[JakeJ]

Hi All,

I tried this question elsewhere but didn't get a response.

Is a big part of the sonic difference between two amps due to the difference in input impedance/sensitivity?  I know the input impedance on my Dyna/Curcio amps is 490K Ω with a sensitivity of 1.8 volts for full output and my VAC PA-160s have a very low sensitivity of 0.3 volts for full output.  Or is it due primarily to a difference in components in the circuit?  Capacitors and resistors.  I know that some part of it is due to the difference in circuit topology itself but I am wondering if  some cap rolling might help open up the Dynas as they sound dark and somewhat rolled off on the top.  I'm guessing the answer is quite a bit more complex than I realize. 

Thanks,
JakeJ

[Niteshade]

The shortest answer to your questions is YES to all of the above. You're describing circuit differences and the circuit is what's responsible for at least 80% of the amp's voicing. Gain stages can add noise and there's also the question of how much feedback may be used and whether it's global or local. The other 20% or less of the voicing qualities comes from materials used to make the amp.

[JoshK]

I would have to agree with Blair.

[JakeJ]

Good Morning, Mr Lamphear,

Thank you for answering my question.  All the little lights went on my brain cavity.

The shortest answer to your questions is YES to all of the above. You're describing circuit differences and the circuit is what's responsible for at least 80% of the amp's voicing.

So, if I understand correctly, ±80% of an amplifier's voicing (overall sound character) is due to the design and implementation of the circuit.  How the topology is laid out and how all the components network electrically.

The other 20% or less of the voicing qualities comes from materials used to make the amp.

The components chosen such as capacitor, resistor, inductor, wire, pcb (if used), and tube types and the materials they are made of contribute the remaining ±20% of the voicing.

Gain stages can add noise and there's also the question of how much feedback may be used and whether it's global or local.

This helps me better understand how the amps I mentioned can sound so completely different despite the fact they currently run similar output tubes and the input and driver stage tube types are not the major contributors to the sound character of the two. 

Short take, I can change caps all I want in the Dynas but will most likely not make them sound like the VAC amps.

Thanks again,
JakeJ

[Niteshade]

Hi Jake-

Your are correct. The voicing of an amplifier is primarily developed by the circuit. Changing tubes, resistors, etc... creates subtle tweaks compared to the big picture.


BTW: I like you new circle! 

[anthony a.]

in my many years of trying out different gear, i always found that synergy between amp and preamp was paramount to sonics.  when i say synergy, 90% of the time it had to do with mathcing gain and sensitivity.  too high gain on an amp caused hiss, too low and you hit distortion.  it was a fine balance to find and match an amp and pre.

[JakeJ]

in my many years of trying out different gear, i always found that synergy between amp and preamp was paramount to sonics.  when i say synergy, 90% of the time it had to do with mathcing gain and sensitivity.  too high gain on an amp caused hiss, too low and you hit distortion.  it was a fine balance to find and match an amp and pre.

I agree with this observation also.  I hit the Audio Lottery when I matched a highly modified McCormack ALD-1 SS preamp and the VAC PA160 tube amps.  Unfortunately, I broke one of the fancy WBT RCA jacks the other day when trying to insert an I/C plug that had its locking collar closed.  I really don't care for locking RCA plugs as this hasn't been the only problem I have experienced.  I'm using my backup Quicksilver full function preamp and it's just not the same. 

Despite the fact that the ALD-1 and PA160s play well together I am still considering a VAC preamp for that last bit synergy.  aa

Niteshade - Thanks for the kudos on my circle.  I find your contributions to AC of high value and look forward to learning more from knowledgeable folks like yourself. 

Recently I have been rolling 6SN7 tubes in the VACs and found this to be a subtle form of tone flavoring.  Many times people use the term "tone control" but I associate that term to tone controls in older equipment that emphasizes or de-emphasizes treble or bass.  With tube rolling and using the term "tone" I associate that with the way a musician uses the term in reference to one brand of instrument versus another.  This applies to violins, trumpets, and triangles as well as guitars.  YMMV on this statement. 

Cheers,
JakeJ

[avahifi]

I will try and answer the original question raised here, which is a good question.

How will the amplifier's input impedance affect the sound?

The input impedance is most of the load that the preamplifier attached to the amplifier will have to drive.  The rest of the load essentially is the distributed interconnect cable capacitance.  As the input impedance goes up, the load driving requirements for the preamplifier are reduced.  This is probably not an issue with solid state preamplifier which typically have a low output impedance and reasonably good output drive current capacity (to charge and discharge the distributed cable capacitance linearly on each duty cycle).  However the typical vacuum tube preamplifier has a much higher output impedance and much less output drive current capacity, lowering its ability to drive a load.  A much lower than normal input impedance for the power amp (less than 10,000 ohms or thereabouts, and a long and rather high capacitance interconnect cable can actually represent a load that causes the tube preamplifer to get into big trouble and distort.  The effects are subtile, normally a reduction in gain, a reduction in dynamics, the sound getting flatter and the highs getting a bit edgy.  These effects are not well understood by many, and often the advice to use long expensive multistrand braided interconnect cables from a tube preamp to a low input impedance solid state amp is contrary to good engineering practice.  So in general, if you are using a tube preamp, a good idea is to select a power amplifier with a relatively high input impedance, at least over 20,000 ohms, and drive it with short, low capacitance interconnect cables (we like Blue Jeans cables for well engineered ones).

Input sensitivity is a separate issue.  In general the more sensitive the amplifier is (the more it amplifies the input signal - - - a typical number is about 25 times) then the less input voltage will be required from the preamp.  This can be a good thing if the preamp output circuit is pretty feeble, but it is a bad thing with respect to the signal to noise ratio.  The output noise from the preamp line section is fixed.  It originates from the active line circuit which is normally after the volume control.  Thus as the drive signal goes down, the relationship between the music signal and the fixed noise signal will go down  - - -  the signal to noise ratio gets worse.  In general, a high input sensitivity amplifier connected to very efficient speakers is to be avoided as both the amplifier and the speakers will amplify the fixed noise from the system much more than normal.  In addition, the amount of useable volume control will be reduced as only a small fraction of the available output signal will be needed to drive the system to clipping.  The volume control is designed to provide functional use from off to nearly full on, a bad combination of equipment will allow only 1/4 of that rotation or worse, and may never allow quiet enough operation for your musical enjoyment.  For the best possible signal to noise ratio, and lowest output noise, a standard to low gain amplifier is preferred, along with medium to low efficiency speakers.  Of course this may be contrary to your musical choice in amplifiers and speakers.  Everything is a compromise.  Just stating the facts, as Sargent Friday would request.

One other factor given input sensitivity should be considered.  Input sensitivity is a measure of the voltage gain of the amplifier.  How many volts in are required to drive the amplifier to full output voltage.  Two things affect this number.  First, what is the maximum output voltage of the amplifier?  Second, what is the gain (voltage multiplication factor) of the amplifier?

Remember that power is a squared function, not a linear function.  The normal expression of amplifier maximum output power is the maximum voltage before clipping (where the circuits stop amplifying) multiplied by 0.707 (to change peak power to continuous power ratings) and that result squared (times itself) and that result divided by the load resistance (usually 8 ohms).

For example if the amplifier can make plus and minus 10 volts peak into an 8 ohm load, the power rating would be (10 x .707) squared, divided by 8 which equals 6.25 watts.  In general this measurement is done with both channels driven with in phase signals both into 8 ohm load banks.  Driving only one channel yields an inflated number as the common power supply will only need to supply one channel and normally will increase the power rating number unrealistically.  Note that if we double the maximum voltage to 20, then the power is 25 watts, four times as much.  Interesting, we only doubled the maximum signal size, from 10 volts to 20 volts, but we quadrupled the power!  Things are not always what they seem to be.

As an aside, gas mileage should be measured in "gallons per hundred miles' rather than as "miles per gallon" because miles per gallon is also a exponential factor, not linear with the same misleading numerical results.  But that's another story.  Math majors out there, think about that please.

Now back to sensitivity.  In the two cases above if both amplifiers had the same normal gain (25) then the input sensitivity of the the 10V amp would be 0.4V as it would take 0.4V in to get to 10V out.  The 20V amp would have an input sensitivity of 0.8V because it would take 0.8V to get to full 20V out.  Interesting again, because both amplifiers have the same gain of 25.  The input sensitivity does not provide useful information unless you know the gain and the maximum voltage possible too.

Finally to stir the mix a bit more, the overall gain of an amplifier is usually set by the closed loop feedback.  In general the open loop gain is the max gain of all the devices as used in that circuit configuration.  Because the absolute gain of each device is different, trying to build an amplifier with no closed loop feedback is a nightmare operation.  Each circuit would have to be individually trimmed by hand to make up for the gain of each in depended tube or transistor used.  Repeatability is nearly impossible to achieve, and as the actual gain of the devices drifts with heat and age, it is very difficult to keep the operational parameters of the circuit constant over time.  Using closed loop feedback does two things.  First it reduces the gain of the overall circuit to a very consistant and predictable value.  The feedback loop is set with stable passive devices in general, and the gain remains constant over big variations in device gain changes.  The closed loop feedback, within reasonable design considerations, also reduces the overall harmonic distortion of the circuit.  In general, more feedback reduces the overall closed loop gain of the circuit and reduces overall harmonic distortion.  Less overall feedback increases the overall gain of the circuit and increases harmonic distortion.  Since the sensitivity number is the result of knowing the closed loop gain and the maximum voltage swing, that number without the knowledge of the other factors is not very informative, other than knowing that a high sensitivity amplifier, all other things being equal (and they rarely are) will yield worse overall noise performance in the system.

Finally, with the feedback, there is no free lunch.  Higher feedback can increase transient intermodulation distortion, as the amount of feedback generates an impossible to correct signal and overloads the amplifier input circuits.  Lower feedback reduces that effect, but increases harmonic distortion.  High TIM often can be heard as a hard, grainy, bright overall sound.  Low feedback can make the sound muddy, bass boomy, and the clarity suffer.  These are interesting design issues for the electrical engineer to deal with.

So there is no way to predict just what the input sensitivity of your amplifier will mean in the overall scheme of things, other than high sensitivity will likely mean higher noise, especially with very efficient speakers. However, with a tube preamp a high sensitivity amplifier may be a positive factor by reducing the load on the preamp.  Your results may vary.

There are exceptions to every general rule of course, but I hope this helps your understanding of these issues.

Best regards,

Frank Van Alstine

P.S.  As suggested above "parts sound quality" is an almost negligible issue in relation to the design issues expressed herein, as long as the parts selected are appropriate for the application (and that simply requires good engineering decisions too).  Second guessing the engineers without an understanding of the design issues are likely futile and expensive exercises.

[JakeJ]

Good Morning, Mr. Van Alstine,

Thank you for answering my first question and explaining what input impedance is and how it differs from input sensitivity as well as their collective effect on an amplifier's overall sound character.

Mr. Lamphear's answer dealt more with my second question but the two are intertwined and I have a much clearer understanding of what is happening in my system.

In addition, the amount of useable volume control will be reduced as only a small fraction of the available output signal will be needed to drive the system to clipping.  The volume control is designed to provide functional use from off to nearly full on, a bad combination of equipment will allow only 1/4 of that rotation or worse, and may never allow quiet enough operation for your musical enjoyment.  For the best possible signal to noise ratio, and lowest output noise, a standard to low gain amplifier is preferred, along with medium to low efficiency speakers.  Of course this may be contrary to your musical choice in amplifiers and speakers.  Everything is a compromise.  Just stating the facts, as Sargent Friday would request.

With the VAC PA160 amps I am suffering from a bit too much gain and the result has been a loss of fine volume control. Noise levels however are excellent and I do not have any hum or extraneous noise coming from the speakers at normal listening volume with no music signal and my ear right next to the drivers.  After rereading the manual some of this may be attributed to operator error.  I had selected a lower than recommended feedback setting for the tube type and mode of operation.  I'll need to play around with that some more and see if I can find a "happy place" for my combination.

Again, I thank Frank and Blair for shedding a great deal of light on a subject that my question asked a bit too simply for an easy answer.  Of course, electronics are always too complex for there to be any real easy answers.

I wish there was an electronics course available at the local community college so I could learn and understand networking.  I can read a schematic and build a circuit from it but I don't truly understand what is going on within it. 

Cheers and many thanks gentlemen,
JakeJ

[Audiovista]

P.S.  As suggested above "parts sound quality" is an almost negligible issue in relation to the design issues expressed herein, as long as the parts selected are appropriate for the application (and that simply requires good engineering decisions too).  Second guessing the engineers without an understanding of the design issues are likely futile and expensive exercises.

 

Especially last sentence.

[charmerci]

As an aside, gas mileage should be measured in "gallons per hundred miles' rather than as "miles per gallon" because miles per gallon is also a exponential factor, not linear with the same misleading numerical results. 

Best regards,

Frank Van Alstine

They actually do that in Europe, i.e. liters per 100 kilometers.

[doug s.]

jake, if you want more range w/the wolume pot of your preamp, something as cheap & simple as this, inserted between preamp and amp might do the trick:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=300304979582

i use similar rothwell attenuators, and have noticed no change at all in sonics, tho others report a difference.  the goldenjacks are a bit less expensive.  you could also try the premium evs attenuators.

doug s.

[JakeJ]

Thanks, Doug, I have scored a set of the Rothwell attenuators and they came just after my days off so haven't had a chance to try them yet.  I did notice they have a recessed female jack that may give me some connection problems but will have to put them in to see.  Dunno if I can trim the barrel down and will communicate with the seller before I do.

Jake

[vett93]

I am not sure I would agree that "parts sound quality" is an almost negligible issue. From my experiences, the coupling caps have significant impact on the sound quality. Replacing "normal" coupling caps from a well designed amp with "high end" caps like V-caps or Mundorf Silver/Gold & Oil usually yields better sound quality.

[turkey]

As an aside, gas mileage should be measured in "gallons per hundred miles' rather than as "miles per gallon" because miles per gallon is also a exponential factor, not linear with the same misleading numerical results.  But that's another story.  Math majors out there, think about that please.

I'm not seeing this.

Gallons are measured linearly, and so are miles. (They don't go 1 gallon, 2 gallons, 4 gallons, 16 gallons.)

Gallons per hundred miles is equivalent to gallons per mile, just the ratio differs.

Gallons per mile or miles per gallon are equivalent too.

1 gallon per 100 miles equals 100 miles per gallon. (And also equals 1/100 gallon per mile.)

2 gallons per 100 miles equals 50 miles per gallon.

3 gallons per 100 miles equals 33 1/3 miles per gallon.

4 gallons per 100 miles equals 25 miles per gallon.

10 gallons per 100 miles equals 10 miles per gallon.


I do agree that gallons per 100 miles is a better way of stating things. It's then very easy to look at a car that gets 3 gallons per 100 miles and see that if gas is $2 a gallon you're going to pay $6 per 100 miles or 6 cents a mile.

If you state it as 33 1/3 miles per gallon, it is less obvious (for me anyway) to see that right off as 6 cents a mile.

I also tend to look at long term use of a car. 100K miles is a good figure to look at. With that car that gets 3 gallons per 100 miles, it's easy to see that it will use 3000 gallons over its lifespan, and that could cost $6000. That makes it more easy to compare cars and see what they're really going to cost you.

But where's that exponential factor? I'm just not seeing it. I guess that's why I wasn't a math major.

[Niteshade]

What type are you calling normal?

I am not sure I would agree that "parts sound quality" is an almost negligible issue. From my experiences, the coupling caps have significant impact on the sound quality. Replacing "normal" coupling caps from a well designed amp with "high end" caps like V-caps or Mundorf Silver/Gold & Oil usually yields better sound quality.

[markC]

I'll call "normal" polyester caps. I've found that replacing polyester coupling caps in CD players and DACs a very noticeable improvement in clarity and detail. Good polypropylene's seem to provide a more "open" sound.

[turkey]

I will try and answer the original question raised here, which is a good question.

How will the amplifier's input impedance affect the sound?

Frank, this was a very interesting article. Good job!

[turkey]

I'll call "normal" polyester caps. I've found that replacing polyester coupling caps in CD players and DACs a very noticeable improvement in clarity and detail. Good polypropylene's seem to provide a more "open" sound.

I thought that coupling caps don't really charge and discharge during operation of the circuit. That would lead me to believe that a "superior" dielectric is not going to work any better here, providing the cap itself is suited for its intended use (high enough voltage capacity, within the recommended temp range, physical size, cost, etc.)

[vett93]

You may want to try replacing the coupling caps with some "high-end" caps. You will be nicely surprised.....

Also, the couple caps do get charged and discharged around a DC point (anode voltage).