What are the advantages/disadvantages of different amp topologies?

[Hear Here]

If you want to get the most out of your amplifier investment, its always going to be best served by a speaker that's not low impedance and has higher efficiency.

I think I'm OK:

Impedance - 18 ohms

Sensitivity (sort of what used to be called efficiency) - 107 dB

 

[Atmasphere]

There is always an engineering problem , first things first , define higher Efficiency ..?

Let's put it this way: if the speaker is 4 Ohms and 86dB, its efficiency (1 Watt/1 meter) is 83dB. That's low enough that in most rooms you'll need 1000 Watts if you want lifelike volume levels. 'higher efficiency' would be something over 90dB. In the old days when tubes were king, likely more like 97 to 100 dB (since the speaker was likely 8 Ohms or more, sensitivity would be 97 to 100 dB, but if 16 Ohms 3 dB less, although in practice that makes no difference to a tube amp if it has 16 Ohm taps; that is why efficiency is a more useful spec if talking about tube amps).

 

[andromedaaudio]

I d love to hear a good Atmasphere OTL set up , i had a graaf GM 20 and GM 200 0TL 17 years ago ( ran very hot though )
I left Solid State in 2014 and see no reason to go back .
The breath of life is simply not there , the more i hear it the more i understand why i left and never looked back
And the solid state alternatives are expensive .
Robert Koda , Gryphon Apex ,FM acoustics ( may be Dartzeel )

 

[Atmasphere]

I d love to hear a good Atmasphere OTL set up , i had a graaf GM 20 and GM 200 0TL 17 years ago ( ran very hot though )
I left Solid State in 2014 and see no reason to go back .
The breath of life is simply not there , the more i hear it the more i understand why i left and never looked back
And the solid state alternatives are expensive .
Robert Koda , Gryphon Apex ,FM acoustics ( may be Dartzeel )

Thanks.

Times do change FWIW; technology marches on. I've been playing a class D amp at home that I think sounds as smooth as any tube amp and just as involving.

So if a solid state amp could give you all that the tube amps do, except perhaps the heat and having to deal with tube failure, would that make it worth it?

 

[andromedaaudio]

So if a solid state amp could give you all that the tube amps do, except perhaps the heat and having to deal with tube failure, would that make it worth it?

Possibly , everything is possible .

 

[Hear Here]

I d love to hear a good Atmasphere OTL set up , i had a graaf GM 20 and GM 200 0TL 17 years ago ( ran very hot though )
I left Solid State in 2014 and see no reason to go back .
The breath of life is simply not there , the more i hear it the more i understand why i left and never looked back
And the solid state alternatives are expensive .
Robert Koda , Gryphon Apex ,FM acoustics ( may be Dartzeel )

That's an interesting notion, although I find quite the reverse is true in the real 2024 world.

I was running tube amps for 2 decades (including a Graaf GM-20 but mostly SETs), until I realised how much these beasts were costing in keeping them in tip-top condition (new tubes, etc) and their appalling fuel consumption. I decided to look for a solid state amp, but only if it matched or bettered the sound quality I was getting from my SETs. A long series of purchases (new and used) or dealer demo amps over a year or so led me to conclude that you can get ss amps as good sounding as tube ones - and at a very much lower purchase and running cost and with far higher S/N ratio As a result I'm listening to even more enjoyable music and much more of it. A well-chosen Class D was my eventually purchase despite my expectation that it would be Class A that would win me over.

I now have 2 great Class D amps - one with Purifi's Eigentakt technology (though not a cheap one using the Eval board) and a pair of Atma-Sphere Class D monos. Would I go back to tubes? - not a hope in hell chance!

 

[andromedaaudio]

I d rather put my $$$ into digital

 

[Audiohertz2]

Let's put it this way: if the speaker is 4 Ohms and 86dB, its efficiency (1 Watt/1 meter) is 83dB. That's low enough that in most rooms you'll need 1000 Watts if you want lifelike volume levels. 'higher efficiency' would be something over 90dB. In the old days when tubes were king, likely more like 97 to 100 dB (since the speaker was likely 8 Ohms or more, sensitivity would be 97 to 100 dB, but if 16 Ohms 3 dB less, although in practice that makes no difference to a tube amp if it has 16 Ohm taps; that is why efficiency is a more useful spec if talking about tube amps).

Ok so you consider 90+ db for sensitivity as high with 97-100db as being very high ?

So low would be below 87db ..?

Its been my experience that very high sensitivity Speakers tend to be unbalanced for sound unless very complex and with multi driver. Usually by the time you get a proper balance top to bottom you have lost a lot of the sensitivity. Many today make it up by running powered subs if not it becomes pretty complex to make it all work.

With multiple drivers you lose efficiency as the impedance is gonna drop but sensitivity stays up , this is very observable in the wild , as you can see many SOTA loudspeaker attempts are with multiple drivers and complex as you half thd every-time you double driver area ..!

Since its not 1940 and high powered amps both toobs and SS are available i would say an 87db loudspeaker of multiple drivers is acceptable but not in a 2 driver 2 way design which will have high thd attempting to reach live levels ..!

Regards

 

[Republicoftexas69]

That's an interesting notion that I find quite the reverse is true in the real 2024 world.

I was running tube amps for 2 decades (including a Graaf GM-20 but mostly SETs), until I realised how much these beasts were costing in keeping them in tip-top condition (new tubes, etc) and their appalling fuel consumption. I decided to look for a solid state amp, but only if it matched or bettered the sound quality I was getting from my SETs. A long series of purchases (new and used) or dealer demo amps over a year or so led me to conclude that you can get ss amps as good sounding as tube ones - and at a very much lower purchase and running cost and with far higher S/N ratio As a result I'm listening to even more enjoyable music and much more of it. A well-chosen Class D was my eventually purchase despite my expectation that it would be Class A that would win me over.

I now have 2 great Class D amps - one with Purifi's Eigentakt technology (though not a cheap one using the Eval board) and a pair of Atma-Sphere Class D monos. Would I go back to tubes? - not a hope in hell chance!

I absolutely agree and I am thinking of selling my X200 as it rarely gets fired up anymore.

 

[DasguteOhr]

Tube amplifier power has always been expensive which is why vintage speakers prior to the solid state era tended to be horn-loaded.

But the simple fact is that any amplifier benefits from not having to work hard for a living! Put another way you might have a power house solid state amp that you can weld with, but making it drive a difficult load (low impedance, low efficiency) isn't the same as saying that amp is sounding its best.

The problem you run into with higher powered tube amps (unless its an OTL) is the output transformer limits bandwidth and introduces distortion. So you run into a practical engineering problem that is best solved by simply using a speaker that's easier to drive, thus allowing for a lower powered amp that can have wider bandwidth and lower distortion.

I don't seen an upside to hard-to-drive speakers- they don't seem to have greater resolution or bandwidth... If you want to get the most out of your amplifier investment, its always going to be best served by a speaker that's not low impedance and has higher efficiency.

All correct, but a transistor amplifier with a low internal impedance has more control over the loudspeaker (damping factor and usually high negative feedback) even under heavy load. It simply copes better with fluctuating low impedance of the loudspeaker. There are RCL correction* elements for a tube amplifier. Well-designed loudspeakers have something like this to offer more choice in amplifiers.

*You lose some efficiency, but then a tube amplifier can work cleanly.

 

[Atmasphere]

Ok so you consider 90+ db for sensitivity as high with 97-100db as being very high ?

So low would be below 87db ..?

Its been my experience that very high sensitivity Speakers tend to be unbalanced for sound unless very complex and with multi driver. Usually by the time you get a proper balance top to bottom you have lost a lot of the sensitivity. Many today make it up by running powered subs if not it becomes pretty complex to make it all work.

With multiple drivers you lose efficiency as the impedance is gonna drop but sensitivity stays up , this is very observable in the wild , as you can see many SOTA loudspeaker attempts are with multiple drivers and complex as you half thd every-time you double driver area ..!

Since its not 1940 and high powered amps both toobs and SS are available i would say an 87db loudspeaker of multiple drivers is acceptable but not in a 2 driver 2 way design which will have high thd attempting to reach live levels ..!

Regards

Since all measurement these days is done by sensitivity rather than efficiency, you have to include impedance. 87dB at 4 Ohms is low efficiency. 87dB and 8 Ohms is moderate and that continues to about 96dB where i consider that to be higher efficiency if 8 Ohms as well.

The impedance of the speaker matters to the distortion of any amplifier. All amps make less distortion driving higher impedances. In addition, speaker cables become far less critical! Finally, thermal compression is a thing and you can't solve it by throwing more power at it. So low sensitivity/efficiency speakers also with low impedance are always at a disadvantage if the best audio reproduction is your goal.

Put another way, if you could keep things equal and make a 4 Ohm speaker 8 Ohms, it would immediately sound better because the amp was making less distortion. You might think the difference in distortion is too little to be heard, but its almost entirely higher orders so it is audible even in trace amounts. Distortion obscures detail and alters tonality; if your goal is the best reproduction, your amplifier dollar investment is best served by a higher impedance speaker that is easy to drive.

All correct, but a transistor amplifier with a low internal impedance has more control over the loudspeaker (damping factor and usually high negative feedback) even under heavy load. It simply copes better with fluctuating low impedance of the loudspeaker. There are RCL correction* elements for a tube amplifier. Well-designed loudspeakers have something like this to offer more choice in amplifiers.

*You lose some efficiency, but then a tube amplifier can work cleanly.

The fact is though that a solid state amp will still make higher distortion with low impedance loads and the speaker cable will be far more critical to the best presentation- keeping the cable very short will be mandatory if you want the best bass impact, definition of vocals and the like.

 

[Audiohertz2]

Since all measurement these days is done by sensitivity rather than efficiency, you have to include impedance. 87dB at 4 Ohms is low efficiency. 87dB and 8 Ohms is moderate and that continues to about 96dB where i consider that to be higher efficiency if 8 Ohms as well.

The impedance of the speaker matters to the distortion of any amplifier. All amps make less distortion driving higher impedances. In addition, speaker cables become far less critical! Finally, thermal compression is a thing and you can't solve it by throwing more power at it. So low sensitivity/efficiency speakers also with low impedance are always at a disadvantage if the best audio reproduction is your goal.

Put another way, if you could keep things equal and make a 4 Ohm speaker 8 Ohms, it would immediately sound better because the amp was making less distortion. You might think the difference in distortion is too little to be heard, but its almost entirely higher orders so it is audible even in trace amounts. Distortion obscures detail and alters tonality; if your goal is the best reproduction, your amplifier dollar investment is best served by a higher impedance speaker that is easy to drive.


The fact is though that a solid state amp will still make higher distortion with low impedance loads and the speaker cable will be far more critical to the best presentation- keeping the cable very short will be mandatory if you want the best bass impact, definition of vocals and the like.

Err, yes and no ,

Since most domestic listening is done at 1watt RMS and less making all of the above or most mute with modern drivers and coil designs. The 8/4 ohm straw man doctrine really mostly applies to tube amplifiers especially OTL’s which are distortion camps below 8 ohms , as they are typically higher in distortion than their SS counterparts even when driving low Z speakers . An SS amps with appropriate bias and feedback will have low distortion numbers into 2 ohm’s , yes the bad feedback, that great devil driven by those with an agenda about how bad it is to use feedback are now designing class D feedback factories

The first watt or 1 watt rms delivery is very important with tubes as this is the region where they compete with low distortion numbers making them very listenable, above this their distortion output rises exponentially unless with sufficient feedback , unlike SS which usually don't do so until typically 33% of output ..

The benign soft clipping of toobs is what makes them more pleasant to most than their SS counterpart running into clipping..

BTW Loudspeakers Sensitivity and efficiency are two different entities and measured differently ..!

2.83V is the current adapted industry standard @1M for comparative purposes..

At 1-2 watt RMS you will need 200 watts to avoid clipping on high DR recording
( as verified by Bob Cordell ) this very fact is why most under powered SS system owners run to toobs to avoid the dreaded SS tang at clipping .

 

[Atmasphere]

Err, yes and no ,

Since most domestic listening is done at 1watt RMS and less making all of the above or most mute with modern drivers and coil designs. The 8/4 ohm straw man doctrine really mostly applies to tube amplifiers especially OTL’s which are distortion camps below 8 ohms , as they are typically higher in distortion than their SS counterparts even when driving low Z speakers . An SS amps with appropriate bias and feedback will have low distortion numbers into 2 ohm’s , yes the bad feedback, that great devil driven by those with an agenda about how bad it is to use feedback are now designing class D feedback factories

The first watt or 1 watt rms delivery is very important with tubes as this is the region where they compete with low distortion numbers making them very listenable, above this their distortion output rises exponentially unless with sufficient feedback , unlike SS which usually don't do so until typically 33% of output ..

The benign soft clipping of toobs is what makes them more pleasant to most than their SS counterpart running into clipping..

BTW Loudspeakers Sensitivity and efficiency are two different entities and measured differently ..!

2.83V is the current adapted industry standard @1M for comparative purposes..

At 1-2 watt RMS you will need 200 watts to avoid clipping on high DR recording
( as verified by Bob Cordell ) this very fact is why most under powered SS system owners run to toobs to avoid the dreaded SS tang at clipping .

The 1st Watt of any amp is the most important; tube or solid state. Your comment about OTLs only applies to lower powered OTLs. Any tube amp driving 4 Ohms will be higher distortion; that's not limited to OTLs (or for that matter, tubes in general). All amps will have higher distortion driving 4 Ohms as well. Its both measurable and audible. There's simply no good reason for 4 Ohms if high quality audio reproduction is your goal. You do get a weak (IMO) reason for 4 Ohms if you have solid state as you can get 3dB (not much, that's the weak bit) more sound pressure (depending on the amp; that may not be to full power).

I know about sensitivity vs efficiency... the industry standard is 2.83V because into 8 Ohms that's 1 Watt. So 2 Watts into 4 Ohms; since tubes do not double power as impedance is halved efficiency is a far more convenient spec to use with them. You can get a tube amp to behave as a Voltage source, but doing so means it cuts power in half as impedance is doubled rather than doubling it as impedance is halved. When the relatively inexpensive power of solid state showed up sensitivity became the spec to use instead. This transition took about 20 years (1960 to about 1980).

Most solid state amps have that 'dreaded SS tang' (harshness and brightness) at power levels well below clipping. In a nutshell its caused by how feedback is applied, plus the lack of gain bandwidth product to support the feedback at all frequencies. This is what has kept tubes in business the last 60 years since being declared 'obsolete' (back when germanium transistors were the only game in town ...). Lots of people don't like harshness and brightness. You don't have to know any engineering to understand the simple economic fact that tubes are still around, which tells us that solid state didn't bring home the bacon.

Class D offers an easy solution to both feedback problems; putting tube amps on shaky ground.

 

[sparkie]

Hi Guys

Please carry on with your interesting discussion here in this dedicated thread and NOT in the SET-owners thread.

Knock yourselves out.

All I see is just talk so far. I'll post circuit topologies. I'll start with common ones first.

Lets talk about the grounded emitter/cathode/source amps (ie. common emitter, common cathode, common source)

Signal is unbalanced connected from signal input to ground. Amplifying device is connected between voltage supply and gnd, so 100% of the power supply noise is added to the signal. Output signal is 180 degrees out of phase with the input signal. low/mid input impedance to high output impedance. Biased class A 50% so higher idle noise floor and wasted heat. This is also called an inverting voltage amplifier.