John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Anatoliy, you might be surprised. No need of testing 2W SE tube amps only or overload protection function. Try it and you might be surprised.

So, what's the point, in such case, in frequency response curve of the amp loaded on some kind of dummy load that represents some standardized model of some standard speaker? ;)

To show that the amp has very low output impedance? Is it good, or is it bad? What about amps that were made especially with high dynamic resistance, so called Current Drive Amps? Do they sound worse than amps with negative output resistance?
 
So, what's the point, in such case, in frequency response curve of the amp loaded on some kind of dummy load that represents some standardized model of some standard speaker? ;)

To show that the amp has very low output impedance? Is it good, or is it bad? What about amps that were made especially with high dynamic resistance, so called Current Drive Amps? Do they sound worse than amps with negative output resistance?

When you get into specialized topography such as current output amplifiers then you then get very loudspeaker design specific results. (But you knew that.) But most of the amplifiers I have looked at have quite interesting differences in distortion results both at low level and at clipping when run into a real loudspeaker load.

These days I use a dual 18" subwoofer for the load because it was leftover from an arena system. (They only had enough space for 30 of them not the 32 the system was designed to use.)

Even at very low levels where I would not expect any real strain from the load differences show up, there often is a difference. Now is it due to back EMF into a feedback loop, or just due to the V/I phase shift, or some other cause is an interesting consideration and points to what may be weaknesses in the design.

As you may be aware I am starting to look at noise modulation of circuitry and that may be an issue with different loads.

Now at high levels collapsing inductors in loads and other issues pop up. Although most of your work assumes that the load crossover inductors don't saturate, the behavior during saturation makes a very big difference in the perceived sound quality of many pro amplifiers.

This is much less of a problem in amplifiers just used for reproduction of recorded music with purpose built loudspeakers. But unless you are controlling all the users of your products you may get complaints or even a bad reputation among some users.
 
You have a good point, Wavebourn. If anyone wants to check, my JC-1, because of its unique output stage, heavy bias, use of negative feedback, and NO output coil has one of lowest output impedances over audio frequency in the industry. Still, it does not mean that much. Other factors count more, and I use added .3-.5 ohms in series with my amp to drive my WATT loudspeakers. I prefer it that way.
John A, I meant to challenge the inference that we were buying our reviews through placing ads in either 'TAS' or 'Stereophile' . Some people just imply things that have no justification.
 
But most of the amplifiers I have looked at have quite interesting differences in distortion results both at low level and at clipping when run into a real loudspeaker load.

Yes, the keyword is distortion results. I totally agree here. But the point was about frequency response curve. If I want to check frequency response curve of my amp most revealing would be the curve taken on plain resistive load, even 1/4'Th of nominal load resistance. If I want to see frequency response curve with loudspeakers I would add measurement microphone into the equation, but frequency response of the amp loaded on some kind of dummy load is meaningless.

If to compare amps by frequency response curve taken as function of output voltage to input voltage when loaded on reactive load, of course Bob's amp will look always better than Nelson's amp in such reviews because they have different output resistances by design.
 
You have a good point, Wavebourn. If anyone wants to check, my JC-1, because of its unique output stage, heavy bias, use of negative feedback, and NO output coil has one of lowest output impedances over audio frequency in the industry. Still, it does not mean that much. Other factors count more, and I use added .3-.5 ohms in series with my amp to drive my WATT loudspeakers. I prefer it that way.

That means, you need to add a switch: "For Stereophile Review / For Listening Preference". ;)
 
Last edited:
So, what's the point, in such case, in frequency response curve of the amp loaded on some kind of dummy load that represents some standardized model of some standard speaker? ;)

It gives an easily grasped picture of how and if the amplifier's output impedance changes with frequency and whether the inevitable response changes with a typical speaker will be large enough to be audible or not. You can see the circuit diagram and the complex impedance of my standard simulated speaker at Real-Life Measurements Page 2 | Stereophile.com .

John Atkinson
Editor, Stereophile
 
Peter Moncrief said that the role of the reviewer is to help the reader
differentiate between products. There is no such thing as a "best"
amplifier, only ones which best fit a particular customer's needs.

:cool:

Amen to that thought, Nelson, with the proviso that there are some designs that would never fit any customer's needs!

John Atkinson
Editor, Stereophile
 
It gives an easily grasped picture of how and if the amplifier's output impedance changes with frequency and whether the inevitable response changes with a typical speaker will be large enough to be audible or not. You can see the circuit diagram and the complex impedance of my standard simulated speaker at Real-Life Measurements Page 2 | Stereophile.com .

John;
I respectfully disagree with you here.
If the amp has output impedance that changes with frequency it is much more useful to compare result with straight line than with some kind of standardized curve that also depends on output resistance of the amp, even if it is absolutely frequency independent. Also, dips and peaks on impedance curve of loudspeakers are not directly translated into dips and peaks on sound pressure curve.

During one BAF in San Francisco there were 2 listening rooms, one in front of another. In both rooms were installed speakers and amplifiers. In one room were flat panels, like doors, with 1 full-range driver in each, and amps that drove them had relatively high outpt resistances.
In another room speakers were made by professional carpenters, they were complex dipoles with crossovers, amplifiers had very low output resistance.

In your tests amps in room 1 would look much worse than amps in room 2. But music in both rooms sounded nice. I personally prefered reproduction in room number one, because closing eyes I could hear real instruments in the room, but in the room 2 I always heard nice sounding system.
 
Last edited:
Member
Joined 2002
Paid Member
I like Johns measurements consistency is important. One of our techs was testing an Int150 and I couldn't remember a curve when my machine was backing up. Well a quick look at Johns review measurements and all was well.
As an old prof said measure! Measure! Measure! You had to bang the table with a red nut driver while saying it.
 
It gives an easily grasped picture of how and if the amplifier's output impedance changes with frequency and whether the inevitable response changes with a typical speaker will be large enough to be audible or not.

Not only amplitude frequency response, but distortion as well. Sometimes we may have an amplifier with rugged multipair output stage, no output coil, almost no influence of load impedance to amplitude freq. response, and even if this is fulfilled the distortion vs. frequency may vary considerably with dummy load. It may reflect the driver (transistors) circuit design.
 
As an old prof said measure! Measure! Measure! You had to bang the table with a red nut driver while saying it.

Right! Measure, measure, measure; but measure the things that give you an information, what to do in order to improve things! ;)

Here is an example: the company measured amplifiers, and was doing well. It hired professional managers in order to do better. Managers divided profit by cost of assets and said that according to measurements the company needs to eliminate assets. Now the company has no assets, it's financial parameter measures excellent, but can't measure amplifiers because have no equipment. It has better performances, but has no future, because all that left from it is a brand name only...
 
John A, I meant to challenge the inference that we were buying our reviews through placing ads in either 'TAS' or 'Stereophile' . Some people just imply things that have no justification.

Indeed, sadly, this is so. However in an open forum, there is no way to stop people posting their thoughts, regardless of how erroneous their thoughts may be. Also, some people wouldn't let themselves be confused with facts.
 
John; Nelson's amps sound very nice, because instead of run for lower output resistance to satisfy your readers he carefully selects optimal load lines, to satisfy demanding listeners.

Well, Nelson's amplifiers both sound great and measure well. As for the way they sound, I know first hand, for my power amplifier is a Nelson's product.
I assume there is much more to his designs than only selecting optimal load lines.
 
Well, Nelson's amplifiers both sound great and measure well. As for the way they sound, I know first hand, for my power amplifier is a Nelson's product.
I assume there is much more to his designs than only selecting optimal load lines.

I suppose, he uses proper power cable, right Joshua? ;)
Edit: But it does not matter, actually: I suppose Nelson prefers satisfaction of listeners to satisfaction of readers. :)
 
Last edited:
Status
Not open for further replies.