|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
66.115.181.216
In Reply to: RE: What makes a driver High Efficiency? posted by morricab on May 30, 2023 at 05:45:30
Exactly! Direct radiators (K<<1), to have flat response have an acceleration controlled response.
I think the hard part to get is how the motor works, it's a small series inductance and a DC resistance in series with the loss less motor BL, that BL can also be expressed as Newtons of force per Amp of current but it is also a generator who's voltage proportional to BL and the motion velocity and called back EMF. The higher the BL, the greater the back EMF, they are hand in hand.
Follow Ups:
Doesn't this high BL woofer then, as a generator of back EMF benefit from a class D amp like a Hypex with strong feedback and relative load independence down to 2 ohms? IOW, an amp with strong control over the motor vs. a wimpy Single Ended Tube amp with a few watts on a good day?
This is complicated as there is more involved than just motor strength.
What you have is a mass and spring (moving mass and driver plus box compliance or spring) Those two reactance's have a resonance where the two reactance's are equal but opposite and cancel out. AS you approach resonance, you see the parallel L and C's effects cancel out and at Fb resonance the only load you see in the impedance peak is the Qm mechanical losses and a tiny amount of radiation resistance (big ohms).
The motor is connected to this moving stuff and so is the radiator.
The bigger and more massive the moving system, the larger the BL has to be to have the same amount of "control" on the moving system.
The amplifier (in modern times) is assumed to be a low impedance voltage source. It used to be a selling point that the amplifier's output impedance was tiny with respect to the load and the ratio expressed as "damping factor".
In loudspeaker land, the driver has a DC resistance (Rdc) in series with the motor so the amount of "control" is limited by that Rdc (BL and Rdc controlling motor part of the driver Qe). The amplifier appears to be a small resistance in series also but if the load were 4 ohms and the DF 100, the amplifier looks like 4/100 Ohms. Around 20 or 30 DF is where making it larger has very little/ less and less effect.
A series resistance that is significant (say an Ohm) begins to modulate the normal frequency response, where ever there is a low impedance, there is a new small dip or depression added and so the shape of the response curve is altered by the addition of a version of the shape of the impedance curve.
Morricab brings up an obscure but real issue.
The assumption is when you add negative feed back to make a closed loop, you lower the output impedance accordingly as well as lower the distortion. That is true to the degree there is open loop gain left to do those things at what ever frequency it is.
If one connects a capacitance across an amplifiers output and drives it with an impossible to follow signal like a square wave, theoretically what one would see is the square wave rounded off in an R/C low pass filter (amplifier output R and load C).
In reality, normally what one gets is a ringing square wave showing not a simple but a more complex and reactive source, showing an amplifier over shooting and damping often with a definable Q.
In my own fooling around, what found was if one made an amplifier with a low output Z without -feed back, it did behave like an R/C.
I think what one can say is in a closed loop reactive loads can use up the open loop gain that normally gives the good measurements.
How big is this stuff?? I don't know but at least with an amplifier, it's possible to set up a null test where one can listen to the difference between the input signal and output signal, much harder with loudspeakers but the generation loss test works.
Tom
Matti Otala showed that back EMF has more impact on the sound quality of high feedback amps as that back EMF is pushed through the feedback loop and re-amplified. A SET without feedback will just dissipate the back EMF as heat in output transformer as there is no backwards path through the amp without a feedback loop. A high sensitivity driver with a low Qts doesn't need or want a high damping factor amp...- my 99dB Supravox drivers with a 1.9T Alnico magnet and 7 gram MMS and Qts of 0.21 love SET above all else.
Hypex Class D is a different animal opposite of an SET amp. Not saying it's better, and I know not why it takes full, tight, control of all the voice coils regardless of the size of Magnets.
All I have practiced since 1976, is what Saul Marantz told me when I talked to him at a Detroit Audio Show. Solid State on the Bass, Tubes on the mids and highs. He was President of Dahlquist then and sold his name to Superscope for a tidy sum, I imagine.
I have found it is better for coherence to have the same type of amp on each driver. You can hear that different amp topologies are at play when done like you suggest.
> You can hear that different amp topologies
You shouldn't, since where they differ is mainly in the midrange, where subs don't operate. SS where more power was needed and tubes where it wasn't was a common approach in the early days of SS. Things have changed greatly since then. I doubt anyone finds a Bryston or Powersoft to be wanting compared to tubes today, other than where confirmation bias determines the result.
Oh but you can hear the difference easily. I was at a friend's yesterday and he has two amps, one SS (Plinius SA103, which is a Class A PP SS amp) and one SET (Ayon Helios, which has 2 x 6C33C tubes per channel in Parallel SET). We listened to both digital (Ayon Stratos) and Analog (Transrotor + Transfiguration Proteus and Chinook phono).
We started with the Plinius and the sound was, not too bad. Decent tone, ok dimensionality (both soundstage and imaging) and tight powerful bass. Two things stood out; 1) the bass sometimes seemed a bit dominant in the overall sound signature and 2) vocals were quite present but the rest of the instruments in the mix "sat back" a bit.
Then we switched to the Ayon Helios...the difference was not subtle. Much more accurate tone color of instruments, better dynamics, instruments that were there but not recognized popped out. Dimensionality in all directions and imaging 3d improved drastically. Finally, and importantly, the bass was now less prominent but seamlessly integrated into the whole, where before it stuck out and called attention to itself.
I have heard this effect in the past as well. I once took a KR Audio VA350i to a guy running a Gamut D200 amp. With the Gamut the speakers sounded like a collection of drivers, each kind of doing it's thing. With the KR it sounded like they were all working together for the whole sound.
So, in essence I think you are completely wrong to assume a Bryston or Powersoft will be difficult to tell from a good tube amp...nothing could be further from the truth...it is blindingly obvious.
Read my post again: "where they differ is mainly in the midrange, where subs don't operate."
As for a Bryston or Powersoft not equaling or bettering a good tube amp, show us the evidence. That means measurements and...shudder,shudder,shudder..double blind testing.
99% of Subjective words without data to back them up deserve the Paul Klipsch Yellow Button.
"Confirmation Bias" and no AB/X test box (I was there for the unveiling of that rarely utilized device when I was an AES member at a local chapter meeting in Michigan).
Also no measurements to back up comments from 90% of people on the web (I could be understating it in their favor).
Knowing I am with fellow cheapskates here (a distinguished group I might add)..................who's gonna spend a Kilobuck on a multitap autoformer with a rotary control for only ONE friggin driver????
Go active crosovers with 2, 4, 6, 8, 10 outputs, since we are talking high efficiency, you only need TI 3255 chip amps and derivatives to do the job. They are available RCA or Balanced XLR instead of buy 1 Horsepower Class A/AB amps like a friend just did Times 2 because they are monoblocs.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: