Why do tube amps roll off high frequencies?
- Thread starter MadFloyd
- Start date
I'm no expert but from what I've read and heard the output transformer is responsible for the roll off. I'm sure it depends on the bandwidth of the particular transformer though. Then there are OTL designs that don't use an output transformer. Of the OTL's I've heard I have not notice rolled off highs.
I've also noticed rolled off highs on the McIntosh solid state amps that use an output transformer. They call it an autoformer I'm not sure what the distinction is?
Maybe Atmasphere will chime in. I'm sure he can explain it better.
I've also noticed rolled off highs on the McIntosh solid state amps that use an output transformer. They call it an autoformer I'm not sure what the distinction is?
Maybe Atmasphere will chime in. I'm sure he can explain it better.
Look up inductance and transformers on Wikipedia or some such. Inductors pass DC and low-frequency but block high frequencies (capacitors do the opposite). Transformers are coupled inductors; two or more inductors that work together. There are transformers that work into the microwave range, but audio transformers tend to be large to handle the power, and have lots of windings to convert from the high-impedance tube plates to low-impedance speakers. That all reduces their bandwidth. An autoformer is a tapped transformer that is often configured as a single continuous winding instead of isolated coupled coils. It is also used to provide impedance transformation.
I'm surprised you hear a roll-off on Mac solid state gear. They are typically down 3dB at 100 KHz and completely flat well beyond 20 KHz. That also goes for their tube amps.
Well to be honest that was my impression. Can't say I ever measured the FR. Their tube amps seemed even more rolled off to me. I'm guessing it's how they are reacting to a particular speaker load. I can't say I've ever heard/read anyone claiming Mac amps are bright.
As I understand it too, there's a relationship between power rating, the construction of the output transformer and bandwidth eg. the bigger power tube amplifiers have a wider bandwidth than lower powered tube amplifiers.
Myles-I can’t understand why a bigger power amp (and I’m assuming you mean higher power) would have a higher bandwidth than a lower powered tube amp. I’m thinking that a high-power high-bandwidth output transformer would be much harder to wind than a low power high-bandwidth output transformer. I believe that it was David Manley that talked about the “magnetic window” that output transformers have to be designed to fit in and wound in. In reality, the properties for an output transformer designed to reproduce deep bass are different than an output transformer designed to reproduce the upper frequencies and thus the “window” so you don’t throw one or the other out of it with a conventional output transformer. We have ample proof backed up by measurements that very wide-band output transformers do exist.
Getting back to the OP’s question, it was always my understanding that the reason high frequencies of tube amps aren’t ruler flat is the output impedance of tube amps is high and interacts with the impedance of loudspeakers which depending on the impedance curve of the loudspeaker can cause deviations in the FR which can be heard.
Getting back to the OP’s question, it was always my understanding that the reason high frequencies of tube amps aren’t ruler flat is the output impedance of tube amps is high and interacts with the impedance of loudspeakers which depending on the impedance curve of the loudspeaker can cause deviations in the FR which can be heard.
Usually high power amplifiers have many power tubes wired in parallel and this decreases the output impedance of the power stage, decreasing the primary to secondary turns ratio of the transformer. This can extend bandwidth. But there are many other parameters is transformer design, including the core, the area allowed for the windings and their segmentation. The Audio Research REF150 specifications say : Power bandwidth (-3dB points) 5Hz to 80kHz; Frequency response (-3dB points at 1 watt) 0.5Hz to 120 kHz. Why should we want an higher bandwidth?
^^^ +1.
I think the BW issue with tube power amps ws much more significant many years ago, and I doubt it was true for all amps even then. Transformer technology has evolved, along with (IMO w.r.t. BW) more minor circuit improvements through the years.
I still think the biggest difference between tube and SS amps (OK, most of them) is the output impedance.
I think the BW issue with tube power amps ws much more significant many years ago, and I doubt it was true for all amps even then. Transformer technology has evolved, along with (IMO w.r.t. BW) more minor circuit improvements through the years.
I still think the biggest difference between tube and SS amps (OK, most of them) is the output impedance.
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |