Ayre wants $1.5K for DSD'ed QB-9

[barrows]

I have found the volume control in the ESS 9018 DAC chip to work very, very well in my set up. But, I have my system pretty much optimized around using it: Because I have a DIY DAC, I am able to adapt the gain of the DAC's output stage to match my system, hence I play audiophile recordings such Reference Recordings (which are recorded at lower average levels) usually at only -2 dB to -5 dB, and the more typical high level recordings at no more than -15 dB to -20 dB. I listen at relatively high levels almost all the time, so for me and my system, DAC direct and digital volume control works perfectly. But, for those with a DAC which might have, say 4V-6V output at 0 dB, one would have to use a lot more attenuation with most amp/speaker combos, then there could be problems. I guess those folks would be best served by getting an Ayre KX-R, or the sweet new integrated amp. I will also say that not every DAC output stage seems to have enough current drive to mesh well with every amplifier input stage and cable combination, I have heard some combinations which seem to lack both body and dynamics.

With all digital systems we are at a turning point from traditional Hi Fi, which has always required a pre-amp to be the control center of a system. Without the need for phono preamps (for many listeners) or even multiple source components (for many listeners, a computer itself can select multiple sources) going DAC direct to amp (or integrated amp with VC) is an excellent choice for most. I, personally, if building a system from scratch, would rather apply money saved from not having to purchase a preamp, on better speakers, where the differences in performance are large. Remember also, even with the Ayre QB-9 one can use use the volume controls in Pure Music, Audirvana, HQplayer, etc, to good effect, especially if the system is well matched from a gain perspective.

[Defender]

Actually this discussion is soo much informative - as much it made me register for computeraudiophile (only to find out I registered years before)

 

It is impressive to have Charles share his ideas and thoughts so open - I already saw him writing at the Hi-Rez Highway discussing with Alex Peychev which I respect as much as him. But Charles brings High End more to the "normal" user.

 

I had a small affaire with computer audio about 3 years ago - bought NAS, ethernet switch and LINN Majik as it was said and "tested" to beat even the best CD/SACD Players out there. Since than I dont buy audio "blind" anymore just because of good tests in newspapers or because people say it sounds good - you have to hear it in your own home/room, in your system with your music.

 

It could not even beat a 10 year old Marantz CD16. Since than I have some resistance to computer audio. But I still like the idea of NAS-Streamer-Stereo System - not so much Computer-USBDAC - Stereo System.

I might be just old school but computer and audio shouldnt be in the same room and computer should not directly connected to audio. If you do over usb you have to invest in better power supplies and you might also have to buy a dedicated computer for audio - that adds up to the initial costs.

I agree with Charles on the advantages and disadvantages of both systems - but for me the LINN system was easy to set up - it just sounded not as perfect as I hoped.

 

Charles I was looking at the QB9 at that time too but the LINN experience made me not further evaluate it because it was USB and no DSD and I really like the sound of the small bunch of SACD´s I have.

 

Since you seem to think about something streaming - I am really interesting to see/hear that soon because I like your analog audio concepts and your filter approach.

 

One thing makes me wonder with all that PCM vs. DSD discussion. Ed Meitner is a strong believer in 1bit and converts everything to DSD - his MA-1 is a perfect example for a nice sounding product. There is something he must do right.

 

Charles if you get close to the performance of the MA-1 with the new QB9 I will be stupid again and buy it blind - because you have the better output stages ))

[dca58]

Hello Silver,

The Classe uses a Cirrus Logic IC-based volume control with an internal high-feedback op-amp. The best thing that one can say about it is that it offers a lot of positions (128) for a low price.

 

No, it's a PGA2310 from Burr Brown.[h=1][/h]

[SoNic67]

No, it's a PGA2310 from Burr Brown.

It's pin-compatible with CS3310, with (slightly) better specs. A good volume attenuator.

[barrows]

Actually, if you want a good resistor based VC on a chip, best choose the MUSES part and implement it by bypassing the internal opamp and using a well designed discrete circuit. Still, it is doubtful that an IC like this will outperform a no compromise discrete resistor VC using .01% TX-2575 resistors and shallco switches.

[Audio_ELF]

With respect to those in this thread... It really annoys me when people who should know better (especially other manufacturers) start pulling a product apart because of a particular component used here or there.

 

Yes "everything matters" but actually all that matters is how the complete product sounds.

 

Barrows (and Charles before) you have discredited a product which (afaik) you don't own and maybe never demo'd based soley on its use of a volume IC!!!! I'm sorry but that just isn't right.

 

Eloise

[One and a half]

One thing makes me wonder with all that PCM vs. DSD discussion. Ed Meitner is a strong believer in 1bit and converts everything to DSD - his MA-1 is a perfect example for a nice sounding product. There is something he must do right.

 

 

Playback Designs use the same approach, converting any PCM fs on the input to DSD.

[barrows]

With respect to those in this thread... It really annoys me when people who should know better (especially other manufacturers) start pulling a product apart because of a particular component used here or there.

 

Yes "everything matters" but actually all that matters is how the complete product sounds.

 

Barrows (and Charles before) you have discredited a product which (afaik) you don't own and maybe never demo'd based soley on its use of a volume IC!!!! I'm sorry but that just isn't right.

 

Eloise

 

Ummm, what! I did not mention any product, I only pointed out that if one is looking for a really good VC IC, the Muses part is likely the best available. I made no mention of any other product, or any manufacturer.

[Audio_ELF]

Ummm, what! I did not mention any product, I only pointed out that if one is looking for a really good VC IC, the Muses part is likely the best available. I made no mention of any other product, or any manufacturer.

I'm sorry Barrows. But that doesn't fly. Either you are niave or trying to blind us. Your comment followed a thread on the IC used in a Classe product following a discussion on Ayre's preamps...

 

I don't know what industry contacts you still have; but you regularly promote specific companies method of design while subtly criticising others... Never quite crossing the line perhaps but still there is a bit of a pattern!

 

I would copy and paste to show the sequence of comments but I'm on my phone and everyone can scroll back 3/4 posts I'm sure...

 

Eloise

[barrows]

I'm sorry Barrows. But that doesn't fly. Either you are niave or trying to blind us. Your comment followed a thread on the IC used in a Classe product following a discussion on Ayre's preamps...

 

Eloise

 

I assure you neither is the case. I have no bones with Classe. I am not Naive, although I try to be as much as possible ;-)

 

Since I pretty much summed up my feelings on this topic, I am out. I hear Hope Sandoval calling me....

[barrows]

I can understand if you are no longer monitoring this thread, but if you are I have a few tech questions on the new QB-9:

 

1. I understand that you are running the ESS chip synchronously, that is with master clock and bit clock synchronous, disabling any intervention from the onboard ASRC and DPLL, right?

 

 

2. I also understand that you are running your own OSF (on FPGA), and I would guess that this is the same set of filters as in the previous QB-9, outputting 705.6 kHz and 768 kHz to the ESS chip, with your minimum phase/slow roll off "listen" option, right?

 

 

3. Given if my understanding above is correct, I suspect that you are using fixed oscillators at 45.184 and 49.152 as master to the FPGA and ESS chip, right?

 

 

4. And, that the OSF in the ESS chip is off, eliminating the onboard OSF, right?

 

 

Thanks again for sharing so much information here, I liked the previous QB-9, and I bet the new version sounds fantastic!

[Charles Hansen]

No, it's a PGA2310 from Burr Brown.[h=1][/h]

 

I haven't seen the inside of it, but Alan Clark who designed it said it was the Cirrus. But maybe he made a mistake, I don't really know. I think he was more in charge of the project than actually selecting which components to use.

 

Best,

[Charles Hansen]

With respect to those in this thread... It really annoys me when people who should know better (especially other manufacturers) start pulling a product apart because of a particular component used here or there.

 

Yes "everything matters" but actually all that matters is how the complete product sounds.

 

Barrows (and Charles before) you have discredited a product which (afaik) you don't own and maybe never demo'd based soley on its use of a volume IC!!!! I'm sorry but that just isn't right.

 

Eloise

 

Hello Eloise,

 

I understand what you are saying, but I strongly disagree. The way that we select components such as volume controls or switches is to compare them in a bypass test. This is done in a very rigorous fashion. We take a male connector and female connector and wire them back-to-back. The ground connection is done directly. The signal connections are done with the highest quality wire we can find.

 

We make two sets of these. The "Control" set has the two wires soldered directly together. The DUT set has the switch or relay or whatever DUT we are evaluating soldered between the wires. So now EVERYTHING is COMPLETELY identical. The same number of connectors, the same type of connectors, the same wires, the same solder, the same number of solder joints, the same EVERYTHING.

 

We are comparing the sound of the DUT to the sound of NOTHING.

 

This is a very difficult test.

 

By far the best sounding switches are the Shallco mil-spec switches with solid silver contacts and solid silver wipers. These sound almost exactly like the direct connection. There is perhaps 1% loss of resolution and that is all.

 

In contrast, I have never heard ANY relay that sounded anything at ALL like the direct connection. They ALL mangle the sound. I can assure you that any product that uses relays in the signal pat is absolutely not a neutral component. That doesn't mean that you won't like it. It just means that it changes the sound irreparably.

 

We test resistors in the same way. We make a 10:1 voltage divider with a 10 kohm series resistor and a 1 kohm shunt resistor which provides 20 dB of attenuation. This gives a nice listening level with most discs when a CD player is connected directly to the power amp.

 

It will absolutely DESTROY the sound of 99% of the preamps available on the market. We make it so that it plugs directly into the rear of the power amplifier with no cable to drive. In this case it is impossible to perform a bypass test, so all that one can do is to compare one brand and mode of resistor against another.

 

When one sets up a very carefully controlled experiment like this, it becomes trivially easy to "rank" the resistors in terms or their sonic performance.

 

I have no problems at all when I say that such and such a component sounds good or bad. I know because I have listened to it in the most rigorous circumstances possible. That is how we select the parts we use in our products. I have never listened to $100 capacitors and $10 resistors, as I think that is rather silly. I have listened to $20 teflon capacitors. I have listened to the effects of cryogenically treating them. I have listened to $3 foil resistors. But one has to draw the line somewhere. I have zero interest in making a $200,000 power amplifier, so I don't listen to parts that are silly expensive.

 

Best regards,

[Charles Hansen]

It's pin-compatible with CS3310, with (slightly) better specs. A good volume attenuator.

 

Burr-Brown makes one part that is a copy of the CS3310. These both have a huge limitation in that the maximum signal is only 3 volts RMS. Burr-Brown makes another one which at least will handle a normal sized signal with some degree of headroom. But I have never heard an op-that I liked, either inside or outside of a volume control.

 

Everything is a compromise. If I were trying to make a $500 product, I would probably have to use op-amps. But when I see a $10,000 product with op-amps, I think it is a joke. In essence the designer is saying, "I have no idea how to design a circuit. I will leave that up the the engineers at the op-amp company that never, ever listen to anything they make." To me that is pathetic.

 

Best regards,

[Charles Hansen]

I can understand if you are no longer monitoring this thread, but if you are I have a few tech questions on the new QB-9:

 

1. I understand that you are running the ESS chip synchronously, that is with master clock and bit clock synchronous, disabling any intervention from the onboard ASRC and DPLL, right?

 

 

2. I also understand that you are running your own OSF (on FPGA), and I would guess that this is the same set of filters as in the previous QB-9, outputting 705.6 kHz and 768 kHz to the ESS chip, with your minimum phase/slow roll off "listen" option, right?

 

 

3. Given if my understanding above is correct, I suspect that you are using fixed oscillators at 45.184 and 49.152 as master to the FPGA and ESS chip, right?

 

 

4. And, that the OSF in the ESS chip is off, eliminating the onboard OSF, right?

 

 

Thanks again for sharing so much information here, I liked the previous QB-9, and I bet the new version sounds fantastic!

 

Hello Barrows,

 

Yes, yes, yes, yes, and I think so. YMMV. It's always best to listen for yourself especially with recording with which you are extremely familiar.

[barrows]

Thanks Charlie. I really prefer the sound of the ESS with synchronous clocking with a good source. I am suspect that many people will be very pleased with new QB-9...

[nyc_paramedic]

Hello Eloise,

 

I understand what you are saying, but I strongly disagree. The way that we select components such as volume controls or switches is to compare them in a bypass test. This is done in a very rigorous fashion. We take a male connector and female connector and wire them back-to-back. The ground connection is done directly. The signal connections are done with the highest quality wire we can find.

 

We make two sets of these. The "Control" set has the two wires soldered directly together. The DUT set has the switch or relay or whatever DUT we are evaluating soldered between the wires. So now EVERYTHING is COMPLETELY identical. The same number of connectors, the same type of connectors, the same wires, the same solder, the same number of solder joints, the same EVERYTHING.

 

We are comparing the sound of the DUT to the sound of NOTHING.

 

This is a very difficult test.

 

By far the best sounding switches are the Shallco mil-spec switches with solid silver contacts and solid silver wipers. These sound almost exactly like the direct connection. There is perhaps 1% loss of resolution and that is all.

 

In contrast, I have never heard ANY relay that sounded anything at ALL like the direct connection. They ALL mangle the sound. I can assure you that any product that uses relays in the signal pat is absolutely not a neutral component. That doesn't mean that you won't like it. It just means that it changes the sound irreparably.

 

We test resistors in the same way. We make a 10:1 voltage divider with a 10 kohm series resistor and a 1 kohm shunt resistor which provides 20 dB of attenuation. This gives a nice listening level with most discs when a CD player is connected directly to the power amp.

 

It will absolutely DESTROY the sound of 99% of the preamps available on the market. We make it so that it plugs directly into the rear of the power amplifier with no cable to drive. In this case it is impossible to perform a bypass test, so all that one can do is to compare one brand and mode of resistor against another.

 

When one sets up a very carefully controlled experiment like this, it becomes trivially easy to "rank" the resistors in terms or their sonic performance.

 

I have no problems at all when I say that such and such a component sounds good or bad. I know because I have listened to it in the most rigorous circumstances possible. That is how we select the parts we use in our products. I have never listened to $100 capacitors and $10 resistors, as I think that is rather silly. I have listened to $20 teflon capacitors. I have listened to the effects of cryogenically treating them. I have listened to $3 foil resistors. But one has to draw the line somewhere. I have zero interest in making a $200,000 power amplifier, so I don't listen to parts that are silly expensive.

 

Best regards,

 

Charles,

 

Have you ever heard of the Placette Audio RVC? It's a passive that uses the Vishay S-102 resistor for every single volume setting. With these resistors, the input impedance is a constant 9k at every volume setting.

 

I use the RVC with your QB-9 and it has been the most transparent device I have ever listened to. No loss of dynamics, rock solid imaging and no forwardness as you increase the volume.

 

Since your QB-9 is a single input device and the RVC only accepts one input, the pair is a minimalist's dream.

[Audio_ELF]

Actually Chatles, the jist of what I said was that it was bad form for one manufacturer to discredit another manufacturer...

Hello Eloise,

 

I understand what you are saying, but I strongly disagree. The way that we select components such as volume controls or switches is to compare them in a bypass test. This is done in a very rigorous fashion. We take a male connector and female connector and wire them back-to-back. The ground connection is done directly. The signal connections are done with the highest quality wire we can find.

 

We make two sets of these. The "Control" set has the two wires soldered directly together. The DUT set has the switch or relay or whatever DUT we are evaluating soldered between the wires. So now EVERYTHING is COMPLETELY identical. The same number of connectors, the same type of connectors, the same wires, the same solder, the same number of solder joints, the same EVERYTHING.

 

We are comparing the sound of the DUT to the sound of NOTHING.

 

This is a very difficult test.

 

By far the best sounding switches are the Shallco mil-spec switches with solid silver contacts and solid silver wipers. These sound almost exactly like the direct connection. There is perhaps 1% loss of resolution and that is all.

 

In contrast, I have never heard ANY relay that sounded anything at ALL like the direct connection. They ALL mangle the sound. I can assure you that any product that uses relays in the signal pat is absolutely not a neutral component. That doesn't mean that you won't like it. It just means that it changes the sound irreparably.

 

We test resistors in the same way. We make a 10:1 voltage divider with a 10 kohm series resistor and a 1 kohm shunt resistor which provides 20 dB of attenuation. This gives a nice listening level with most discs when a CD player is connected directly to the power amp.

 

It will absolutely DESTROY the sound of 99% of the preamps available on the market. We make it so that it plugs directly into the rear of the power amplifier with no cable to drive. In this case it is impossible to perform a bypass test, so all that one can do is to compare one brand and mode of resistor against another.

 

When one sets up a very carefully controlled experiment like this, it becomes trivially easy to "rank" the resistors in terms or their sonic performance.

 

I have no problems at all when I say that such and such a component sounds good or bad. I know because I have listened to it in the most rigorous circumstances possible. That is how we select the parts we use in our products. I have never listened to $100 capacitors and $10 resistors, as I think that is rather silly. I have listened to $20 teflon capacitors. I have listened to the effects of cryogenically treating them. I have listened to $3 foil resistors. But one has to draw the line somewhere. I have zero interest in making a $200,000 power amplifier, so I don't listen to parts that are silly expensive.

 

Best regards,

[Charles Hansen]

Charles,

 

Have you ever heard of the Placette Audio RVC? It's a passive that uses the Vishay S-102 resistor for every single volume setting. With these resistors, the input impedance is a constant 9k at every volume setting.

 

I use the RVC with your QB-9 and it has been the most transparent device I have ever listened to. No loss of dynamics, rock solid imaging and no forwardness as you increase the volume.

 

Since your QB-9 is a single input device and the RVC only accepts one input, the pair is a minimalist's dream.

 

Dear Paramedic,

 

First of all, thank you for the job you do every day. Being a paramedic in NYC can NOT be an easy job -- but every day you save people's lives. When I was struck by the motorcycle, I suffered double pneumothorax. If I had not been taken to the hospital within two to three hours by the paramedics, I would have died.

 

I have heard a lot about the Placette, but never listened to one myself. It is nice when two components work together well. We use complementary JFET inputs on all of our power amps. With this setup, we can choose the input impedance at will, anywhere between 10 kohm (about as low as you can go and still maintain compatibility with virtually all products) up to 1 Megohm (about as high as you can go and not have any problems from gate leakage current). I have found that the input impedance affects the sound significantly. I believe that this is due to the loading presented to the interconnect cables.

 

I prefer the sound with the higher input impedance, so over the years we have increased the input impedance of our amplifiers from 10 kohm to 100 kohm and now 1 Megohm (per phase). But every little detail matters. A part that is firmly mounted to a circuit board sounds quite a bit different from the same part mounted only by its leads, as in a point-to-point wired product. And the mechanical properties of the PCB have as much effect on the sound as the electrical properties do.

 

There are so many variables that it can take a long time to find the right combination of components that will deliver sound that makes you feel good when you listen to your favorite music. I am glad that you have found a combination that you like. Please try the DSD updated version when you get a chance. I think you will like it even more!

 

Best regards,

[Charles Hansen]

Actually Charles, the gist of what I said was that it was bad form for one manufacturer to discredit another manufacturer...

 

Hello Eloise,

 

Yes you are right. I am not a very diplomatic person and sometimes it causes problems for me. Nobody is perfect, and I am probably further away than most.

 

But some thing are really not disputable. A car may seem wonderful to someone even though it has solid rubber tires. But if the manufacturer changed to pneumatic tires, I would be willing to bet that both the people that liked the old version and the people that didn't like the old version would like the new version more than the old version. YMMV.

 

Best regards,

[labjr]

Charles,

 

Do you still use Toshiba JFETs? I think I recall you were disappointed the 2SK170/ 2SJ74 parts were discontinued and no comparable replacement was available? Linear Systems was supposed to start making them but more than five years later, both complimentary parts are still unavailable .

 

Some manufacturers bought a lifetime supply but some didn't. And what will new manufacturers use for JFETs? Do audio manufacturers have any influence in keeping parts available? I don't know what else those JFETs were used for but audio?

 

Thanks.

[labjr]

Edit duplicate message.

[barrows]

Charles,

 

Do you still use Toshiba JFETs? I think I recall you were disappointed the 2SK170/ 2SJ74 parts were discontinued and no comparable replacement was available? Linear Systems was supposed to start making them but more than five years later, both complimentary parts are still unavailable .

 

Some manufacturers bought a lifetime supply but some didn't. And what will new manufacturers use for JFETs? Do audio manufacturers have any influence in keeping parts available? I don't know what else those JFETs were used for but audio?

 

Thanks.

 

I do believe that Ayre owns a large stock of the discontinued Toshiba JFETS to which you refer. The high end audio market is generally speaking not big enough for audio companies like Ayre to exert significant infuence on the large semi conductor companies. Especially when one considers how many so called "high end" companies seem content to use IC opamps in place of discrete circuit designs. This is a shame, but it is reality. At some point designers are going to have to make do with the parts currently available, but I believe Ayre has enough of these NOS parts left to last our lifetime...

Of course, there are still other JFETS available, but there are those designers who feel that the discontinued Toshiba parts are the best...

[Jud]

Of course, there are still other JFETS available, but there are those designers who feel that the discontinued Toshiba parts are the best...

 

Kinda thinking I recall reading that Nelson Pass is among those who share a liking for the old Toshiba parts.

[Charles Hansen]

Do you still use Toshiba JFETs? I think I recall you were disappointed the 2SK170/ 2SJ74 parts were discontinued and no comparable replacement was available? Linear Systems was supposed to start making them but more than five years later, both complimentary parts are still unavailable .

 

Some manufacturers bought a lifetime supply but some didn't. And what will new manufacturers use for JFETs? Do audio manufacturers have any influence in keeping parts available? I don't know what else those JFETs were used for but audio?

 

Hello Lab,

 

Yes, we still use the Toshiba JFETs. Back in the '70s and '80s, "consumer electronics" meant two things:

 

a) Televisions -- CRT only. Sony owned the market with their "Trinitron" picture tube. Consumers Reports tested it and said it had the best picture of any TV. Overnight the stigma of buying Japanese TV's disappeared. Zenith, GE, Magnavox, Sylvania, RCA, and the rest of the US manufacturers went out of business almost overnight. If you wanted the best, you bought a Sony. If you wanted to save money, you bought a Panasonic or Toshiba or Mitsubishi or whatever.

 

b) Stereos -- Everybody wanted to get a new "solid-state" receiver that had 100 watts per channel, didn't need tube replacements and had unmeasurable distortion. The only US brands that sold were Marantz and Dynaco. Everbyody else bought Sansui, Pioneer, Sony, Yamaha, Technics (Panasonic), et cetera.

 

It was such a huge market that ALL of the big Japanese semiconductor companies were having contests to see who could build the best transistors. You could get complementary JFETs from NEC, Hitachi, Sony, Toshiba, Yamaha, et cetera.

 

Then came VCRs. Next was computers, which was 10x bigger than stereos, although Japan pretty much missed out on that. It went from IBM and Compaq to Gateway and Micron, to Taiwan. Only Sony sold any Japanese computers. Then the Walkman hit the market and the CD. There wasn't any room in a Walkman for discrete transistors, and CD players were designed by digital engineers that just put op-amps everywhere an analog circuit was required.

 

After that was DVD, CRT projectors (front and rear). There was one last hurrah by Sanyo, who made some killer discrete transistors for the horizontal deflection coils in computer monitors. But everything else was headed to op-amps.

 

One-by-one they all dropped out until the only man standing was Toshiba. They made a great product, but they had the market totally to themselves for 15 years. But that wasn't enough. The JFETs weren't officially imported into the US as they would only bring a part in if they could sell $300,000 per year per part. If you wanted JFETs, you had to buy "gray market" parts that were bought by a couple of specialty companies that bought from Japanese distributors.

 

Since we weren't buying through "official channels", we were not notified that the low-noise P-channel parts were being discontinued. We had two months to figure out what to do. Linear System was working on a P-channel part but weren't having a lot of success. (They've made an N-channel part for over a decade now.) We ended up buying a half-million 2SJ74's which should last us between 50 and 100 years.. About a month later the head of Toshiba discrete products for the US flew out to Boulder to see why a dinky company they had never heard of bought so many parts.

 

I asked him if they would sell the process to another manufacturer. He said, "No, just the legal part would cost more than it was worth." I asked him if we got a consortium of audio companies together if he would make them for another year. He said, "No, a Canadian company (non-audio) had wanted to purchase 1,000,000 parts per month and the factory said "No bid. We are shutting the product line down."

 

We knew it was only a matter of time before they stopped making everything. It turned out that a few years later they stopped making ANY part in a TO-92 (through hole) package. Nowadays "consumer electronics" means iPhones, iPods, and iPads. They can only fit the tiniest parts in there. They still make one or two SMT JFET's but N-channel only. No complementary pairs. A few years ago Fairchild still had a pretty good selection. We tried every one that they made but the all basically sucked next to the Toshiba parts.

 

So we bought another half-million each of the low-noise N-channel parts, the high voltage complementary pairs, and the through hole bipolar transistors. They still make the bipolar transistors in SMT packages, but they can only handle half the power of the through-hole parts. Basically what we did was the equivalent of Audio Research buying a 100 year supply of US and European tubes in 1970 before all of the US and European tube companies shut down. Now all you can get is half-assed Chinese tubes and OK Russian tubes. But before the Berlin wall fell, there was about a decade when American companies could only buy tubes from Yugoslavia.

 

The only reason that ANY tubes are made at all anywhere in the world is because of guitar players. The market for tubed guitar amps is 10,000 times bigger than the market for high-end stereo tube amps. But none of today's tubes are as good as the NOS tubes from the '30s, '40s, and '50s. Why do you think that Audio Research products are only about 1/4 tubes and 3/4 solid state?

 

Thanks,

Charlie Hansen

Ayre Acoustics, Inc.

 

I do believe that Ayre owns a large stock of the discontinued Toshiba JFETS to which you refer. The high end audio market is generally speaking not big enough for audio companies like Ayre to exert significant infuence on the large semi conductor companies. Especially when one considers how many so called "high end" companies seem content to use IC opamps in place of discrete circuit designs. This is a shame, but it is reality. At some point designers are going to have to make do with the parts currently available, but I believe Ayre has enough of these NOS parts left to last our lifetime...

Of course, there are still other JFETS available, but there are those designers who feel that the discontinued Toshiba parts are the best...

 

Yes, there is nothing that is even close to the Toshibas since NEC dicontinued the 2SJ44 and 2SK163 almost 20 years ago. And even Fairchild and NXP (formerly Philips) are discontinuing most of their FET products. Linear Systems has a hard time building the P-channel parts. Nobody can supply good wafers any more. Siliconix used to make their own wafers. Then about 20 years ago, they started outsourcing them and the quality went to hell very quickly. If you don't use JFETS, you pretty much have to put coupling caps in the signal path and all of the music is colored by the caps. But like you said, most companies don't care. They just put out a bunch of features that nobody needs. make one small change every year or two and say that it has been "completely redesigned". Okie-dokie....

 

Best,