Impedance Question: Tube Preamp with SS Power Amp

Having researched this issue with the archived threads, I am still having difficulties understanding the problems of impedance mismatch between Pre and Power amps. As a budding audiophile moving onto my first separates Pre/Power, I want to know more about impedance matching. I have several questions that I hope veterian audiophiles can help me out with.

1)Is the 10:100 ratio between the output impedance of preamp and input impedance of the power amp an absolute rule?

2)What would happen if impedance mismatch do occur? Will the sound be any less satisfying?

3)I understand that Tube preamps generally have higher output impedance values compared to SS, so does that make it problematic to use it with a SS power amp even though tubed preamps are often priased over their SS counterparts?

I ask these questions because I recently purchased a tubed Preamp (Musical Design SP-2B)that has an impedance of 2K Ohms. I want to match that with a SS power amp with an impedance value of 20K Ohms. Am I cutting it too close? (exactlly 10:100 ratio) Or are all of these issues irrevalent and I should just listen to that setup and hear for myself if it would work or not? Thanks in advance for the answers.
Robert Harley's book "The Complete Guide to High End Audio (original version in paperbsck) has an excellent definition of input/output impedance on pages 395-397.

I use a Cary tubed pre-amp & a Musical Design SS power amp without problems. Would suggest you e-mail John Hillig of Musical Design to ask question with impedance value of SS amp. He clarified my concerns.
Krell cautions against the use of high impedence pre-amps with thier amps, but they add that some caps can be changed that would allow for the impedence miss-match.
Sonic Frontiers preamps have output impedance almost the same as SS preamps. Same with AudioResearch ones.
Answers by the number -

1) No its not an absolute rule is just a general guide

2) Rolloff of the highs and lows - probably some softing of the bass impact

3) I can see no reason why these shouldn't work in theory, but most folks would really want more leeway, like money in the bank so to speak. I've successfully used a 10K impedence amp with a tubed pre in which the manufacturer recommended a minimum of 20K (which was a 40 to 1 ratio). But as a practice, to keep my options open I usually only buy amps with imput impedences of 47K or more. In your place I would want to hear the amp - preamp combo before I commit.
Hovland HP100 into Pass X150. Output Imp. 2K Ohm quoted by Manufacturer but rise to 4700 Ohm towards 20Hz per stereophile test report. Input Impedance of Pass Lab approx 22K Ohm. This yields a 1/10 ratio at most freq but 4/10 at base. Theoretically should lead to base rolloff. In practise, HP 100 does not give up much base comparing to X1 preamp. 1/10 is a only a general rule. If system works, I wouldn't worry about numbers.
(If thread help, don't mind grammar)
You can help stack the deck in your favor by keeping the interconnect cable lengths SHORT. 1 or 2m is fine, but if you had to go to 15 ft of interconnect cable with this heavy of a mismatch, I'd say fuhgedabowdit.
Many tube preamps will also have a higher gain, which can offset any power losses from high output impedance. High frequency roll off can be an issue if the input circuit is shunted with a filtering capacitor (R-C impedance) but amps with resistive only impedance are not affected.
In well designed equipment, all impedances should match. This would give us the greatest power transfer, highest levels of linearity and least amount of signal reflections and the associated ringing that takes place.

Since we aren't dealing with well designed equipment, most manufacturers recommend using an amplifier with a much higher input impedance than the output impedance of the preamp being used. If one doesn't follow that basic "recommendation", the end result is typically a softening of highs and bloated, muddy, uncontrolled bass response. How severe this will be will depend on how stable the source or preamp is. The more stable that device is, the less noticeable these problems will be. In some systems with bright and lean sounding digital sources, the colouration / distortion caused by the improper loading conditions mentioned above may actually help the system to sound more bearable.

By playing games with various interconnects and altering the feedpoint impedance that the source or preamp sees, one can reduce these sonic drawbacks somewhat, but probably not completely correct them. As such, if one wants to "go beyong the territory of recommended guidelines", they have to be willing to "explore new horizons" on their own. That's because each situation / exploration will probably be different from the last and you'll have to "make things work" all over again. As such, get ready to do a lot of cable swapping i.e. "band-aiding" of the system to compensate for the selection of mismatched components. Sean
So I'm looking to hook up my AR Ref one MK2 with a pair of Krell 350mc's. Is this a mismatch? Must I do the internal jumper changes that Krell recommends?


The Krell amps are "direct coupled" - that is they don't put
capacitors in the signal path. This is an advantage since
"coupling capacitors" can have a negative impact on the quality
of the signal.

Krell amplifiers also have DC protection circuitry which will
shutdown the amp if too much DC signal is seen at the input.
Too much DC can be damaging.

Unfortunately, many tube amps output a signal with a DC
component above that which will trigger Krell's protection

In order to use a tube preamp with a Krell amplifier - Krell
advises that the normally inactive coupling capacitors in
the Krell amp be activated by a Krell dealer.

Dr. Gregory Greenman
So where do I go to find out if my AR Ref One MK2 is one of those tube preamps that puts out too much DC ???

I'd contact either Krell or Acoustic Research or both.

I'm sure you're not the first to hook an AR pre to a
Krell power amp.

Perhaps the two manufacturers have some experience - either
directly or garnered from their customers.

Dr. Gregory Greenman
Just to follow up - I spoke to Leonard at ARC and he clarified for me the fact that the Ref One has DC capacitors in it's output to filter DC from the Amp. So unless a tub blows or some such crisis the Krell's should be ok. He thought it to be a reasonable gamble to leave the jumpers in the Krell so as to keep the capacitors out of the signal path.

Thanks all
Impedance mismatches in this way affect (diminishes) the very low bass; some of those who say they don't have any problems may be due to the fact that they don't own really full range speakers.
Vladimir you're piqued my curiosity.
Can someone walk me through how to assess the impedance charactreristics of both my Krell 350mc's and the ARC Ref One Mk2. I'm sure I can go to the owner's manuals but I just don't have a clue as to what I'm looking at.


On page 26 of the current FPB-series manual; under the
column labeled FPB 350Mc - the input impedance is
specified as 100 kOhms.

In fact, all the amplifiers, both stereo and mono; and all
powers have an input impedance of 100 kOhms.

Dr. Gregory Greenman
Ok. Thaks for that. And if I go to my ARC manual, in the Spec section under Output Impedance it states:
400 ohms Balances, 200 ohms SE main(2),
20K ohms minimum load and 2000pF maximum capacitance.

In the Rated output section it states:
4V RMS (2V RMS SE) into 100K ohms balanced load.

My proplem is I don't know how to make sense in comparing the two.

AGH!!!!, so you wanted to be a carpenter...grasshopper



The Krell only has balanced inputs - so the balanced output
of the ARC is 400 ohms. [ Ignore SE = single ended ].

ARC states that it needs a minimum 20K ohm load. The Krell
is a 100K ohm load to the ARC - so it meets the minimum.

In the ARC output section, it states it can output 4 Volts RMS
to a 100K ohm balanced load - which is exactly what the Krell

So as far as impedance matching - the Krell is an acceptable
load for the ARC - the ARC is "happy" with the Krell.

The next question is whether the Krell is happy with the
ARC. The Krell doesn't want to see much DC at its inputs.

A previous poster stated the ARC has filter capacitors in
its output lines - which will block the DC the Krell doesn't
like. As long as one of the two puts the filter capacitors
in the line - the Krell will be happy.

Since the ARC already filters its output - you probably
don't need the Krell to filter its input. The Krell
monitors its input and will shutdown if it is not happy
with the input signal. In the unlikely event that this
happens - the jumpers in the Krell can be removed so that
the filter capacitors in the Krell are in the circuit path.

You should be OK - but if you have any doubts - call Krell.

Dr Gregory Greenman
Thank you. It all makes sense once it's explained to me.
They sure sound great together.