Graham Slee Power Amplifier

enjoying following the development of this. what's next...

Got the PCBs back from the manufacturer and now I'm working on the case. I had always wanted to make an amp small and compact enough to fit on the back of a small monitor loudspeaker, and saw it as a challenge to use the same extrusion we have made here in England for many other items in our range (you are allowed to use it on the hi-fi rack/shelf as well! ).

The heatsink for the time-being is a "donor" item bought off the shelf. I hope to be obtaining "cast" lengths of a similar extrusion having slightly lower profile, and that runs full length of the case, and have it machined so that no fixings are seen. Colour will match the case (natural/silver anodise - don't worry, it will dissipate the heat).

The PCB is inverted (in the top of the case) with components under-slung apart from the power transistors. I guess some could see it a drawback in that the connectors are mounted upside down but I'm not going to let that put me off.

With professional connectors and balanced input (which can be wired unbalanced) it could be attractive to the professional market too.

Graham should you offer a version w/o volume pots? I mean some may want to use these as power amp letting their current pre-amp take over the volume selection function.

More news on the Proprius mono block power amplifier...

Like I said in my last post, "it'll do will not do!"

I believe for an amp to sell it's got to make the speakers "disappear" and that means a rock solid sound stage. You hear a lot of talk about slew rate but maybe some are confused by it?

Slew rate put simply is how fast the amplifier will react to a step change in output. To be honest, a nice sound is obtained from amps in the 50 watt region for only around 2 Volts step change in a micro-second. The problem is that amp will be struggling to do a believable sound stage.

It's the fast changes the amp needs to do and with less than adequate slew rate those fast changes are going to be out of phase with the rest of the music - it's not about us being able to hear the high frequencies that make the start of a note fast and snappy - it's the stability it's done with.

Phase differences introduced by a slow amp trying to react to a fast step change cloud or veil the image - sometimes it's heard as distortion or if there's sounds like the tide breaking on the shore or rain drops (often found in popular genres) it sounds more like frying fish.

However, too high a slew rate puts the amp on the very edge of stability and there's insufficient phase and gain margin, which, guess what? Yes, it's going to lead right back to the fish fryer!

Somewhere in-between there's a happy medium. With the Proprius mono blocks the original prototype sounded really nice but remember it was a single channel and I was listening in mono - no sound stage! In stereo it was a different story. OK there was a sound stage but not the jaw dropping sensation that there was no discernible output from the speakers but sound from everywhere else, so further investigation was needed. As a result I found I could double the current required to drive the "miller" stability capacitor of the voltage amplifier stage (VAS) and not run into stability problems - that's a property of the unconventional type of voltage amp stage I'm using.

In fact, with a conventional VAS the amp often only has 60 degrees phase margin in its quest for maximum slew rate - a far better sound is obtained nearer to 90 degrees, but if you read up on amplifiers they'll always tell you 90 degrees sounds slow and sluggish. What they're really saying is at 90 degrees they have to compensate so much that there's insufficient bandwidth to have a good slew rate. It's all down to developing enough current to drive the "miller" stability capacitor.

What's also good about the Proprius mono blocks is the giant ultra fast output transistors that have a transition frequency (where they run out of gain [when they reach a current gain of 1]) of a staggering 30MHz!

This means we can achieve something in the order of 4MHz where the entire amp's output falls to a gain of 1 and the fall-off is a nice stable and gradual 90 something degrees. Even so the educated amongst us will be able to work out a slew rate of only 1.2V/uS. But here the VAS which is emitter degenerated (in other words a good thing) brings that up 15 times higher - in the region of 18V/uS.

Now that's only OK if the input doesn't try to "hunt" (or overtake) the output and that means an input filter (which the educated reader will have heard of) to slow down the input slew rate to below that of the output. In the case of the Proprius mono blocks the result is an input slew rate of 8V/uS measured on the scope - that's 4 times faster than the minimum required - and the output is faster still. And that's why after only a few days burn-in the prototypes are starting to make the speakers "disappear".