[TimS]

I’m currently reading about a design for a Shunt Regulator from DIYaudio (http://www.diyaudio.com/forums/power-supplies/153715-my-take-discrete-shunt-voltage-regulator.html) by ikoflexer which sounds promising; it only uses a handful of parts and boards will be available for around the US$8-10 mark.  At the moment the regulator design has a "Normal" voltage version recommended specs of:
* output voltage 8V to 28V (plus or minus)
* max current limit: 4A (with adequate heat sink)

And a "Low" voltage version recommended specs:
* output voltage 3V to 28V (plus or minus)
* max current limit: 1.5A - 2A (with adequate heat sink)

What I’m interested in is that there is potential for the regulator to reach voltages of 60 V with beefier mosfets and large heat-sinks so am I right in thinking that there is a chance it could be used as a shunt regulator power supply for my AKSA 55W N+ that requires (from memory) +/- 36 Volts per channel? 
I'm no EE, so wonder what the inherent problems are with using a shunt regulator for a power amp – if it's possible it would be one hell of a  ‘tweak’ !!

For those who don't know exactly how a shunt regulator works (as I didn't) ikoflexer wrote the following after I queried if the regulator might work with the amp:

".....I think it would work. You'd need huge heat sinks though, to dissipate 55W continuously. This is one of the things about a shunt regulator, it needs to burn all this energy to have the current supply when needed by the load. Basically, if I can draw an analogy, think of it as a pipe that splits in two, one pipe goes to the load, which in your case is the power amp, and the other pipe goes to the shunt mosfet. The total "water" from the wall gets distributed to the load if it is needed; when the load doesn't "consume" this water, it gets redirected to the shunt mosfet, the other pipe. As soon as the load demands it again, the circuit redirects the water from the shunt mosfet to the load. But all the time, whether your load/amp consumes or not, the water flows. Wasteful, for sure. But able to supply on demand better than other circuits.

So, it's really like building another amp."

[AKSA]

Tim,

I concur with Ikoflexer's comments (his name is Relu, we have corresponded, very bright man).

Actually, it's like building a mini Class A amp to power your Class AB amp.....  how cool is that?

In truth, it's not cool, as you have to dissipate the heat, and if you choose to use no fan, you must use another two heatsinks of the same dimensions as the one which presently handles two AKSA55 modules.  So it's a big project thermally, tripling your heatsinking.

BUT, would it sound good?  I vaguely remember doing a shunt power supply for the AKSA55 some years ago, and deciding the cost/benefit was adverse.  This is not to discourage you, I actually use a shunt regulator of sorts on the GK1, so I'm in agreement with the general principles.  My thoughts are that your invested capital and time would be better off going for an upgrade, but then, I would say that I guess!!

Good luck, and keep us posted, we are very interested in your project, thanks for informing me!

Cheers,

Hugh

[TimS]

Thanks for your reply, Hugh.  I'm still interested in trying it but I do have a couple of questions off the top of my head:
I've read (if I've understood correctly) that the regulator needs an extra 10 to 15 volts more (from the transformer) than the desired output voltage to make it work, so that would mean using another higher voltage transformer than the recommended 25-0-25 or is it possible to use the same 25-0-25 transformer and run the amp at a lower voltage? 

Also, rereading the AKSA 55 assembly instructions, after the shunt regulator has been connected to the amp modules at the +/- 36V connections, would I still need to connect the earth connection on each amp module back to the star earth point on the power supply PCB?

Tim

[AKSA]

Hi Tim,

Yes, you would need to connect both regulators to Star Earth, no way out of that one.

No, 10-15V is too much.  You need only sufficient extra voltage to accommodate the maximum sag of the amp at full tilt.

Let us assume that a 55W AKSA will do say 30V peak into 6R.  This would mean that the supply would draw a maximum current theoretically of 5A, about right for a single pair of output devices.  Thus, the shunt reg should be designed for 5A continuous.

The shunt element will pull the conjugate current;  that which is NOT drawn by the amp, with the total being 5A.  This means that we need only a small margin over the 36V nominal rail voltage, but in fact at full power the standard supply would drop to about 33V under load.  So, giving the shunt element a full 7V of margin, we must use a raw supply of 40V, call it 42V as this is a 30-0-30Vac trafo.


Now, with 42V, we can sink to 36V under load with a six volt drop, which will be across the shunt resistor.  The shunt element will pass up to 5A at this 36V at all times, falling with increasing amp output.  So, this means it must be set up to dissipate 180W PER CHANNEL in the shunt elements alone.

Now, this is an improvement on dropping 10-15V, BUT, do you see the problem?   You will be dissipating 360 watts through the shunt elements, spread over four devices - though more properly eight devices, giving you 45 watts at each device.  In addition, you will be using a resistive shunt element to drop this seven volts;  this will dissipate a constant 5A x 7V =  35watts, and you will need four of them, too.

I would very strongly recommend that you use mosfets for this task.  They are VERY robust thermally.

Cheers,

Hugh

[RonR]

Hi Tim,

As a sort of half-way house, there are some interesting developments in the Naim-tweaking world at the moment on PFM, where 2 power supplies are being used per Power Amp Channel.
By splitting the early stages off onto their own Traffos and PSUs (Being Class A and low current, these could easily be shunt regulated), the "voltage sag" in those stages is eliminated, and the high-current output stages can be left unregulated.
I haven't tried this yet, but reports from others are good so far.

Cheers,

Ron.