|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
4.246.253.92
In Reply to: RE: More, more posted by Charles Hansen on February 05, 2009 at 10:39:33
Hey,
>>The analog output stages are (naturally!) zero-feedback and fully balanced. The prototype at CES had circuitry similar to the CX-7e MP CD player, but next week we are going to listen to a new discrete circuit I came up with. We'll see which one sounds better.<<
Regarding the analog outputs on a couple of your products, if I have it correct, the C-5 has a discrete circuit, the CX-7 is op-amp based and you're still deciding for the USB DAC.
>>The cost of the two approaches is (surprisingly enough) not all that different.>>
Any chance the CX-7 in the future might get discrete analog outputs?
Thanks,
Scott
Follow Ups:
From a 2003 interview (linked below):
We are now using ICs (integrated circuits) in some of our less
expensive products designed by Dr. Barrie Gilbert, one of the
"godfathers" of analog circuit design. Now, it’s very important to
differentiate between the terms "integrated circuit" and "op
amp" (operational amplifier). We’re using integrated circuits,
which simply means that there are many transistors created at
once on a single piece of silicon. An op amp, on the other
hand, uses negative feedback as its fundamental operating
principle. You cannot use an op amp in a zero-feedback circuit.
The topology of the IC we use in the less expensive products is
very similar to the topology of our discrete circuits. We've fig-
ured out how to implement these monolithic ICs with zero feed-
back. By using two of them together, and modifying their char-
acteristics, we can get within spitting distance of the perform-
ance of our discrete circuits at a much lower cost. The critical
advantage here is that the monolithic design of the IC means
that all of the transistors are matched extremely closely. In con-
trast, with our discrete circuits we have to spend a lot of time
measuring, sorting, and matching transistors, which translates
to a more expensive final product. By using an IC in a way that
hasn’t been done before, we’re able to achieve a real break-
through in performance at a real-world price point. It's great to
make an all-out assault on the state of the art, when the only
limit is your imagination. But the real challenge is to bring those
lessons back to the real world where more people can enjoy
the fruits of our labors.
We are very happy with the analog section of the CX-7e MP. We are not planning to make any changes to it.
Charles,
Thanks for the clarification and the link. Now I get it; I should have known "zero-feedback" and "op-amp" were mutually exclusive.
Just getting familiar with some of your products and rather enthused about the MP filter.
Scott
> > rather enthused about the MP filter < <
We are too, now that all that hard work is done!
We had originally planned to listen to the filterless "non-oversampling" designs many years ago. Unfortunately circumstances prevented it at the time. It turned out to be a good thing in the end, though.
The original impetus for listening to the digital filters was simply to check out the "non-oversampling" option and compare it to our existing filter. But then we decided to also listen to Peter Craven's "apodizing" filter. After that, the project took on a life of its own as we kept thinking up new ideas to try out -- it was months before we finished.
Even when we *thought* we were finished, we realized at the last minute that there was one configuration option in the DAC chip we hadn't listened to. This realization occurred the afternoon before we were packing up the entire soundroom to go to CES. So three of us dropped what we were doing and met down at the factory to do some listening tests. We ordered a pizza and ate there while we were listening. When we finally finished at 1:30 AM, we were pleased with the work that we had accomplished. But one of the crew (Ariel Brown, Ayre's "secret weapon") only got about 3 hours of sleep before he had to get up and pack for the road trip to Las Vegas...
But if we had listened to the "non-oversampling" years ago, we could have only compared it against the off-the-shelf digital filters. At that time we didn't have the tools to make our own custom filters using the DSP capabilities of the FPGAs. Plus we knew about, but didn't have the software tools to design the "minimum-phase" digital filters that turned out to sound better than the standard "linear-phase" filters. All of the filter design software packages just assume that everyone *of course* wants "linear phase"! This feature was only recently added to our software package, a $2,000 tool.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: