I get how power amps can make really significant changes to the sound of a system. And of course speakers have an even bigger effect. And then there is the complicated relationship between the speaker and power amp. But I wonder about pre-amps.
In theory a well designed preamp should just act as a source switch and volume control. But does it add (or ruin) magic? Can a pre-amp color the sound? Alter pace and timing? Could you take a great sounding system and spoil it with the wrong preamp? Stereophile once gushed (while reviewing a preamp that cost as much as a car) that the preamp was the heart of the system, setting the tone of everything. Really? Some people don't even bother with a preamp, feeding their DACs straight into the power amp. Others favor passive devices, things without power. If one can get a perfectly good $2K preamp, why bother with 20K?
It would seem to me that attenuation would be best implemented remotely at the input of the amplification stage and all other control duties be placed further upstream.
In theory a well designed preamp should just act as a source switch and volume control. But does it add (or ruin) magic?
it also matches (and thus affects) source and output impedance to connected components and cabling, with sometimes significant sonic benefit (or cost)...
The preamplifier is at the heart of any great hi-fi rig. A bit of history may be in order here. In the 60's - the 80's nearly all preamps were full-featured. They included a phono stage, a line stage and a power supply all in one box. There are real advantages to this approach. The installation is simpler, cable requirements are simpler - it has a concentrated purpose. All preamps of that era had single-ended circuit topology. That was the standard and had the very significant advantage of simplicity of the signal path. More recent marketing efforts promote balanced connectivity. Balanced topology imposes twice the circuitry into the signal path. Abandoning single-ended is not an advancement. Goals or attributes of a great preamp: 1) Straight line with gain without contributing to the sound thus preserving micro and macro details, etc. 2) Lo distortion 3) No noise (good signal to noise) 4) Two outputs 5) Adjustable gain and cartridge loading in the phono stage 6) Multiple stages of power supply filtering/regulation and a processor for control 7) Multiple line stage inputs 8) Line stage input with unity gain pass thru for Home Theater 9) Simple remote control 10) Able to drive long interconnects This is not easy to accomplish by anyone's measure. The market is full of preamps that fail pretty badly at either one or more of these goals. After all, there are no standards and specifications are unlikely to give you any scene or what the preamp will sound like. Auditioning several will be best solution for choosing. All of you in the "balanced only amplifier world" will have to seek out a "balanced" preamp.
That was the standard and had the very significant advantage of
simplicity of the signal path. More recent marketing efforts promote
balanced connectivity. Balanced topology imposes twice the circuitry
into the signal path. Abandoning single-ended is not an advancement.
@perazzi28 This statement is false. The best way to implement fully balanced circuitry without transformers is to go fully differential. This does not require 'twice the circuitry'. Add about 50% and you will be closer. And there are advantages: for example for a given stage of gain, you can have up to nearly 6dB less noise, and distortion will be reduced as well. In addition the circuit is far less sensitive to power supply noise and of course can reject noise at its input caused by hum fields and like impinged on the interconnect cable.
The cost of the parts does not seem to be a consideration; there are a good number of balanced line preamps available that are less expensive than single-ended preamp with which they easily compete.
Finally the balanced line system was created with two purposes; the first to eliminate ground loops, the second to eliminate the artifact of interconnect cables (if you've ever heard a difference between interconnect cables you know what I'm talking about). To make this happen there is a standard (which does not exist for single-ended connections so its a bit of the wild west 70 years on since hifi was created). It is known as AES48. FWIW, most 'high end' balanced audio products don't support the standard, which means that not all the benefits of going balanced will be realized with such equipment. If you've based your comments on hearing such gear, its no surprise you've come to your conclusion. But once you've heard balanced line executed correctly there's no going back. The improvement isn't subtle.
The best way to implement fully balanced circuitry without transformers is to go fully differential. This does not require 'twice the circuitry'. Add about 50% and you will be closer. And there are advantages: for example for a given stage of gain, you can have up to nearly 6dB less noise, and distortion will be reduced as well. In addition the circuit is far less sensitive to power supply noise and of course can reject noise at its input caused by hum fields and like impinged on the interconnect cable.
audio research figured this out and thus implemented differential amplification circuits long long ago...
A preamp can make a big difference - positive or negative - in my experience.
This weekend I auditioned some equipment for my system. I am interested in purchasing a more powerful amplifier.
My situation is I have a McIntosh C 2600 preamp and McIntosh MC 302 amp with Moon 280D DAC and Rega P8 turntable.
I chose to listen to McIntosh MC462 and Moon 330A (with the intention of buying 2 and using as a mono block.
The test system was McIntosh C2700 with the Moon 280D and Rega P8. I also had Moon 390P available to listen to the impact of the improved DAC and the difference of a solid state preamp. My rationale was to test any synergies of using the same brand.
My experience was unexpected:
1. I preferred the Moon amplifier. I felt it provided more details especially with delicate sounds or when there's a long of different music going on such as part of David Bowie's Let's Dance and during Black Sabbath War Pigs. I felt the highs were a bit 'brighter' with the Moon. Potential for harshness so I listened to a variety of recordings and my conclusion was the McIntosh provided a bit more mids such as around voices and there was less clarity when things were busy. The Moon was better and having congruent yet details in all the music.
Given that I did a sound comparison between using just the Moon amplification.
2. Tested the sonic difference between Moon 280 D and Moon 390 just as a DAC. The 390 has more detail - aligned with the product placement.
3. Used the 390 as a preamp instead of the McIntosh. The sound impact was more details and delicacy. I recently upgraded from Sonus Sonetto V to Olympica Nova V - the clarity in details was that significant. Note preamps are different philosophy tube versus solid state.
My take away- preamps make a difference as does the whole system (Millercarbon states it regularly and I observed it as well).
Can a quality integrated amp sound great? Yes and better is going to be a personal preference.
Thank you all for the super interesting responses, just what I was hoping for.
I liked the point that a perfect preamp would do nothing to the sound, but what would it cost? Maybe my 1K preamp is far from what it could be.
A general theme appeared to me that people found running without a preamp made their system sound a bit weak, undramatic. I had noticed this myself years ago, but thought that was just me. Now I have long interconnects (balanced) between pre and power so I can see that I need something to drive that cable.
I think my next treat will be a decent preamp. I am going to hunt around for one with a MC phono section but without an inboard DAC (as I think that technology is still changing fast, and I like the idea of keeping the digital stuff out of a mainly analog box). Something that feels classy to use (as the preamp is mainly what one touches).
Now I have long interconnects (balanced) between pre and power so I can see that I need something to drive that cable.
I don't know of a preamp that cost only $1000 that supports the balanced standard. Since you are driving long cables what you might consider is getting a set of Jensen transformers to sit at the output of the preamp and then have them drive the cables properly. You might be surprised at how well this can perform. You can do this with a single-ended preamp too.
I'm assuming that the amplifier has a balanced input. Its easier to build a balanced input that supports the standard than it is to build a preamp output section that does the same thing. So you may not need transformers at the input of the amp(s).
If you do try this, the output of any line transformer has to be loaded for best results. Jensen Transformers can advise you in this regard with their product, as should any manufacturer of a quality line transformer.
Passives seem to have the edge with faster transients and a somewhat leaner sound. We need a contribution here from georgehifi.
Back from holiday. Look at it this way if "going direct", "going passive", "going active" all drive the amp/s perfectly, with no impedance or voltage restirictions.
The one that sounds closest to a piece of wire is the best, that’s the "direct" next best is the "passive", then the "active". Going direct imitates a piece of wire with no colourations, next is passive, last is active as it has the most colourations/distortions.
The one that sounds closest to a piece of wire is the best, that’s the "direct" next best is the "passive", then the "active".
As we all know from the existence of the high end cable industry, sounding like a 'piece of wire' might not be the way to go- since obviously different pieces of wire sound different- hence the cable industry.
The solution to preventing wires from sounding different, and getting rid of their artifact, is a technology that was originally developed for the telephone industry, but was rapidly adopted by the recording and broadcast industries in the early 1950s- the balanced line system, which is implemented with a set of standards. A passive volume control cannot support the balanced line standard, and so is less capable of being neutral.
As far as “standards” go; it would seem to me that if amplifier manufacturers would adapt a standard for single ended input sensitivity of say between 2 V (CD) and 2.2 V (HDCD) and a balanced input of between 4 V and 4.4 V for full power output, it would go a long way towards making direct connection from digital sources, more accessible. This would permit full range of the volume controls and reduce the risk of bit stripping, especially with the now ubiquitous digital sources with extra bits that can be applied for this very function. Doing so would reduce the amount of cables and the corresponding concerns being attributed to them.
Addendum; Another consideration would be to include volume controls at the input of the amplifiers (this would probably require remote control) so that full output would be driving the full length of the cables. And yet another consideration would be, at least for Class D amplifiers, to have digital inputs. From which a conversion of PCM to PWM would take place, and thus eliminating the traditional DAC altogether.
Many amps are 1.5v or less input sensitivty for full output, and most dacs today are well over 2v output these days, so there's no problem with being able to voltage drive. And most dacs today have very low output impedance output stages, that are equal to or better than many solid state preamps to drive any capacitive interconnects with, essecially bettering most tube preamps. So it's a furphy to say a preamps drive better than dacs can direct. And like I said if a passive pre has no voltage or impedance issues then it is the next best way.
Sure balanced has it's place in over 3mt interconnect runs, for noise cancelation only, below that there is no need for it. As single ended is just as good, and better in many cases better, as the balanced inputs of many poweramps are just a "balanced opamp" then leading the into the real single ended input of the amp which will sound better.
@unsound You really do have to wonder what the designers of Redbook were thinking. In all cases of a Redbook output driving an amplifier, the voltage has to be knocked down, often quite a bit.
Many amps are 1.5v or less input sensitivty for full output, and most
dacs today are well over 2v output these days, so there's no problem
with being able to voltage drive.
-And this is false for the simple reason that there is a problem when you knock the signal level down. If a passive device is used, its an impedance that sits between the circuit driving the interconnect cable in the source, and what is being driven (the amplifier). That is why passive volume controls are so sensitive to cables: they are about as high an impedance as you can have for driving the cables and so are far more sensitive to the vagarities all cables have.
IMO/IME its a pretty poor line stage that can't outperform (and by that I mean be more musical and more neutral) a passive volume control. Put another way, if the passive seems better then line section you're using for comparison isn't all that great to begin with.
If a passive device is used, its an impedance that sits between the circuit driving the interconnect cable in the source, and what is being driven (the amplifier). That is why passive volume controls are so sensitive to cables:
This is false.
To explain to members with the maths of it. As a 10kohm passive has an outuput impedance of around 2.5kohm, the "cheapest interconnect cables" can be <100pf capacitace per ft and less. It would if be use you 2mts!! of it, still be only 300pf!!! And this 300pf capacitance with the passives 2.5kohm output impedance, will still only be only -3db down at 212khz!!!! in hf rolloff, a dog would have trouble hearing this.
And that’s why it’s false info to post and say passive pre’s are cable sensitive, sure to a dog or bat maybe!!.
To explain to members with the maths of it. As a 10kohm passive has an outuput impedance of around 2.5kohm, the "cheapest interconnect cables" can be <100pf capacitace per ft and less. It would if be use you 2mts!! of it, still be only 300pf!!! And this 300pf capacitance with the passives 2.5kohm output impedance, will still only be only -3db down at 212khz!!!! in hf rolloff, a dog would have trouble hearing this.
A 10K passive may well have an output impedance of nearly 10K, depending on the position of the volume control. This is in series with the source impedance, which often includes an output coupling capacitor. When you put a capacitor and a resistance in series, you create a simple filter. Filters can and do color the sound. This is easily heard. The most common complaint leveled at passive controls is a loss of bass and overall impact when the volume is anything less than full. This complaint has been mentioned by others earlier on this thread.
Its simple physics. It is possible for a thing to be too simple, and a volume control by itself is an example.
Another example of the sort of problem passives can cause is actually laid out in George’s example above. Let’s go with that 10K value. What that means is that the source (perhaps a CD player or DAC) is driving that 10K control. Many sources simply aren’t going to be happy with a load like that and will have higher distortion and often a loss of bass (since the 10K load interacts with the output coupling cap, creating a low frequency roll off). The formula for the cutoff frequency thus generated is: f = 1,000,000/R x C x 2Pi where f is frequency in Hertz R is resistance in Ohms (in this case 10K) C is capacitance in microFarads 2Pi is 6.28. (Normally this equation is seen with a ’1’ instead of 1,000,000 but algebra allows us to do this since normally C is would otherwise be Farads rather than microFarads. Farads are not a useful value in audio applications.) What this formula shows is that as R goes down, C must be increased. You want the cutoff frequency to be less than 2Hz so that phase shift is not introduced at 20Hz. This means that the minimum value for C is going to be about 6.8uf. Its a simple fact that due to inductance created when a capacitor is rolled that a cap that big will introduce coloration simply on account of the inductance. So many manufacturers choose to avoid values like that on that account.
George will (and has in the past) point out that a direct coupled output from the source will not have this frequency pole. But a load like that will affect its distortion- and limit the kinds of sources with which such a control will be successful! The obvious thing to do is increase the value of the control; by doing so you can run into Miller Effect problems with the input stage of the amplifier being driven. Miller Effect is caused by the input capacitance of the amplifier and is governed by the same formula, only now were talking about a high frequency effect. To avoid this problem, usually the control is buffered by some sort of active circuit. The simple fact is that passives will only work right in limited situations. This is why you see so much conversation about this, because different people have different systems. So we are often talking about apples to oranges when it comes to passive controls.
As an audiophile maniac for decades, I am a believer that the pre amp is the pulsing singing heart of the entire rig. It all comes through the pre amp. This is where I would (and have) place the most budget you can afford. My magic is solid state amp and tube pre amp combo. Neal
I would go with what Art Dudley owened! I'm so surprised to find so many high price systems can't do tone! For example roon source with YG speakers and Gryphon amps. Dynamic yes but tone no! Cheers!
The pre DO make the signal large. If the theory is talking the pre is just let the signal pass,it is wrong!
The signal from the DAC is not large enough for amplifier, unless the DAC has a pre inside.
The early stage matter! Both the DAC and pre have huge influence on the sound, as a little change in the early stage, there is huge change after the signal pass the amplifier.
I have a audio lab 8000s integrated to the speaker TANNOY 385,and this integrated has a power in RCA input , so I use audio research LS25 MK I as a pre to the audio lab 8000S as a power amplifier. Huge different! Even the gain is high than before.
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
