What ohm to set amp

Thanks so much Al. But as I said, I'll keep my day job. ;>')

As to Ralph's comment about tonal balance problems, I suppose using a tube amp that has a high output impedance implicates the mismatch issue more so than using a tube amp that has a lower output impedance.

You've got it correctly!

Best regards,
-- Al

Al, thanks again for the education. So to restate what you said above, "[t]he tighter the regulation, meaning the smaller the +/- number, the LESS the voltage will increase as load impedance increases, and the LESS the voltage will decrease as load impedance decreases."

"An ideal voltage paradigm amplifier will have a regulation of +/- 0.0000 db. Its output voltage will not change at all as load impedance varies. And, per Ohm's Law, for a given output voltage the amount of current drawn by the load will vary inversely with the impedance of the load. So since power is proportional to voltage x current, and equals voltage x current in the case of a purely resistive load, the power delivery of a voltage paradigm amplifier will NOT "correspond to the magnitude of the input signal."

And as to Ralph's comment about tonal balance problems, I suppose using a tube amp that has a high output impedance implicates the mismatch issue more so than using a tube amp that has a lower output impedance.

Last and final point: I think I'll keep my day job.

And Al .... thanks again for the education.

Bruce

I gather that output regulation relates to how closely a tube amplifier's voltage output will vary so that the amount of power (watts) presented to the speaker will correspond to the magnitude of the input signal, thus compensating for varying speaker impedances which change as a function of FR. In short, if impedance increases, voltage will also increase. This in turn will increase current. All of this is accomplished through NF.

No, that's not right Bruce. Output regulation in this context refers to how little or how much the amplifier's output voltage will vary as a function of load impedance, for a given input signal. Period. The magazines define it based on the range of impedance variation of a particular simulated speaker load. The ARC specs define it based on a load variation from 8 ohms to infinity ohms (i.e., an open circuit). The tighter the regulation, meaning the smaller the +/- number, the LESS the voltage will increase as load impedance increases, and the LESS the voltage will decrease as load impedance decreases.

An ideal voltage paradigm amplifier will have a regulation of +/- 0.0000 db. Its output voltage will not change at all as load impedance varies. And, per Ohm's Law, for a given output voltage the amount of current drawn by the load will vary inversely with the impedance of the load. So since power is proportional to voltage x current, and equals voltage x current in the case of a purely resistive load, the power delivery of a voltage paradigm amplifier will NOT "correspond to the magnitude of the input signal."

And a voltage paradigm speaker is, by definition, one that is designed based on the expectation that it will be driven by an amplifier which behaves in that manner. Such as the small two-way I described in my previous post.

If a SS amp is asked to feed current into a high impedance segment of the speaker's FR spectrum, somehow the amp must increase its current output or else power (watts) will decrease.

In that situation power WILL decrease, but if the speaker is (correctly) designed with the expectation that it will be used with a solid state amplifier, having negligibly small output impedance, the acoustical output of the speaker will nevertheless be correct. In other words, a voltage paradigm speaker whose impedance varies significantly as a function of frequency will require less power at frequencies for which its impedance is high than at frequencies for which its impedance is low, to produce a given amount of acoustical power.

If I am tracking so far, what I don't understand is how a so called zero NF amp (tube or SS) can properly respond to impedance values which change as a function of FR in order to maintain constant power through a speaker's FR spectrum.

So at this point it should be clear that a voltage paradigm amp does NOT maintain constant power as load impedance varies. That is the behavior that a power paradigm amp, having relatively high output impedance and generally minimal or no feedback, will approximate to some (usually loose) degree.

As far as feedback is concerned, an amplifier whose output impedance is negligibly small in relation to the load impedance, at all frequencies, will behave as a close approximation of an ideal voltage source, maintaining an output voltage that does not vary significantly with load impedance (as long as it is operated within the limits of its voltage, current, power, and thermal capabilities). Given a low enough output impedance (as in the case of the Ayre amplifiers) it will do that to a close approximation regardless of how much feedback it does or does not use.

Two closing points:

1)Feedback, as it is usually applied in this context, does not cause output power to "correspond to the magnitude of the input signal." It causes output VOLTAGE to more closely correspond to the magnitude of the input signal (multiplied by some gain factor).

2)As Ralph has pointed out in the past, tonal balance problems result from using speakers with amplifiers that are not of the same paradigm.

Best,
-- Al

Thanks Al. Would you kindly clarify my understanding of the Voltage and Power Paradigm area just one more time.

At least with my tube gear, ARC publishes so-called output regulation stats. In the case of both the Ref 150 and VS-115, output regulation off the 8 ohm tap is +/- .8 db and probably 60% of that number off the 4 ohm tap. See Stereophile and Soundstage bench test reports.

Perhaps I am misinterpreting what the term "output regulation" means. So please help me here. Based on the explanations provided in the two magazines, I gather that output regulation relates to how closely a tube amplifier's voltage output will vary so that the amount of power (watts) presented to the speaker will correspond to the magnitude of the input signal, thus compensating for varying speaker impedances which change as a function of FR. In short, if impedance increases, voltage will also increase. This in turn will increase current. All of this is accomplished through NF.

As I understand the Voltage Paradigm, SS amps generally maintain constant voltage and increase or decrease power in response to speaker impedance changes. But here again, if a SS amp is asked to feed current into a high impedance segment of the speaker's FR spectrum, somehow the amp must increase its current output or else power (watts) will decrease. Here again, I would think that NF would throttle up or down as the case may be the amount of the amp's gain in order to compensate for these variations.

I surmise that since Voltage Paradigm amps naturally increase power in low impedance loads, it is easier for the SS amp to deliver current and ultimately power into the bass regions. In contrast, if presented with a high impedance segment of a speaker's FR spectrum, the amp will need to deliver much more current in order to "power" its way through the speaker's impedance mountain, and may choke. Again this is where NF comes into play ... I think.

If I am tracking so far, what I don't understand is how a so called zero NF amp (tube or SS) can properly respond to impedance values which change as a function of FR in order to maintain constant power through a speaker's FR spectrum. I would have guessed that some type of NF, either global or local would be needed. It would seem that the only way to avoid NF is to use a speaker that has extremely flat impedance curves and phase angles.

Btw, I was reading some of the tech data on the Ref 150. Can't say I understand it, but from what I gather, ARC has somehow directly coupled the power tubes to the output trannies in order to regulate power output. I don't know if this is just another term for local feedback, but that's as much as I understand. I recall ARC used the term partial cathode following, or something like that.

Sorry for getting back into the tech stuff again, but all of this touches on the OP's question and the area is confusing, albeit interesting, to this lay person.

Kalbi, I couldn't find a complete impedance curve for the Liuto, but I did find the following comment in this review:

... a nominal impedance of 6 ohms would be more representative than the quoted 8 ohms.

Measured data was presented indicating a minimum impedance of 4.5 ohms at 124 Hz, a current-hungry phase angle of -54 degrees at 68 Hz, and an "equivalent peak dissipation resistance" of 2.2 ohms.

And the measurements in this review indicate a minimum impedance of 4.1 ohms at 100 Hz, at a phase angle of -23 degrees.

Given that those impedance minima occur in the bass region, where lots of energy is typically required, and given also that the manual for the amplifier specifies that the 4 ohm setting should be used for a 6 ohm speaker, it all seems to add up to the 4 ohm setting being best.

My guesses are that the automatic switching mechanism that was referred to may involve circuitry whose primary function is protection, and that its operation may not be sonically seamless.

Bruce, regarding your question consider the case of a small two-way speaker. The combination of small woofer size and small cabinet size will result in low efficiency in the bass region. It is not uncommon for that kind of speaker to have an impedance of close to 4 ohms in the bass region, and close to 8 ohms at higher frequencies. Generally that kind of speaker will conform to the voltage paradigm, and will be intended for use with a solid state amplifier. The near zero output impedance of the amplifier will result, for a given input signal level, in twice as much current and power being delivered at low frequencies as at high frequencies. That doubling of power at low frequencies will compensate for the low efficiency of the speaker at low frequencies, presumably resulting in a flat overall frequency response. If the amplifier were to deliver the same number of watts to that speaker at high and low frequencies, for a given input level, the response would be decidedly bass-shy.

Best,
-- Al

Al, I am somewhat familiar with the Ayre being touted as a zero feedback amp. So if it doubles power if the load drops in half, and halves power if load doubles, how does the amp manage to regulate the amount of current being shown to the speakers so that the speaker's acoustic output is flat?

It seems to me that regardless of whether voltage remains constant or current remains constant, ultimately the SPL generated by a speaker is based on power (i.e., watts, or the product of volts and amps). In cases where a SS amp or tube amp uses NF to throttle either volts or current, one way or the other it seems logical that the amount of watts presented to the speaker's voice coil should in some way correspond -- or perhaps more accurately mimic -- the frequency and amplitude of the wave form presented to the amp's input.

I recall reading a member's post some months ago that expressed some doubt that the Ayre is able to dispense with all forms of feedback, albeit local or global. I believe the member may have expressed some puzzlement how such an amp could adjust to changing impedances as a function of FR and thereby maintain a flat acoustic presentation. Unless of course the speaker in question has a ruler flat impedance curve and zero phase angle across the entire frequency spectrum.

Thank you everyone for taking the time to research and post. The SF speakers I have are the Liuto towers and the matching center the smart. I talked to a rep at Mac, and he told me doesn't matter if I have it set at 4 or 8 ohm, as the amp would switch on its own when needed. But he also asked if the amp was running hot at the 8 ohm setting, since this is my first Mac, I didn't know how hot it is suppose to run. I'm wondering why would it matter if it switches on its own.

Does that permit the inference that a SS amp that operates as a Voltage Paradigm amp would be expected to use NF of some type (local or global) to regulate actual power output into loads which vary as a function of FR in order to maintain output power that corresponds to the signal presented at the amp's inputs??

Usually voltage paradigm characteristics (i.e., maintaining constant voltage, not power, into varying load impedances) go hand in hand with the use of some amount of feedback. But that is not always the case. For example many and perhaps all of the Ayre amplifiers use zero feedback, yet have output impedances of a small fraction of an ohm. As a consequence of that low output impedance they will deliver essentially the same voltage and very close to twice as much power into 4 ohms as into 8 ohms when operated within their maximum power capabilities. And, in addition, those maximum power capabilities double into 4 ohms relative to 8 ohms.

Best,
-- Al

That's kinda what I figured. In a sense, the 4 ohm switch is protective.

So .... as you mentioned if the impedance switch is set to 8 ohms and the amp is pushing power into a 16 ohm load, the power output should be roughly one-half the amount which can be delivered at 8 ohms.

Al, does that permit the inference that a SS amp that operates as a Voltage Paradigm amp would be expected to use NF of some type (local or global) to regulate actual power output into loads which vary as a function of FR in order to maintain output power that corresponds to the signal presented at the amp's inputs??

Btw, sorry. Sometime I confuse you and Ralph. The last sentence of my previous post was directed to vous mon frer.

Bruce

Thanks Bruce; I've learned a new word today :-)

As I indicated, the MC205, as well as the MC solid state amplifiers that use autoformers (which the MC205 does not use), have negligibly small output impedances. The MC205 is spec'd as having a damping factor of 140, corresponding to an output impedance of 8/140 = 0.06 ohms. So it will act as a voltage source with respect to the interaction of its output impedance with the speaker's impedance vs. frequency variations.

Which means that for a given signal level at the input to the amplifier it will output twice as much power at frequencies for which the speaker's impedance is 4 ohms as it would deliver at frequencies for which the speaker's impedance is 8 ohms, and half as much power into 16 ohms as into 8 ohms.

PROVIDED, however, that the maximum limits of its power capability are not exceeded. What will NOT double into 4 ohms, compared to 8 ohms, is the MAXIMUM amount of power it can put out. And I suspect that if the rear panel switch were set to 8 ohms and the impedance of the speaker is close to 4 ohms in most of the bass region, its MAXIMUM power capability into that 4 ohms would be significantly LESS than the 200W it could deliver into 8 ohms at that setting.

Basically, what the switch is apparently doing is optimizing internal voltage, current, and thermal operating conditions for whichever of the two load impedances is selected, such that a maximum of 200W can be delivered into either impedance.

Best,
-- Al

Thanks Al. Glarff means an inchoate stream of consciousness. Going back to the OP's Q re the MAC SS amp, I am just guessing here, but perhaps the 4/8 ohm switch avoids power doubling at 4 ohms if the 4 ohm switch is used. Perhaps the amp is more stable operating in that power configuration??? Dunno.

Of course the trade off is that many speakers have roller coaster impedance curves as a function of frequency response. So, I further surmise that if for discussion purposes, the OP's SF speakers have an impedance rating of 12 or 16 ohms at the midrange/tweeter x-over point (just guessing), the MAC's ability to produce power at that frequency point might be considerably less (maybe 25% if the x-over impedance is 16 ohms) based on the Voltage Paradigm approach. I guess that's where NF comes into play to ramp up power output or else the speaker's acoustic output will make the listener dizzy.

Sorry Ralph, more glarff.

I took a look at the literature on the MC205 at the McIntosh site. It is a solid state amplifier that has neither output transformers nor autoformers nor multiple output taps. It has a rear panel switch for selection of either 4 or 8 ohm speaker impedances. Apparently the function of the switch is to select higher internal voltage rails (the DC power supply voltages that are provided to the output stage) when an 8 ohm speaker is connected, compared to when a 4 ohm speaker is connected. The manual recommends using the 4 ohm setting if a 6 ohm speaker is being used. The power rating of the amplifier is 200 watts into both 4 and 8 ohms, assuming that the load impedance corresponds to the switch setting.

Kalbi, what specific models are your Sonus Faber speakers? Knowing that would hopefully allow us to look up their impedance curves. In any event, if the dips into the 4.1 to 4.7 ohm range occur in the bass region, where a lot of energy is typically required, it seems clear that you SHOULD use the 4 ohm setting, as Schubert indicated. And likewise for the 6.2 ohm center speaker, although the selection is presumably less critical with respect to that impedance, and in any event the amplifier just provides one overall selection between 4 and 8 ohms, not independent selections for each of its channels.

Bruce (Bifwynne), I don't know what a glarff is, but I'll comment on it anyway :-). Both this amplifier and the Mc solid state amplifiers that have autoformers have output impedances that, as with most solid state amplifiers, are a tiny fraction of an ohm regardless of tap or setting. So they will act like typical solid state amplifiers with respect to the interaction of their output impedance with the impedance vs. frequency variation of the speaker.

As to why using a tap on a tube amplifier that is intended for an 8 ohm load can sometimes be preferable to using a tap intended for a 4 ohm load, as you realize some speakers conform to what Ralph refers to as the power paradigm, and are designed to sound their best when used with an amplifier having relatively high output impedance. Also, a mismatch between the actual load impedance and the impedance a given amplifier output tap is designed to drive can, if severe enough, degrade the sonic performance of the amplifier, as Ralph indicated above.

Best regards,
-- Al

Kalbi .. , don't know much about the Mac or SF impedance curves. So I only add this comment because it might help focus your discussion with MAC tomorrow.

As I've come to better understand with my own gear, using the lower impedance taps on my tube amp enables me to drive my speakers with a lower output impedance source. Of course, my amp uses output transformers and negative feedback, but doesn't have built-in autoformers like some MACs.

Frankly, I am a little puzzled why one would want to use higher impedance taps if the price is higher output impedance. Most tube-heads may say use whatever taps sound better to you. That's fair advice ... but what sounds better to you might be acoustically colored playback especially if the amp's output regulation varies more widely with respect speaker impedance which varies as a function of frequency.

I've killed this issue a zillion times already and if I start it up again, I'll start a negative feedback war. Yes ... NF comes with a cost. Read Ralph's white paper on the Voltage and Power Paradigms. But understand that a tube amp that uses NF does not perform like a pure Power Paradigm source. It's kinda' a hybrid which is good in some respects, but also inserts TIM and odd-ordered harmonic distortion into the output -- not so good. But life, and amplifier design, is about compromises. You takes the good with the bad.

Cheers,

Bruce

Perhaps Ralph or Al will edit my glarff so it makes more sense.

Thanks guys, I called a SF rep and he said I should run them at 8 ohm since they are 8 ohm speakers. I'll contact Chuck at Mac and see what he saids.

Yes , and you should if a speakers dips to 4.1

I went over to the Mac site and the amp is rated for 4&8 ohms. If you'd like detailed info or any info really, e=mail Chuck Hinton of Mac. He can answer most any question you have.

If you run an 8 ohm speaker on the 4 ohm tap there will be no reliability consequences, although it may not sound right (which is a different matter).