How important is the pre-amp?

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.

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.

@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.

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.

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.

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.

So in the end, would a high quality integrated be the best way to go (assuming one likes the sound of the unit) due to one less pair of cables, electrical cord, etc?

Note the word 'best' in the question above. Integrated amps are a set of compromises. They might be excellent, but if we are talking about best then they will not be able to outperform separates that are built with the intention to be the best. This simple fact has been a thing in audio for the last 70 years and hasn't changed at all in that time. 

So in the end, would a high quality integrated be the best way to go (assuming one likes the sound of the unit) due to one less pair of cables, electrical cord, etc?

If you're looking for ultimate sound quality, no. Integrated amps have both power amp sections sharing a common power supply; in rare exceptions they might have dual power supplies (so that's a thing to look for) but you have a lot of other circuitry on the same chassis and finally, to make it work it has to sit between the speakers (since for best results the speaker cables should be kept short). That may not be the best place in the room since vibration can play a role in system performance. Integrated amps often share ground connections, which has the same effect that you hear with a 3-wire headphone hookup as opposed to a 4-wire hookup. Its nice to keep the left and right hand ground circuits separate- that gives you the opportunity for lower noise and possibly less ground loop potential.


Separates have more chassis real estate for things like extra power transformers, regulation and the like. The don't share power cords (unless you plug them all into the same power strip) so there are less voltage drops in the AC supply.

I firmly believe that for 99% of applications, a true, discrete, stepped attenuator using a minimal number of resistors, great contacts, and good (not awesome) resistors is pretty darn transparent.

It is.

The thing that is keeping the preamp as a thing in high end audio is the simple fact that most sources don't have the ability to handle the interconnect cable used all that well. A good preamp does. A preamp that colors the sound more than a passive control isn't a good preamp in my book. A good preamp will be less colored than a passive. This is simply on account of the interconnect cables themselves (if you've ever heard differences between cables you know what I'm talking about). Cables will exhibit the greatest amount of coloration with passive devices.

I disagree that the preamp’s sonic performance is always tightly correlated with the price.

Funny- I was just commenting about this on another forum. I agree the price doesn't have much to do with performance. Those that price to a formula will be less expensive generally speaking. My philosophy runs very similar to yours.

Transformers on the other hand, where resistors have problems transformers have benefits. There is a reason after all why we use them all over the place for isolation and power conditioning.  

You solve one problem; you introduce others. Then it gets tricky.


The issue with transformers is loading and inter-winding capacitance. All transformers have inter-winding capacitance. If the transformer is not loaded correctly (usually too lightly) the capacitance may play a role causing the frequency response to be less than flat. This coloration can be easily heard.

Transformers are called that because they transform impedance. It goes both ways; you're not insulated on the input from the output. So if the output load is reduced, so is the impedance on the input side.

What happens when you have a transformer with multiple taps used as a volume control? Since you are working with a variable turns ratio, to keep the transformer linear you need to have the correct load at both the input and output. You can't, as a designer, expect to always see the same source impedance and you certainly don't see the same load impedance since all power amps are different in that regard. On top of that you have the moving target of variable turns ratio.


So you'll need a lot of switch contacts, not just the ones to change taps but another set with resistor to correct for the changing load impedance as the turns ratio is varied. 

This has to be done right, else the unit can introduce coloration.

Do you believe it’s a point of tremendous debate that certain preamps color system sound? If so, that point of view would be surprising. One has only to look at twenty years of Audiogon preamp discussions to reach the conclusion that among many it’s not debatable, and in fact is a consideration when selecting a preamp.

@tvad

Nope- not debating that.

Or, do you rather believe it’s a point of tremendous debate that some folks like coloration that certain preamps provide?

Not debating that either.

I was talking about the endless debate between passive and active linestages.

I’m in the active camp because I can get the interconnect cables to drop out of the system equation and I can keep the distortion so low in the active linestage that coloration isn’t a thing. Not all active preamps can do that so coloration happens. Also, no passive or transformer system can do that either so you can expect coloration from them too.


To my knowledge you can either throw money at the interconnect cables in the hopes of minimizing coloration, or you can get something that was designed to minimize the coloration of cables. That latter problem was solved in the 1950s with the balanced line system (which is based on standard) but is mostly unused in high end audio. The latter part of the last statement is true because most high end audio companies either ignore the balanced standard or are ignorant of it; most high end balanced preamps I’ve seen don’t support it. The Backert is one of the few I’ve seen that does.

 

I am in agreement that Balanced is not always better than single ended. It's like wine , it depends on your taste.

Its not a matter of taste, its a matter of most high end products don't support the standard. If they did there would be no debate, no matter of 'taste' in this regard. IOW if you've heard a balanced setup but the standard isn't supported, you may hear colorations that cause you to prefer a single-ended setup you heard.

Some folks like coloration certain preamps provide.

This is of course a point of tremendous debate. I prefer the lack of coloration afforded when the interconnect cables cease to alter the sound of the system. You can't do that with a passive setup. But to be clear a lot of active preamps can't insure that either.

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.

This is how important: no matter how good your amps and speakers are, they cannot make up for a shortcoming upstream. The preamp is vital in this regard. Sometimes you can get by with a simple volume control and a switch (to do so, plan on keeping your interconnect cables really short).

I have my preamp sitting with my source equipment in an equipment stand that is well away from my speakers. To increase system resolution, I run short speaker cables and long (30') balanced interconnect cables. But my preamp supports the balanced line standard so I can run cables that long with no degradation and the kind of cable is irrelevant to the sound of the system (since that is one of the advantages of the balanced line system).  In a case like this you can see how important the preamp is, since most sources do not have the ability to drive long cables.