Resurrecting the Preamp

[roscoeiii]

As for the lightspeed, Ralph from Atmasphere commented on it positively on Audiogon, but also gave tips for it to sound its best (short interconnects of what length I forget). You might want to search hi comments over there (user name = atmasphere).

 

General rule of thumb for impedance is input impedance being 10x higher than the output impedance feeding it.

 

[4est]

Roscoe is correct in his 10x statement, but it does not tell the whole story. In my case, my DAC2 has 50 ohm out, and my amp a 47000 ohm in. That is a 1000x difference- should be good eh? Furthermore, my TVC has an output impedance of significantly more than 50 ohm, and yet sounds better. I assume teb_b's line stage has significantly more than the 50 ohm as well as it would be difficult to get that low with a tube circuit.

 

FWIW, the Concert Fidelity retails for about 18k as stated in a review.

 

[SteveChicago]

I've just plunged into a tube preamp and it's taken things to an amazing new height. :-)

 

[kana813]

I've had the PS Audio PWD and Weiss DAC202 in my system directly driving my amp.

 

Both sound excellent, but I wish they had a balance control.

 

[SteveChicago]

The Weiss DAC202 is extremely good, IMO. I auditioned one with and without a preamp. If you can, try auditioning a preamp of comparable quality to the DAC. You will be amazed. You might try a good tube pre...

 

[Panelhead]

The Lightspeed is best used with low output sources and high input impedance amps. The one's the George makes are about 7k or so. Many dacs(tube output) will not like a 7k load. A 50 ohm source like the Weiss 202 will drive it fine.

Most dacs have output impedances in the 100's of ohms. They will not like Lightspeed.

I used a Lightspeed with an amp with 100k input. All sources were 50 ohms or less. It was startling, better than any TVC or other passive. But the factory one are for one source. And not for all components.

My current preamp has opamp buffered inputs, PGA4311 volume control, and opamp outputs. Works with EVERYTHING. I remember someone in Eastern Europe building a preamp with opamp buffer on inputs, PIC controlled optio isolators, and opamp output. He said it was better than a Lightspeed in most applications. But was five times the cost at 4,000.00 USD.

The reason I parked the Lightspeed type volume control was going 2.1 and balanced. Matching up the optio-isolators is a PITA. There are 2 pairs for a stereo single ended. Two quads for balanced, and then another two pair for 2.1.

But in the right application the results can be amazing. It is a better way to adjust volume.

 

George

 

[rom661]

I expected to find everyone saying that a preamp isn't necessary. Instead there are a lot of thoughtful comments discussing the pros and cons. I have gone both ways (that came out wrong) but ultimately find I prefer a really good pre. Even more ultimately, yes I know that's nonsensical, I have no choice since I also use vinyl on a regular basis.

 

Admittedly I'm a bit spoiled on the preamp front. An Audio Research 40th Anniversary preamp with Sutherland Phono Blocks for phono stage.

 

So, preamp enthusiast that I am, I have to say that, assuming you have adequate output and no impedance matching issues, I would rather hear a no-pre solution than one with a poor quality preamp. I know preamps are supposed to be simple; switching, impedance matching, volume, but over the years I have found many companies who make very respectable power amps but whose preamps are under performers by comparison.

 

Again, nice thread. I always enjoy the discussions in the forum that don't seem to have an agenda and exhibit some thoughtfulness and an open mind.

 

Thanks

 

Rick

 

[One and a half]

The Grace Design m903 has a headphone amp built in, and I can't tell if I listen to that amp or my main amplifier and its headphone socket. One is the equal of the other, and there's a balanced cable in between plus may other components.

 

The output impedance for the headphones is 50 Ohms, for the line outputs the m903 expects a pre-amp as the impedance is 100k on the output.

 

You can also program the m903 to boot up to a preset analog output value, so if you drive this into a power amp directly at 0db, there will be some noise....so a good pre-amp is essential even just for that, or some other mishap.

 

[barrows]

Going direct without the preamp, and while there were some gains (slightly improved transparency) the losses outweighed the gains (lost body, weight, dynamics). I tried two DACs, one with ~150 ohms per phase output impedance, and the other with 100 ohms per phase. The amp is 11K in per phase, and my preamp is 55 ohms out per phase. I do not believe the problems of the no pre set up are due to using a digital volume control (my gains are matched such that 10 dB of digital cut is enough in most cases) but are caused by the DAC's output stge not driving the input stage of the amp as well as the preamp.

In my system, my preference for running the preamp is not caused by pleasant colorations, or other euphonic effects. And I believe impedance matching is not the soul source of the problem-more likely that current drive of the DACs output stage is a big part of what is going on. One DAC I have heard that loves driving amps directly is the PS Audio PerfectWave. It is probably not a coincidence that this DAC features an output stage designed to have large amounts of current drive capability.

 

[NickG]

I'm hoping I can get a few basics on impedance matching.

 

The previous post says that the Grace m903 won't serve as a preamp due to having a balanced input of 103k ohms. However, I am under the impression that is the output impedance that must be matched properly with the amp, with the amp having an impedance that is at least 10x that of the source.

 

The Grace m903 has an output impedance of 95 ohms (balanced) and 47.5 ohms (unbalanced). If I have this right, that would mean that the Grace could serve as a pre for most amps? Mine has an input impedance of 28k ohms. So, the Grace should be able to easily pass the maximum current to the load. If I have this wrong, please let me know, as I was thinking I would try the Grace as a pre and a DAC.

 

Meanwhile, does anyone have experience with the Grace or the Calyx?

 

[sandyk]

"I have tried...

Going direct without the preamp, and while there were some gains (slightly improved transparency) the losses outweighed the gains (lost body, weight, dynamics)."

I have found the same.

Another major problem is that the vast majority of power amplifiers appear to have a sensitivity of 600mV to a little over 1V RMS, for full power output.

In a normal domestic situation, much of the time we wouldn't be

using much above 1W of power into the speakers.

To use >10dB of digital attenuation with a typical DAC results in too high a loss in resolution. 10dB or less seems q good compromise, especially with 24 bit files.In the case of some of the Oppo SACD/DVD-A players, there is an initial setting of 10dB max. of attenuation via SPDIF using the remote control.

SandyK

 

 

 

[Exeric]

As far as the impedance goes, it is not really a question of matching. All else being equal you always want the lowest impedance output from the dac as you can get and the highest impedance input on the power amp as you can get. The analog output from the dac should be considered a voltage source, not a power source, so no impedance matching is involved like there is on the output from the power amp going to the speakers.

 

The reason the above is important is because the capacitance in cables or lowish impedance input on a power amplifier etc can load down the output from the dac at high frequencies if the power amplifier's input impedance is not high enough. The dac ends up being a power source then instead of a voltage source and has to supply current into those loads before the amplifier "sees" the input signal. Not good. If the dac doesn't have a low enough output impedance to supply that instantaneous current you then end up bypassing part of the high frequencies to ground through that capacitance.

 

This isn't always bad. If one's dac or power amplifier is harsh on the highs this may improve things subjectively. Or not. If not then sometimes a preamplifier can improve things. But all else being equal you should not need a preamplifier as long as your dac and amplifier are up to snuff. If you need one to improve the sound it tends to indicate you have a problem somewhere. By the way, the DCS dacs have an amazingly low output impedance of less than 1 ohm. This means you generally always get the intrinsic sound of DCS dacs as far as frequency response into the amplifier regardless of the amp.

 

As a rule of thumb transistor outputs in dacs and tube inputs in power amps have the most advantageous ratio of impedances. Of course there are exceptions.

 

[Jim Smith]

Me too.

 

Including the very expensive dCS Paganini, which is claimed to be able to drive a power amp directly.

 

All the audiophile terms were fulfilled (transparency, soundstage, clean, etc.), except for two things. It just didn't seem to have as much drive, and more importantly, it just wasn't as involving musically.

 

Same for my Ayre QB-9, when using PM as the volume control, running at 24 bit resolution.

 

[rom661]

This may not be pertinent to today's world but I used to use a Meridian 518, what John Atkinson referred to as a digital Swiss Army knife, between my Levinson transport and DAC. The 518 did reclocking, noise shaping, dithering and digital volume control, if enabled. I went straight from the output of my DAC to the inputs of my monoblocks.

 

When I first attempted this, the Levinson No.30 was a 16 bit DAC. It lost sound quality immediately when attenuating, way before anything like 10db. When I later had it upgraded to the 30.6 it became a 24 bit DAC. I reclocked the word length in the 518 to 24 bit and tried using the direct connection again. At 24 bits by the time I really noticed loss of resolution, it was getting down close to background levels.

 

I know some of the products, Amarra included, claim to use dither to prevent loss of resolution. Others are simply attenuating the analog output. Although I use Amarra, I have not bothered with a comparison with my preamp.

 

Rick

 

[SteveChicago]

... isn't there something to the idea that a good quality device that is designed for one function often outperforms another that's intended to multitask? E.g., few if any receivers approach the performance of separates.

 

[NickG]

Yes, that is the conventional wisdom. And it used to be the conventional wisdom that you couldn't get high quality playback from a computer. :-)

 

[Exeric]

I'm kinda surprised at your experience, Jim. I'm wondering what the input impedance, or more importantly, the amplification factor of the power amps were that you were driving directly from the Paganini. Did you try the 6 volt output setting from the Paganini instead of the 2volt normal output?

 

Sometimes one doesn't have speakers or power amps that are well matched. Synergy is definitely part of audio. If you don't have it sometimes you have to add the complication of an added component, such as a preamplifier just to bring the entire system into balance. But it is always best to select the combined system with the least number of components in the chain that has that synergy. There is just less between your ears and the original recording in that case.

 

If amplifier manufacturers wanted to they could build brilliant integrated amplifiers by just adding one or more voltage amplification stages into the front of their power amps. It is really stupid not to do that. But they have figured out how to separate the two components and make more money that way. Maybe they get a kickback from the interconnect manufacturers to do it that way.

 

[Exeric]

Removed duplicate message.

 

[One and a half]

yes, I misread the output impedance, it was at maximum output level at some + 27dbu for 100k.

For normal levels, the output impedance for a balanced line is 95R which is fine cause balanced systems work off 110R, nominal.

So if you crank up the gain, the amount of load impedance goes high until it hits the wall at +27dbu, which is over the top anyway, so your load would have to be high (200K+) to match. This is extreme, prolly never happen, but at least the output stages won't be overloaded and cause the distortion that goes with it.

Headphone output impedance is 1R, close to a short circuit, so that's more then enough for headphones.

Unbalanced just asks for noise introduction, I wouldn't dream of using that system, especially on the input to a power amp.

 

The integrated I have can accept a direct input into the power amp, I can experiment on the weekend, see how it goes.

 

[SteveChicago]

... but that's in large part because pieces were made / improved for computer playback even with older gear, not because they built a better "all in one."

 

[barrows]

question, that DACs could be designed and built to drive power amps directly, with no compromise. All this requires is that the DAC have an output stage equal in capability to the output stage of a (well matched to the amp) preamp. If one has issues with digital volume control, a DAC could also have a properly implemented discrete resistor volume control (the Aesthetix Pandora DAC offers a analog, discrete resistor volume control as an option).

Having a really good output stage would not add a whole lot to the selling price of a DAC, but a good, balanced, discrete resistor volume control would add a lot to the selling price of a DAC, as parts quality and proper implementation of a volume control like this is not simple. Of course, really good preamps are costly as well! I would love to see someone build a really good buffered, 0 dB gain, volume control, with just a single balanced input and output. A well designed component like this could outperform any preamp (input switching adds distortion/noise, as does gain), and still give top quality analog domain volume control, and the output buffer would provide the necessary drive to properly couple with the amplifiers input.

 

[Jim Smith]

I'm kinda surprised at your experience, Jim. I'm wondering what the input impedance, or more importantly, the amplification factor of the power amps were that you were driving directly from the Paganini. Did you try the 6 volt output setting from the Paganini instead of the 2 volt normal output?

 

I was using amps with an input impedance of 20k (Pass Labs XA30.5), and up to 100K - various tube amps.

 

Most have a gain structure of around 20 dB or so - in other words on the low side - which is good, as I typically use high-efficiency speakers.

 

The last thing I would want to do is use the 6 Volt output and then throw away tons more resolution.

 

I wasn't comparing the DACs to a tube preamp - in these instances - instead it was a Pass Labs X1 - hardly a colored sounding preamp.

 

Just like some of us didn't know that computer audio could outperform redbook CDPs, I'm not sure we know why a technically excellent DAC doesn't have the drive capability of a well-designed preamp. For all sorts or reasons, I WANTED the DAC output to be superior, or at least as good.

 

It did have that ultra-clean, see-though factor. But when it comes down to sheer musical involvement - which is so often related to dynamics - it was simply unacceptable. And lots of knowledgeable listeners agreed.

 

[Panelhead]

I have used a couple and found the older Pass amps require almost an amplifier to be driven properly. The output stages used in Pass preamps uses the same mosfets as many amplifiers. They potentially can output several amprees with the right load.

I do not know why the dac had trouble with the tube amps. Those are usually easy to drive.

I have a line stage that uses 6AS7 output devices. 80 ma standing current. I thought it could drive anything from an amp to headphones. It lacked a lot with three different Pass amps, all SE design.

Back on track, I found a preamp added nothing compared to the simple pots on my active speakers and sub. It only took away. But nothing tried cost over 2K or so. But that is a limit in my mind for a linestage. A line stage should not cost over two grand.

 

George

 

[barrows]

while I agree with you regarding the older Pass amps, the X .5 series and XA .5 series amps are not difficult to drive properly. The .5 series was changed specifically to address this issue, and the input stage transistors were changed to easily driven JFETs. People are now having plenty of success driving the X .5 series and XA .5 series Pass amps from any preamp, including high z out tube designs.

Jim's XA 30.5 is a current Pass model, balanced, class A, with the more recent input stage design.

 

[Jim Smith]

... as well with a straight feed from the DAC, when compared to using a preamp, no matter how expensive the DAC, or how strong the claims that are made to the contrary:

 

ARC

Art Audio

Audiopax

Avantgarde Acoustic

Ayre Acoustics

BAT

Bottlehead

Cary

Conrad-Johnson

E&E The Lars

FirstWatt

Goldmund

Lamm

Levinson

Quicksilver

Various RAM models

Pass Labs Aleph 3, XA 30.5, X250.5

Tom Evans Linear A

Vivas

Wavelength Triton, Duetta, others

And others that I can't remember just now.

 

Never heard a DAC work at its most engaging level when directly driving any amplifier, period.

 

Haven't heard the PS Audio piece, though.

 

Would love for a DAC mftr. to design an excellent analog output stage that would allow bypassing a preamp.

 

For example my Ayre QB-9 with a real linestage & volume control from an Ayre preamp would be awesome (well, at least to me).