[G Georgopoulos]

You obviously haven't heard Nimbus, Derecho, Arcus, or Perfect Storm George, or you wouldn't have made that last statement.   

We had Nimbus running on a 3.5W 2A3 SET and a 35W 805AT SET at RMAF a couple years ago and it filled a very large conference room without a problem.  And at Axpona 2014, we had Derecho running on 18W SET and were easily hitting 100dB in the room.

https://www.youtube.com/watch?v=li-NBI6kP2I

Your formula is of the form R || R/R.  R/R has no units, so I'm not sure it's completely correct.

Damping factor is simply the ratio of Zout and RL and it doesn't matter what class you're running.

pete i'm sorry i mentioned vapor speakers,what you do with them is your business

i'm out of this thread,you dont get my point

kind regards

[Pete Schumacher]

pete i'm sorry i mentioned vapor speakers,what you do with them is your business

i'm out of this thread,you dont get my point

kind regards

Please explain your point then George.  No need to be sorry.

[G Georgopoulos]

ok Pete
i'll try my best
this only occurs with class-a not class-b/ab amps
the Zout is constant
because in class-a both semiconductors are always on
there is always a dynamic resistance in parrallel with Rload
this resistance is found by Vout / Ibleed
this has the tendancy to load the Zout and reduce the Vout abit
and hence the dynamic impact
without this resistance a class-b/ab amplifier
is loaded only by Rload so you are correct and the damping factor
is df= Rload/Zout
to put it another way if you load a certain source resistance with more
current there is gona be less voltage across the load
and less current into the load thus less dynamic impact
this is what happens with class-a there is this shunt dynamic resistance
that loads Zout and there is less voltage across Rload

kind regards

[Steve]

ok Pete
i'll try my best
there is always a dynamic resistance in parrallel with Rload
this resistance is found by Vout / Ibleed
this has the tendancy to load the Zout and reduce the Vout abit
and hence the dynamic impact
without this resistance a class-b/ab amplifierthis only occurs with class-a not class-b/ab amps
the Zout is constant
because in class-a both semiconductors are always on

is loaded only by Rload so you are correct and the damping factor
is df= Rload/Zout
to put it another way if you load a certain source resistance with more
current there is gona be less voltage across the load
and less current into the load thus less dynamic impact
this is what happens with class-a there is this shunt dynamic resistance
that loads Zout and there is less voltage across Rload

kind regards

I do not mean to but in, but this statement

this only occurs with class-a not class-b/ab amps
the Zout is constant because in class-a both semiconductors are always on

is not true. AB can easily have a constant or very near constant output impedance (Z). However, as the output
approaches class B status, the output Z does vary with signal level. One needs to observe the plate
resistance vs the plate current or grid voltage.

There are a variety of factors that affect sound, but Pete's graph clearly demonstrates that the damping factor is one of
those factors, and does affect sonics.

Amplifier current is not really a factor until the speaker driver, usually the woofer, becomes capacitive in nature, at very low frequencies.
Above such, it is resistive and inductive.

Cheers
Steve

[Folsom]

But don't forget crossovers, Steve.

[Steve]

But don't forget crossovers, Steve.

First, I should have been clearer, I meant cone type speakers in my previous post. Next, it will depend upon some factors, and what we mean by high current.

For two way cone speaker, the inductor is basically in series with the woofer (6db xover) and caps (12db/octave xover), which limits the current.

The full range driver used for midrange/tweeter, the cap is in series with the driver (6db) and also inductor (12db). (two way design).
Current is again limited.

In three way design, all capacitors have some sort of inductance and/or VC inductance/resistance in series.

Here is a link to lots of xover designs to check out. I did not exhaustively search, but saw no huge capacitance shunting the amp's output to ground, requiring huge currents. (Speaker cables could have some small capacitance.)

http://www.bing.com/images/search?q=3+way+speaker+crossover+diagram&id=64F52EFB8203C15F73FDB3CE09A74E39E250B30C&FORM=IQFRBA

Again, I guess it depends upon what is meant by high current. At very low frequencies, the inductive reactance does become quite low, so more current due to cap to ground (12db/octave xover). With 6db/octave xover, the driver becomes capacitive.

Of course there is always the direct connection to the driver method when using an active xover.

Cheers
Steve

ps. I hope Ryan and Pete do not mind me posting even though I am retired.

[Folsom]

I'm not retired, so I'll just egg you on to do the goodsplaining, Steve.

[G Georgopoulos]

I do not mean to but in, but this statement is not true. AB can easily have a constant or very near constant output impedance (Z). However, as the output
approaches class B status, the output Z does vary with signal level. One needs to observe the plate
resistance vs the plate current or grid voltage.



Cheers
Steve

ha...
to put it another way if frequency response is flat Zout is constant
steve you always disagree with me i dont know why?

kind regards Steve ...

[G Georgopoulos]

when it comes to tubes amps
two things matter most

1)good power supply
2)good output transformer

for dynamic impact!

these are not secret technology
but i'll mention here once again

nfb does little for tubes

lets start with transformers
the magnet wire thickness pri. sec.
the transformer core size
the winding method

now power supply
the transformer va rating
the filter caps size
the power transformer voltage

all these enhance the Zout capacity of the tube amp

feel free to disagree

That's my understanding of this,please correct me if wrong

[Pete Schumacher]

when it comes to tubes amps
two things matter most

1)good power supply
2)good output transformer

for dynamic impact!

these are not secret technology
but i'll mention here once again

nfb does little for tubes

lets start with transformers
the magnet wire thickness pri. sec.
the transformer core size
the winding method

now power supply
the transformer va rating
the filter caps size
the power transformer voltage

all these enhance the Zout capacity of the tube amp

feel free to disagree

That's my understanding of this,please correct me if wrong

negative feedback, judiciously applied, works even for tube amps.  In fact, most tube amps employ local negative feedback in each gain stage.  Overall negative feedback will stabilize output impedance over the frequency range.

Good power supply applies to all amplifiers.

Good output transformer is also essential, especially if no overall feedback is used.

[G Georgopoulos]

ahh...you're a smart guy Pete,your points i agree with 100%...

kind regards
George

[Steve]

ha...
to put it another way if frequency response is flat Zout is constant

It will depend upon what conditions you are discussing, and how much Zout changes G. I do have a question, comment or two though, if you do not mind. I mean no harm, maybe being a designer for some decades I have to consider all the options possible. I don't know if you are referring to the whole amp or just the output stage.

So my first question is what happens if the input stage of an amplifier (uses no global negative feedback) and the input tube is, say a 12AX7 that drops the FR at some frequency X? The output stage itself can have a constant output impedance (Z) over the entire frequency range (at a given signal level), while the frequency response of the amp is not flat. I want to assume you meant just the output stage FR.

If you are referring to just the output stage being flat response, however, the output Z will be dependent upon the over all design, SET or PP, and operation such as A, AB, B, and signal level itself. The FR and output Z will vary with signal level, so there can be some little relationship, although not the sole relationship by any means. (Hint for newbies: check the OPT FR at 1 watt and full power ratings, and tube plate resistance vs signal level.)

Hope this helps.
Steve

[G Georgopoulos]

So my first question is what happens if the input stage of an amplifier (uses no global negative feedback) and the input tube is, say a 12AX7 that drops the FR at some frequency X? The output stage itself can have a constant output impedance (Z) over the entire frequency range (at a given signal level), while the frequency response of the amp is not flat. I want to assume you meant just the output stage FR.



Hope this helps.
Steve

that's impossible!!!,the 12ax7 plate resistor is say constant,why? because is just
 a resistor (zout) and this resistor feeds another resistor (of the next stage imput z)
unless it is loaded by the next stage imput z "which is impossible",there is no
drop at any frequency...

so there

again,global nfb does little,the main reason is with tubes having less voltage gains
compared say to solid state,you can get away with local nfb however
again as Pete mentioned global nfb judiciously applied could be of some benefit.

kind regards


 

[Steve]

that's impossible!!!,the 12ax7 plate resistor is say constant,why? because is just
 a resistor (zout) and this resistor feeds another resistor (of the next stage imput z)
unless it is loaded by the next stage imput z "which is impossible",there is no
drop at any frequency... 

I think we just have a mis understanding. The frequency response of any stage is limited since resistance, capacitance, inductance are at play.

In our amplifier stage with 12AX7, the FR drops because of plate resistance of the stage combines with stray capacitance around the tube, output capacitance of said tube, interelectrode capacitance of the following tube, and miller capacitance presented. Otherwise the frequency response would theoretically be infinity using any tube, including a 12AX7. In fact, it is quite possible for an open loop 12AX7 stage to have an FR of -1db at 25khz.
 
So without any GNF, while the output stage of the amplifier may have flat FR and constant damping factor with constant signal level, the 12AX7 first stage will have limited FR. The point is the lack of FR in the first stage does not cause the output Z of the amplifier to vary. I hope this is a clearer explanation.

again,global nfb does little,the main reason is with tubes having less voltage gains
compared say to solid state,you can get away with local nfb however
again as Pete mentioned global nfb judiciously applied could be of some benefit.

kind regards

Concerning global negative feedback, the FR is extended, the output impedance (Z) is lowered, the overall gain is lowered. There is also a time delay from input to output, so the global negative feedback signal arrives at the input some X time after the original input signal.

There are two types of local negative feedback.

1) Local current feedback, which lowers the gain of that stage, increases the output impedance (Z) of that stage, and lowers the FR of that stage.

2) Local voltage feedback, which lowers the gain of that stage, decreases the output Z of that stage, and increases the FR of that stage.

With GNF and both types of local negative feedback, one needs an excess of open loop gain, whether tube or SS, as the overall gain of the amplifier
is reduced.

I think we may agree, but explanations can be difficult to pass along clearly. This is my last posting here. If one wishes more information, please PM me.

Cheers
Steve

[G Georgopoulos]

yes i agree with you on those points,i didnt mention them, i just implied, for instance fr the audio band up to 20 khz , tube charactiristics etc,tried to keep it simple!

great write Steve...

kind regards

[rebbi]

Pete, et. al.,
Sort of relevant to this thread, I guess I'll shill shamelessly for my audio blog, where I'm in the midst of a survey of SET-friendly speakers. Suggestions for additions to the list are welcome.
Steve's Audio Blog

[Pete Schumacher]

Pete, et. al.,
Sort of relevant to this thread, I guess I'll shill shamelessly for my audio blog, where I'm in the midst of a survey of SET-friendly speakers. Suggestions for additions to the list are welcome.
Steve's Audio Blog

Depending on how much SET power is available . . .

We had our Derecho model running on 18W SET power at Axpona in 2014.
The Nimbus model was running on 3.5W and 35W SET power at RMAF in 2013 and 35W at Newport in 2013

Perfect Storm and Arcus are also in that category due to rated sensitivity in the 93dB range.

Even Joule ran exceedingly well on the 35W SET power we gave it with its 90dB sensitivity.

[rebbi]

Depending on how much SET power is available . . .

We had our Derecho model running on 18W SET power at Axpona in 2014.
The Nimbus model was running on 3.5W and 35W SET power at RMAF in 2013 and 35W at Newport in 2013

Perfect Storm and Arcus are also in that category due to rated sensitivity in the 93dB range.

Even Joule ran exceedingly well on the 35W SET power we gave it with its 90dB sensitivity.

Hey, Pete!

I was thinking of your garden variety 300B, 8 watt SET, like my ANK Kit 1. But it's pretty cool that you were able to successfully power the Nimbus on 3.5 watts.

I'll be happy to append Vapor to the list in my next post, and/or feel free to respond on the blog.

[undertow]

Not sure exactly how feedback correlates to the results of low impedance speakers directly via the Claytons.

However, I did own these amps, and their whole claim to fame at the time was virtually a "No negative feedback" design. So not  to sure, but honestly it was probably lower than even most tube amps in this case. However, the Claytons I know were of the very few in Class A that can truly drive a 2 ohm load which is helpful, and probably more important in the case of driving standard woofers that will drop into the 2 or 3 ohm range pretty easily especially in modern designs using 2 and 3 way crossovers.

http://positive-feedback.com/Issue40/clayton_m300.htm