What is the Sound of Impedence Mismatch?

As I understand it, you want your power amplifier to have an input impendence much higher than the output impedence of your preamp, at least 10x. Can anyone tell me what the sonic symptoms are of a mismatch? If I'm hovering around 10x, what might I hear that would indicate an impedence mismatch as opposed to, say, a preamp that is simply too bright or whatever?
Lack of dynamics, recessed sound, and different interconnects making big differences to the tonality - are what you would expect if the ratio is too low. But you can also have problems when the ratio is too high. For example I has a ss preamp using BUF03s as an output device, and these things have vanishingly low output impedence, driving a tube amp with high input impedence and the sound was overly hyped and hard-edged. Padding down the input impedence of the tube amp improved the neutrality and overall naturalness, and fleshed the images out.
Dull, lifeless musical rendering, massed strings undifferentiated, often somewhat harsh, grainy sound, no bloom, lack of dynamics . Redkiwi is right to these here ears.
Thank you, gentlemen.

I am playing with a First Sound Presence Deluxe tubed line stage, which has an output impedence of 1700 ohms. It's very happy with my CJ 11a at 100K ohms input impedence. But with the Pass X-150 (spec sheet says 22K balanced, not sure what it is unbalanced, which is the way I am running it), the sound is, if anything, a bit bright and over-etched. No problems with bass or dynamics, definitely not lifeless, just tilted up a little. From your descriptions, it doesn't sound like this is a technical mismatch, however. Could be a listener mismatch :)

Sounds like the Presence has too much "presence". Which tubes is it using? Sometimes a simple tube swap is all that's required to restore a more pleasant tonal balance...
Of course if you are just borrowing the piece and can't change the tubes, that's another matter.
No, I own the piece. To tell the truth, I don't know what tubes are in it. I'll have to open it up and check.
I've read that bass rolloff (happens due to interconnect capacitance) can be caused by such an impedance mismatch where the pre's output Z is too high for it's PA load, so you might have such an issue Drubin.
Have you experimented with different upgrade AC cords though?
No bass problems, actually, so that suggetsts no impedence mismatch. I need to experiment with cables and will also try power cords.

Wouldn't it be great to have a concise set of rules of thumb for system problems? Guidelines for identifying an impedence mistmatch, or cables with capacitance that is too high, too low, etc. Am I dreaming? It's just that it feels like most of the time I'm drifting around rudderless when some basic principles would help the navigation a bit.

While I'm at it, let me offer up my idea for the audiophile test CD I'd like to see. The Chesky Ultimate Demo disk is pretty good, but I'd like to see someone produce a disk that demonstrates the unwanted side of things. This is what grain sounds like. This is what midrange suckout sounds like. This is bad VTA ... azimuth mis-alignment. This is wooly bass. Clipping, even vs. odd-order harmonic distortion. You get the idea.... Even though the playback system would alter what you are hearing, I think you'd be able to get the picture if it is presented relative to a reference. Would be tremedously helpful to me in terms of calibrating my vocabulary with others.

Dan you are a dreamer indeed, but no fault with that! Your suggestions are really terrific. I think for now at least the closest that we'll get to "guidelines" are to be found here on these forums (try this - try that - see what happens - scratch your head some more). The funny thing is that a particular setup for one system may not be at all appropriate for a different one. (One person says that this XYZ cable is great, yet another person is totally dissapointed).
The "test disc" idea seems quite worthwhile; when are you going to get started? I'll buy the first one off the press..
I wouldn't know how to create such a disk, but I would put up some money to get it produced. Anyone interested?

As I thought about this, it occured to me that we might be able to capture noteworthy sonic signatures as well. Going out on a limb here, cause I know this is problematic and full of pitfalls ... if there were a way to convey the "classic" sound of CJ, of SET amps, selected eras from Krell, ARC, etc., the Koetsu sound vs. ??? Not a substitute for evaluating products, but it could have value in helping people who haven't heard it all (like me) to organize their preferences.
The impedance mismatch is only critical when the mismatch causes the preamp to go into distortion. If the preamp's current drain into a lower impedence device puts the preamp into a clipping mode similar to a amp driving low impedence speakers at high volume. So the sound is normal distortion IM and THD. Harsh sound would be the first clue. Some preamps can take a ratio of 1 and others need an amp impedence of 20 time their output impedence because they can't "drive" any current.
Keis are you familiar with the concept of LCR cable-reactance-induced frequency response rolloff? I cannot at all agree with your statement above.
In the Recording and Broadcast Industry, virtually all Audio Wiring is balanced, it is normal practice for Studio Equipment to have an Output Amplifier (Line Amplifier) with a low output impedance, say about 50 Ohms. This amplifier is generally not capable of feeding into a 50 Ohm load but is designed to feed equipment having an input impedance 600 Ohms or greater, very often in the order of 10 K Ohms. The same will apply to Stereo Hi Fi equipment where you can always feed a low impedance source into equipment having a high input impedance. This applies for relatively short run audio wiring or cabling inside a Studio Complex. For long runs, say 50 meters or more, or into Phone Cables going outside the studio, proper Line Driver Amplifiers are used capable of delivering audio into a load of 600 Ohms, these Line Driver Amplifiers are always balanced and capable of driving into 600 Ohms. Early practice was to build the output impedance up to 600 Ohms by adding resistors in series with the output, and make the input impedance of the equipment being fed 600 Ohms. Modern practice is to use the low impedance output of the line amplifier directly to feed into the line. Most equipment inherently have a high input impedance, made low by terminating the input by a resistor, usually 600 Ohms. In most cases, the accuracy of the 600 Ohm Termination resistor is not particularly critical. For Stereo Hi Fi applications, it should in virtually all cases be OK to feed a low impedance into a high impedance. Hi Fi equipment line outputs generally have a low output impedance, but cannot cope with low impedance loads and will distort if required to feed into a low impedance load.
Poulkirk your above treatise is very nicely explained & prompts me to clarify. Loading & clipping is of course a valid consideration. But it is not *the only* consideration of impedance mismatch, which is more generally manifested in freq. response problems than in clipping distortion.
Yes Bob I am familiar with LCR effects in cables. What's your point? I was answering the question about preamp amp interaction without considering cables. Poulkirk brings up the professional 600ohm standard which while correct is likely not answering the question which is probably possed from a RCA connection perspective. I was addressing the loading issue, not a potential frequency rolloff due to inductance effects. You might want to explain the issue of frequency rolloff to Drubin and how the R in the LCR circuit is partially the output impedence of the preamp since you are so keen on it.
Thanks for all the input everybody. It's the damndest thing, but the brightness has gone away. I loaned the First Sound to a friend for a few days, and ever since its return it has not sounded bright at all with the Pass amp.

This hobby can be so maddening sometimes!

And now, the First Sound sounds too dull and soft with the Pass. These days I am using a 4M run of interconnect, which could be a factor. Or I could be losing my mind.

Nine months later, I still wrestle with these same issues. The First Sound sometimes sounds too dull (or soft) and sometimes it's plenty alive. Honestly, I think it is variability in my hearing that does this. How can it be that one week I'm ready to sell all my tube gear and the next I want to dump the solid state? That's why I still have both.

May I suggest you look up the thread: How to match ohms with amps & preamps
Thank you for the pointer to your earlier post, unsound.

I have a simple multimeter. Is it possible to measure the input and output impedences of my components with the meter? (I ask this because there is no publshed spec for my Pass amp in single-ended mode.)
No need to measure: balanced connections will double the input impedance of your component. This is the drawback to using balanced connections (i.e. differential). You get double the noise cancellation but double the input impedance. The stereo works as a system and what is delivered to your speakers is the system transfer function (input/output). The impact of the input/output impedances on this transfer function are in general small in audio applications. Ideally, you will have a product of the preamp TF and the amp TF (e.g. infinite Zin and 0 Zout), no more and no less. However in the typical case, there is some mismatch and so you get less than ideal - this is the nature of transistor output stages. FET amplifier inputs alleviate this greatly with their high Zin but the problem will forever exist. Capacitance matching is more an issue than resistance actually but this problem is often resolved with good interconnects. Modern design is component-oriented which means the amp is designed by itself and the preamp is also. This is fine as long as you can make them work together independantly. The way you do this is to make Zin infinite and Zout zero. If you are more confused now, email me and I will explain in greater detail.
hey again
Sorry I got my lines crossed on my last post! I meant the problem with the balanced connection is that it doubles the OUTPUT impedance (as well as the input but that is beneficial as I described later on) of the component. I should reread my post before submitting but alas..! Now you must be really lost! ;) Arthur
So if the Pass amp is 22K balanced, it is 11K single ended? And with 1700 ohms output impedence on the First Sound preamp, my situation is sub-optimal, right?
Drubin, It appears as though you have a poor match. I hate to confuse you any more but your cables can effect this equation as well. Would you mind posting the rest of your system?
11k would be right (provided Pass did not do something unconventional but I cannot think of what that may be). 1700 ohm preamp is on the high side-11k for an amp is not too bad. If you could hook up your system to an oscilloscope you could find out for sure. Mismatches like this show up as a quasi-sinusoidal wave in the pass band (instead of a totally flat one) which causes excess distortion to the signal - easy to see on an oscope. I would think you are indeed getting distortion. Good luck - Arthur
Thanks guys.

I use HT Pro Silway II 2M interconnects between the two.
Could you give us the complete system?
Interesting coincidence, "sounds" like my problem as well. I posted the thread just below this one. Any help on that thread would be greatly appreciated. Cal Audio Alpha DAC mismatch with Adcom GFP 750 Preamp.

So...I asked Nelson Pass (duh!). He says that, yes, it is 11K single ended. He added that I could insert
resistance in series with the input if I like. The gain of the X150 is about 30 dB, so if I put 22K in series I would still get 20 dB gain.

What do people think of this idea?

By the way, the reduction in gain should not be a problem, at least on the face of it. The lowest setting of the First Sound is sometimes too loud as it is.
The insertion of a series resistance in order to achieve higher input Z may be worthwhile, or may introduce some other undesirable nasties & so could be a tradeoff situation.
In order to do this properly you'll need to get high quality resistors & also find a place to interrupt the signal path, possibly at a point where an input signal coax is running from an input-RCA connector to an input board. If you can desolder the coax connection from the input board & solder in a resistor there then that would likely be a good place. Regarding the proper selection of resistors, if you can get the same brand & type that Pass uses in their equipment then the sonic signature would be compatible with what's already in there. If not then make a call to Michael Percey audio at:
Thank you, Bob. Great info! I think that making this kind of modification is beyond my comfort level, however. Would there be any way to fashion an interconnect "extension" that would add the resistance and let me see, on a diagnostic level, whether there is a significant change in the sound if I increase the impedence?
Two posssible approaches Dan...
A female-to-male add-on extension RCA connector set could be utilized with the series resistance incorporated therein. However just by adding the extension (even without any extra resistance built into it) you would likely change the sonic signature. I suppose that you could listen to the extensions first (without extra series resistance) in order to extablish a baseline, then add the resistors & listen again. But if you're not solder-savy then you would need someone to build this for you & then revise it later.

Another possibility would be to get some inexpensive (disposable)interconnects that allow the RCA connectors to be easily opened up, then add the resistance right inside the connector. Again, listen awhile without the extra resistance, then add it & listen again. Be sure to allow for cable breakin phenomenon too. However by using a cheapie cable, you're probably not going to realize optimum sonic performance either before or after; still you might be able to detect a change? But again if you're not solder-savy then you would need someone to do this for you.
I think the second suggestion is a winner. I'm pretty sure I can handle the soldering, but--and I'm embarrassed to ask this--the resistance should be added to the "hot", right, and that is which, the inner plug or outer jacket on an interconnect? And, are resistors directional?

While I'm at it, I am also embarrassed to discover that I have been misspelling impedance all this time.
Drubin, don't feel bad I misspelled the name of the item I have up for auction, twice! Then spelled the word in the manufacturers name correctly in the text. Talk about embarrassing.
Dan yes: the inner conductor. Direction of the resistor is not an issue.
Be aware however that the input sensitivity of the amp will be reduced somewhat, requiring a bit more drive level from the preamp source component.