Capacitance multiplier problem

[Kari]

Hi,
I've just built a powersupply as the picture below describes. Before i actually bought anything i run the schematics on MultiSim and got a output voltage of 40V which is what i want. However i reality i get 50V out without any load and 45V with load. MY question is: How do i regulate the output voltage to 40V? I really dont want to buy a now transformer (The one i'm having now is a 800VA, 2x42V out).

Thx,

Kari

[millwood]

one way is to put a 40v (actually 41v) zener to the base of the transistors.

The problem you have with the multiplier is that without load, there is no current going through the transistors' emitters, thus no current through their bases. thus no voltage drop over the 1K resistor. Which means your output voltage is just input voltage -0.6v (Vbe for the 3055s).

By adding a zener there, there is always a flow of current going through the 1k resistor. That makes it the simplest form of a regulated PS.

then it is no longer a capacitance multiplier,

maybe you want to think of a reason why you should be concerned about the no-load voltage.

[Christer]

A transformer usually/always has its voltage specified at the
rated max load and will thus give a higher voltage with less
or no load. If your load varies, like in a class B/AB amplifier, you
will have to live with varying voltage due to this or regulate
the voltage, something seldom done in power amps. If you
have a constant load, like in a class A amp, then you will get
some fixed voltage (+ripple and mains variations, of course)
which depends on how much current you draw compared to
the rated max current of the transformer.

Millwoods suggestion is one way, but as he said himself, it
is no longer a cap. multiplier. I think, but haven't really tried
to think about it, that a zener in series with the 1k resistor
would work to drop the voltage and still have a cap multiplier.
However, in this case the voltage would still follow the voltage
variations of the transfomer due to load variations. It might
perhaps work for a class A amp though. You could also go
all the way and build a proper regulator. The problem with
all of these solutions is that you will have to burn away a
lot of power in these transistors. If it is possible, I think it
would be better to see if you can redesign the amplifier so
it can handle the higher supply voltage.

BTW, the parallel connection of the transistors won't work well
since there is nothing that makes them share the load equally.
A small emitter resistor on each of them will help doing this,
but has a negative impact on regulation, of course.

[Kari]

Thx for the help!
I'm actually building a DoZ and i got a little worried about the heat generated at 45V and 1.7A. While i was testing the transistors got very hot (2N3055), i could not touch them at all (well maybe for 0.1secs). I'll put the suggested modifications into MultiSim and check the results.

[traderbam]

Not sure what the benefit of this "capacitance multiplier" is meant to be. I can see a lot of drawbacks, some of which you have experienced with the real circuit.

If you want a power supply for an amp or something then you'd be better off:
1) putting a 42V zener across the base capacitor (as Millwood suggested)
2) turn the parallel transistors into a darlington.

or better still use a regulator IC driving a transistor (you'll need to find a suitable circuit: look up LM317 on the net).

A transformer with 43Vac output will create 61Vdc when rectified. This is rather more than you need and as a result you will waste power regulating to 40V. If you need to use this transformer you will have to use a big heatsink. Roughly, every amp drawn will cause 20W of wasted power, power dissipated by the transistor. Work out the heatsink size so that the transistor case remains below 70C, preferably below 60C, in worst case loading conditions.

[Cobra2]

Sell me your trafo (if it is the ELFA "audiophile" potted type)...

Arne K

[bezobraznia]

I suggest You build amp like i did...
I have 350VA transformer - 43V with load at 2.5A for channel
When i make first experiments with DoZ i use 2N3055 that blow-up with this current situation, then i maker capacitance multipler .. and with help of resistor there i get 39V with 2A for channel...
The only thing that i must say is capacitance multipler go so greasy sound... ... then i use KD503 2Mhz-Chech Transistors and shunt Capacitance Multipler, then with 43V / 2x2,5A load i get realy great sound and more powerfull - almost 36-37Watt... without distortion at highs and without need of more Ampers
The next step is i use MJ15003 They are so fast and tomorrow i will make experiment with current like this - direct from capacitor - 34mili-farads - Sprague powerlytic.
Best Reggards!
Niki/Bulgaria

[Mr Evil]

A resistor in parallel with the base capacitor will reduce the voltage by forming a potential divider with the 1K resistor. Make it a variable resistor and you can adjust the voltage.

BTW, it is greatly beneficial to use Darlington pairs instead of single transistors, since the capacitance is effectively multiplied by the transistor gain. You can get monolithic Darlingtons almost as cheap as a 2N3055.