[modwright]

OK, balanced inputs and outputs are not a problem, but do require either four tubes or four transformers, which adds cost.

Aesthetics will be good, but depending on how sexy, will dictate price!

I am envisioning a steel box with a nice AL face plate. If you really wanted it to look nice, we could also machine the entire thing out of solid Billet, but that adds at LEAST one zero to the price tag! I would prefer to have nicely powder-coated steel with an AL face. Probably silk screened graphics.

I am envisioning something along the lines of the PH 9.0 phono stage, in terms of aesthetics:

http://modwright.com/sources/ph-90-tube-phono-stage.php

Considering that I was envisioning the simplest version, with RCA's in and out, buffer stage and no volume control, gain or transformer coupling, to start at $750, you get an idea of the cost.

I appreciate the input and continued discussion here as it all helps me to shape this into a project scope that will suit everyone's needs as best as possible.

Dan

[Cave]

Function over form for me, I'll take the cheapest box option 

[modwright]

Well, the spirit of this product is that it be more function than form. Of course it won't be ugly, I DON'T DO ugly . But, the money will be spent on design over aesthetics. This is mean to be a product for the mod end of our business, sold factory direct.

Thanks,

Dan

[The Rang]

I’m sure it will be awesome regardless of casework and features

[modwright]

Thank you, this is shaping up to be a great design and I am confident that it will fill a much needed niche for modern systems. It will provide exceptional sound in today’s more simplified and sonically pure modern analog and digital systems.

[Cave]

When you feed this to a passive attenuator, say with a 10K impedance, the output impedance of the attenuator is the same as the input impedance = 10K. So  you are then interfacing 10K with a 50K amp input impedance which is not ideal.

The buffer would be ideal to use between the passive volume control and the amp. It would take the 10K impedance of the pot, seen by the 500K-1M input impedance of the buffer and offer an output impedance of < 100 ohms to the amp's 50K input impedance. This impedance matching, current buffering and handling of the signal by the tubes will give you more body, weight and authority to the sound without any loss of frequency, detail or bass.

Dan

Hi, Goldpoint attenuators:

http://www.goldpt.com/info.html

"Ideal" may need to be more specifically defined.

"Matching the volume control value to the amplifier/pre-amplifier input impedance is not as important as some people think, so you typically don't really have to worry about this. For a 100K input impedance, use a 100K, 50K, 25K, or even a 10K level control.  This applies to passive preamps, active preamplifiers, and power amplifiers when the control is to be installed at the Input."

There's also a lot more on that page that reads very interesting.

[modwright]

I am talking specifically about adding a passive attenuator between source and amp, where there is no active circuitry following the attenuator. In that case, the output impedance seen by the amp IS the impedance of the passive attenuator. True, a 10K passive into a 100K input impedance amp is not bad, but with the rule of thumb ideally being 1:100, not 1:10, you get the idea.

I speak from both experience and engineering knowledge. Before I fully understood why, I used a number of very high quality passive attenuators and the Goldpoint is a great one! It is a great volume control, period! I am just saying that if you use even the BEST passive attenuator between source and amp, there will not be a good impedance match.

My personal experience many years ago was that, while the best passive, always sounded very clean and quiet, it lacked body and weight. I then added an active preamp and found the body and weight that was missing. After this, I always used an active preamp. It has to be a GOOD active preamp however!

I have since learned that it is really a matter of impedance! The use of a good buffer AFTER the Goldpoint or similar excellent volume control/attenuator WILL give you the best of both worlds! The active circuitry must be good, quiet, well designed and not harm the signal of course. I have long since determined that adding circuitry to the signal path is NOT always a bad thing! The same can be said of signal path transformers.

Many of use believe that less is more, always! I don't agree with that 100%. I am not a fan of tone controls and I am not crazy about DSP, because it all involves manipulation of the audio signal, in ways that often creates more problems than it solves. I don't like my analog digitized and I find that when any signal is digitized, manipulated and decoded again, there is too much lost. DSP room correction for low frequency issues IS a good thing, but I would NOT like to have the entire signal treated with DSP.

Now, we are not talking about digital or DSP, but it illustrates a point of adding circuitry for a net positive gain...or not!

Many people feel that adding signal transformers will always limit fidelity. I believe this comes from the fact that tube amp output transformers, especially SET transformers, must be very well designed, in order to not limit bandwidth and add distortion due to core saturation. I will argue in this case, that well designed SET transformers can and will produce the MOST BEAUTIFUL music playback when done well!

Push Pull tube amp output transformers are MUCH less impacted by the issues that SET transformers are because there is very little net DC voltage seen by the core. As long as global feedback is not used, P-P tube circuits can be AMAZING sounding and with very good BW! Our P-P 300B Integrated is flat to 50Khz!

Non-gapped, audio transformers (do not see DC voltage) are an entirely different matter. I use Lundahl line-level non-gapped transformers extensively in our best products! Jeff Rowland has done the same for years, also with great results.  Yes, you are adding a copper coil and magnetic core to the signal path. However, you are also getting galvanic decoupling, inductive reactance and out of bandwidth filtering (say > 60Khz rolloff) which is ideal for digital circuits. Transformers can also attenuate signal level and allow for impedance matching. They are amazingly useful and musical reactive devices! I would much rather create a fully differential balanced output from a single ended circuit with transformers than use op-amps (feedback based) or other active circuitry to achieve the same!

So, to sum all of this up...

The addition of a tube buffer falls into the same camp of addition to the signal path WITH net benefits, not sonic degradation. For the design with gain, we will use transformer coupling and likely the same for balanced input and outputs. The tube circuit offers very low distortion, very low noise and bandwidth flat from 20Hz-100Khz.

Those using Goldpoint or other high quality passive volume controls will find that the addition of our Analog Bridge after the attenuator and between the amp will offer VERY welcome sonic improvements!

Thanks,

Dan

[Cave]

What is a good or ideal impedance match? Specifically please.

[Rusty Jefferson]

What is a good or ideal impedance match? Specifically please.

Preamp output impedance 0 ohms, amplifier input impedance ♾.

[modwright]

1:100 is the ideal ratio. So for 100 ohm out, 10K min input impedance for amp, min. The higher the ratio, the better , but 1:100 is min ideal.

[Cave]

So still the Goldpoint link states that it's not as important as people think. So there's another camp I guess.

Also I can't help but think that a) it can't possibly be good to add more processing in the path and b) the sound will change especially if tubes are involved, which would sort of imply impedance mismatch may have had nothing to do with the new different sound, hope that made sense.

[Cave]

Preamp output impedance 0 ohms, amplifier input impedance ♾.

No trying to be a smart ass, but how did you arrive at those values?

[Rusty Jefferson]

No trying to be a smart ass, but how did you arrive at those values?

It's a theoretical ideal. It would make the current/voltage conversion also ideal.

[modwright]

There are always different views and ideas about things. Impedance relationship is a hard fact and the math will prove this out and it can be verified by test equipment. I won’t dispute that tubes produce their own sonic signature and not everyone likes tubes. I do and our customers have also been very pleased with our products’ sound.

Use of a SS buffer would achieve the same impedance matching and would have a different tone than that of  a tube buffer.

There is no question that a great passive attenuator will produce a very clean sound. I am simply expressing that what I heard when I tried this many years ago was good, but I felt that I knew what was missing, to
my ears, when a good active stage was used between attenuator and amp.

I am not posing this as an argument. Impedance relationship is a fact. We do all hear things differently and as a result, we will have different opinions about what we deem as best sound. Those who know me and our products and how they sound, will like what they hear with this product.

Different opinions are always respected here and by me personally.

[GregC]

I ran an Exogal Comet DAC direct to amps (as an experiment) to see if music would sound as good as without a preamp.  The Exogal Comet has a ridiculously low 18 Ohms for the XLR outputs.  It sounded very clean and had no problems driving my amps.  My amps have a 20k input impedance.

The Ambrosia preamp has a 29k RCA input impedance and 100k XLR input impedance.  The output is 60 Ohms (higher than the Comet Direct).  When the Comet was ran through the preamp the music gained weight, body, and air. 

I guess what I am saying is there is more to the story than just impedance.  Other design elements come into play.  I currently own Dan's PH 9.0 SE phono stage (killer by the way), and I have owned other products from Dan in the past.  Dan has a deft ear for getting good tone with lots of detail, without adding distortion.

[modwright]

True, current buffering is also very much a part of the process. Lastly, adding to the signal path is OFTEN not
detrimental, if the design is done Wright!

[GregC]

True, current buffering is also very much a part of the process. Lastly, adding to the signal path is OFTEN not
detrimental, if the design is done Wright!

Nice, I see what you did there.

[Cave]

Well, if the impedance deal is real, then why not just build or use something like a 2.5K ladder network attenuator, by itself?

[modwright]

Well, a 2.5K ladder attenuator gives you a 2.5K input impedance seen by the source and a 2.5K output impedance seen by the amp. Neither is ideal.

[Sailorboy]

Dan
Just a quick note to say that I am new to this forum and I am looking for this type of product right now, so please count me as another potential buyer in your market calculations
FYI...I am running one of your 100SEs off a T+A DAC8, using the preamp in the DAC, its pretty good, but cant wait to see your product on the market.
Good luck with the design.

Regards
Colin