We are constantly told how bad our SE tube amplifier are when it comes to Damping Factor, some people don't realise or choose to ignore the fact that the actual speakers DC resistance tends to be dominant in most cases.
I tend to ignore all claims of DF importance. But that's just my limited experience and my passion for flea power tube SE amps.
One of many examples. Few speakers are essentially a resistive load.
VAC Statement 452 iQ Musicbloc mono/stereo power amplifier Measurements | Stereophile.com
VAC Statement 452 iQ Musicbloc mono/stereo power amplifier Measurements | Stereophile.com
Is damping factor overhyped to the point where some people are promising more performance than what the laws of physics allow? Yes. Is the idea of a very high damping factor intrinsically snake oil? No.
It's like anything else in the audio world. 90% of what you read or hear is BS. Take the parts that make sense to you and learn how to apply them judiciously to your amplifiers and see if they make a positive change.
If yes, cool. If no, well then it didn't work for you in your particular case and you are now free to go try something else.
It's like anything else in the audio world. 90% of what you read or hear is BS. Take the parts that make sense to you and learn how to apply them judiciously to your amplifiers and see if they make a positive change.
If yes, cool. If no, well then it didn't work for you in your particular case and you are now free to go try something else.
Any damping factor above about 20 becomes as it gets swamped by all the R between the amp & the air.
Damping factor is a poor way of expressing amplifier output impedance, Rout.
As Rout increases (and crosses over to current amplifier) more & more of the loudspeaker’s impedance curve imposes itself on the frequency response.
dave
Damping factor is a poor way of expressing amplifier output impedance, Rout.
As Rout increases (and crosses over to current amplifier) more & more of the loudspeaker’s impedance curve imposes itself on the frequency response.
dave
In the old days (40s-50s?) there were many amps with unity damping factor. There were various reasons, but I just wanted to point out that you can extend the bass lower at the same efficiency and in the same box. It's the only exception to Hoffman's Iron Law I know of.
It’s not an exception to Hoffman. The increase in bass due to the loss of damping only works over a narrow band at resonance - it does not truly extend the low end response. It can however, give you “a lot of bass”. Not generally deep bass, but often satisfying. If you really wanted a increase in bass at 20 or 30 Hz you could use the appropriate woofer where fs in box is that low. Stuff like that wasn’t available in the 60’s but it is now. The box *would be* larger than doing the same thing at the typical 60 or even 70 Hz that was common in the 60’s. Hoffman still works.
SE triodes with no feedback won’t get anywhere near a DF of 20. Doesn’t even necessarily need to be that high to sound good. I’ve heard iron fist bass out of tube amps with DFs of 10, and been satisfied with a DF of 2. You’re not likely to hear it if it does increase from 20 to 30 - and that’s a bigger change than going from 30 to 200.
SE triodes with no feedback won’t get anywhere near a DF of 20. Doesn’t even necessarily need to be that high to sound good. I’ve heard iron fist bass out of tube amps with DFs of 10, and been satisfied with a DF of 2. You’re not likely to hear it if it does increase from 20 to 30 - and that’s a bigger change than going from 30 to 200.
Something like that.
A conversation with a WE engineer who asked me why anyone would want to design an amplifier where Rout = Zloudspeaker that got me off the “higher the damping factor the better” wagon.
As Rout increases the need for a loudspeaker with a fairly flat impedance increase/
dave
Is the DF somehow linked to the power at which we are used to listen the music?
I mean, higher Spl means the speakers will move more, so there will be more mechanical inertia. Is it needed more DF to keep the speakers controlled vs lower Spl?
Can a low DF be used to compensate a speaker weak on bass and highs?
Sorry for the newbie questions
I mean, higher Spl means the speakers will move more, so there will be more mechanical inertia. Is it needed more DF to keep the speakers controlled vs lower Spl?
Can a low DF be used to compensate a speaker weak on bass and highs?
Sorry for the newbie questions
Theoretically an amplifier should be a good AC voltage source , with very low impedance like a power supply . With transistors it is easy to add very large amount of NFB to achieve that ... so the debate is again how much feedback is enough but from other perspective .
Last edited:
Not IME. I hear a substantial difference between 20-30-50. With the particular speakers used in the comparison there were no bass improvements above 50. Interestingly, there was an overall sweet spot around 30 where bass was acceptable and mids not too dry. The DF variance was achieved from positive current FB, so NFB was kept constant.
On the same pair of speakers a DF of 3 sounds just unacceptable.
Depanatoru,
If we must talk about Bipolar Transistors in the Tubes / Valves threads, then . . .
The emitter impedance related to emitter current, is 26/Ie, (where Ie is in mA).
Suppose a bipolar totem pole emitter output stage is standing 52mA.
The output impedance of each emitter is 26/52 = 0.5 Ohm (Parallel of 0.25 Ohm)
When one emitter current goes to zero, and the other emitter current goes to 104mA, the output impedance is 0.25 Ohm.
A PNP / NPN complementary totem pole emitter output stage does not require any negative feedback for the purpose of getting the damping factor low.
0.25 Ohm for an 8 Ohm load has a damping factor of 32, high enough to not be a factor.
Want to make the output impedance be even lower . . . use an NPN emitter follower with 1 Amp quiescent current, and a 1 Amp current source below the emitter (single ended). Now couple with an extremely large and expensive boutique capacitor.
The largest need for negative feedback on such amplifiers is to reduce the distortion.
Just my opinions.
If we must talk about Bipolar Transistors in the Tubes / Valves threads, then . . .
The emitter impedance related to emitter current, is 26/Ie, (where Ie is in mA).
Suppose a bipolar totem pole emitter output stage is standing 52mA.
The output impedance of each emitter is 26/52 = 0.5 Ohm (Parallel of 0.25 Ohm)
When one emitter current goes to zero, and the other emitter current goes to 104mA, the output impedance is 0.25 Ohm.
A PNP / NPN complementary totem pole emitter output stage does not require any negative feedback for the purpose of getting the damping factor low.
0.25 Ohm for an 8 Ohm load has a damping factor of 32, high enough to not be a factor.
Want to make the output impedance be even lower . . . use an NPN emitter follower with 1 Amp quiescent current, and a 1 Amp current source below the emitter (single ended). Now couple with an extremely large and expensive boutique capacitor.
The largest need for negative feedback on such amplifiers is to reduce the distortion.
Just my opinions.
Last edited:
IMHO, DF can be used as a tweaking tone tool to adjust with different speaker types. For most of the cases I've auditioned, mids and highs sounded better with lower DF, but sometimes we lose the sense of control in bass. Remember that bass enclosures can be designed to work nicely with low DF amplifiers.
Lower DF increases Qts. A speaker designer should always put this into the equation, together with the Rdc of speaker crossover components. An increase of Qts can make a midbass region "hump" in quarter-wave or BR enclosures.
I wouldn't say a particular DF value is a nice or bad thing. I'd say we just need to learn how to use and apply it under our circumstances.
Lower DF increases Qts. A speaker designer should always put this into the equation, together with the Rdc of speaker crossover components. An increase of Qts can make a midbass region "hump" in quarter-wave or BR enclosures.
I wouldn't say a particular DF value is a nice or bad thing. I'd say we just need to learn how to use and apply it under our circumstances.
Last edited:
It is snake oil. Originally there was a commercial reason behind as the constant voltage amplifier (zero Zout) is the easier and cheaper for the industry....
By no means is "the best".
The higher DF is not a general rule. As you found out there is a sweet spot. With some speakers it can be near unity...or even near zero (current driven).
@Wavebourn did show his practical method to optimize the Zout with a video a while ago.
Last edited:
Isn't this simply the current (solid state) vs voltage (tube) debate in a snake skin coat?
If your system doesn't need current then DF as a function of impedance to current isn't going to make a blindest bit of difference. If you system, is a constant voltage big current drive then sure, impedance to current will be important.
So where does that leave OTL? 😀
If your system doesn't need current then DF as a function of impedance to current isn't going to make a blindest bit of difference. If you system, is a constant voltage big current drive then sure, impedance to current will be important.
So where does that leave OTL? 😀
Some speakers care about DF a lot more than others. It makes a difference, usually more at low frequencies.
My preference has always been to put tube amps where they sound the best (highs and mids), then throw some chunky solid-state amplifier on the lows (Crest / MC2 / BSS). Even better if it's all done with an active crossover. Tube amps perform a lot better if you aren't asking the output transformer to go down to 40 Hz.
This is less of a concern if you don't listen to music with much low-frequency content.
My preference has always been to put tube amps where they sound the best (highs and mids), then throw some chunky solid-state amplifier on the lows (Crest / MC2 / BSS). Even better if it's all done with an active crossover. Tube amps perform a lot better if you aren't asking the output transformer to go down to 40 Hz.
This is less of a concern if you don't listen to music with much low-frequency content.
Yes. DF is overrated.
Just remember that the efficiency of a speaker is 5% on a very good day. That
means that the electrical and the acoustical side are pretty much isolated by
a huge attenuator. There is not the slightest reason to assume that the coupling
back from mechanical to electrical is any better, so forth & back we have two
such attenuators in series. The idea to CONTROL the acoustic side from
the amplifier via a second order effect is crazy.
The DC resistance of a typical 8 Ohm speaker is ~ 3 Ohms, so the amplifier
may offer stellar 0 Ohm output impedance. But the current is limited by the
voice coil resistance amongst others. On the system level, the DF is
limited to 3 just by this, and it is not 8 Ohm/zero.
You can lead a horse to water but you cannot make it drink.
Long time ago, in a previous life as a student I played with speaker control
in the style of St°ahl. That's an excellent idea. In a nut shell, the amplifier
presents negative 3 Ohms to cancel the voice coil resistance and you get
impressive low bass performance at the expense of efficiency, as long as
the speaker has enough displacement and can survive the stress.
That works for one chassis only per amplifier and speaker/amplifier
are closely coupled. Unfortunately, voice coil DC resistance is a moving
target (one bass drum kaboom and the coil temperature is much higher
for some time) and you must somehow model the other losses.
That was out of my reach 40 years ago, but with current DSP technology
it should be possible to measure and correct that in real time.
This here is DIY, but no more a topic for me. I'll keep my B&W 804.
Gerhard
We're going into the art of marketing. If we imagine, as a constructor you've reached extreme values of one potentially useless parameter and now it's time to make it popular. As mentioned above, you're going to reach high DF with lots of NFB anyway. Afterwards you just need to sell it.
The same is valid for many other parameters (beyond a point).