[tortugaranger]

We've been working on several projects in parallel for the past year including a high performance power amp. For a long time I was big fan of various chip amps which can offer very good performance. But I was looking for something beyond what a chip amp can offer with perhaps a smoother more natural sound. After several prototypes I believe we are close to finalizing an amp design.

After listening to the current prototype for several weeks now I'm really pleased. Smooth and articulate with no harsh solid state artifacts. Works really nice with our Tortuga preamps. This amp is a fully discrete design although it does use a JFET op amp in a servo feedback loop to minimize DC offset.

Aside from the servo the fundamental amp design is fairly old school insofar as it uses 100% bipolar junction transistors in a relatively simple yet highly effective topology. There are some really great high performance and robust power transistors available today that can really deliver great sound in a solid state amp.

Here are a few CAD renderings along with some specs. The enclosure top/front/sides would be the same baltic birch edge grain plus veneer used in our "boathull" speaker designs that are coming along although these CAD renderings don't reflect that look/material. Trying to think outside the metal box...seen enough of those.

No pricing yet but I think this will fall into middle range of high end amps which in my mind means more than a couple of $k but under $10k. 

* Dual channel stereo
* Handles both singled ended (RCA) and balanced (XLR) inputs
* Oversized 400 VA toroidal transformer with independent soft start circuit
* Nominal Power 60W into 8 ohms
* 47k input impedance
* 2.3V/V input sensitivity (gain)
* Class AB power stage with class A voltage stage
* All bipolar junction transistors - all discrete
* Dual parallel power transistors per leg to share load and minimize thermal related issues
* Both AC (coupling caps) and DC direct inputs
* Active servo feedback to minimize DC offset in output
* Output disconnect protection relays for high offset or power supply drop out
* 12V trigger input for remote on/off from external preamp

[TrungT]

Can't wait ... 

[tortugaranger]

A brief update.

After extensive testing/listening to our prototype fully discrete power amp I've concluded that its performance is inferior to our earlier designs using the LM4780 chip amp. As a result we are scrapping the fully discrete design and are proceeding with a chip amp as was our initial plan.

The LM4780 is basically a pair of LM3886's amps on a single somewhat larger/wider chip. I've previously run the LM4780's in parallel which effectively doubles the amperage/power output with great results using the stock design.

Thermally, the LM4780's are challenging relative to the LM3886's because you're cramming twice the current into almost the same die area. The LM4780's are also being discontinued. For both of those reasons we are proceeding with a dual/parallel mono channel board using a pair of LM3886's side by side in parallel mode on a single board.

This will not be a stock design. We've pulled out all the stops to get the best possible performance out of this chip amp. First, we are driving the LM3886s in inverted mode to minimize common mode distortion. To deal with the low inverted impedance we've added a JFET (op amp)  input buffer plus inverter in front of the dual LM3886. Secondly, we'll be using a DC servo for output offset control and thereby eliminate the conventional DC feedback circuit along with its big noisy electrolytic capacitor. The user will have the option of running the input DC direct or AC via coupling cap.

Other upgrades include all Vishay/Dale resistors, polypropylene caps, multilayer X7R ceramics caps, and ELNA silk electrolytics. Basically all high quality components throughout.

Lastly, after much research and back and forth hand wringing, we are going with oversized 200W switch mode power supplies - one for each channel. Thus the amp will actually be a pair of independent monoblock amps in a single enclosure sharing only a common power entry.  For those who hold to the traditional view the all amp power supplies must be big, heavy, inefficient, and linear, with a boatload of copper and rows of caps, I get the nostaglia, but SMPS's for high performance audio have evolved to the point where it makes little sense not to use them. I realize audio enthusiasts will never all agree on this but then again when do we ever.  thumb:

Cheers,
Morten

[Randy]

Morten,
   Have you looked at the Benchmark amp which came out last year to good reviews? It is a small box maybe similar to what you are going for. Any thoughts on how your amp would compare to the Benchmark?

Randy

[tortugaranger]

Morten,
   Have you looked at the Benchmark amp which came out last year to good reviews? It is a small box maybe similar to what you are going for. Any thoughts on how your amp would compare to the Benchmark?

Randy

Based on my last post on the topic I was quite certain that we were going with switch mode power supplies similar to the Benchmark. Then we tried a 3rd party 200W switch mode PS and it was a huge disappointment. I've considered developing our own design but not this year - too many other priorities. And while I'm sure there are better alternative switch mode PS's out there, I made the decision to stay with our muscular 400VA toroidal linear power supply that occupies a lot more real estate not to mention a lot more weight but is a rock solid proven performer. So no light weight smaller box like the Benchmark. The latest prototype boards arrive next week so we'll be testing out the latest PS and amp board soon. Will post an update once we've had some time to test out the new boards.

[AGrede]

Any updates?

[tortugaranger]

Any updates?

Yes and thanks for asking. For most of 2016 we've had no less than 4 projects under parallel development:  1) Version 3 (the "V3") of our preamp controller; 2) A buffer; 3) An amp, and; 4) A speaker. In nudging all 4 (including variations thereon) forward in parallel progress has been overall sloooow....too slow. More recently I put the amp and speaker development on hold and have been focusing on the V3 and the buffer. Once both the V3 and the buffer have been released we'll resume work on the amp and maybe also the speaker. Everything takes twice as long and costs twice as much. Nothing new.