bottom of the capacitor is disconnected from ground, and then reconnected to the anode of the diode, so that the collector of the transistor sees a higher voltage. The capacitor acts as a "second" supply in series with the first 50V supply, similar to a rail switching amp without the extra cost for more power supplies to create a "real" upper tier voltage supply.

==> But how to create the switch?!?! I think it would be a SPDT type of switch, but I'm lost as to a simple way to do it.

[elec-tron2]

Good idea !

I think that two power Mosfet can achieve this. A typical (mechanical switch like a relay) can not be used as it is far too slow.

This schematic can work. Switch M1 by applying correct gate voltage to put C1 to gnd. Then, if you want to have C in parallel with D, release M1 and switch M2. But never switch M1 and M2 at the same time.

Need to try some simulations to see how it work.

I think that power dissipated and Rds is the major problem for reliability.

[Elvee]

Hi

It looks very much like the car booster amps from Philips (now NXP):
http://www.nxp.com/acrobat_download/.../AN10327_1.pdf In principle, any transistor, Bjt, MOS or Igbt could do the job.
LV

[elec-tron2]

It's look like in some way a DC-DC converter.

A good place to start
http://en.wikipedia.org/wiki/Boost_converter

[rtarbell]

==> "It's look like in some way a DC-DC converter."


Yes, very much so... I just thought that it might be easier to implement than a true DC-DC converter because no inductors are used, just a "simple" capacitor that acts as a second voltage source in series with the first source at the right time.

The idea is that when the output of an amp approaches the power supply rail (and thus draws a lot of power to the load), the capacitor doubles the rail. Now, a capacitor is not an ideal voltage source as it discharges, so hopefully the output of the amp would diminish before the capacitor completely discharges, or else the amp will clip.