What amp is better for subwoofer?

V

Vladimir Kuzmin Sr.

New member
Sep 4, 2021
9
1
3
70
Russia
Concerning damping factor (DF). It’s not a secret that DF can be different in various frequency bands. Most commonly the manufacturers give it at 1kHz or so and almost never in lows. However we know that one amp can give very good sound in the bass band and other can produce very doll sound with the “one-note bass” so we can imagine that the amp has very small DF. Especially this concerns to the amps with conventional power supply. We can conclude that this “df” in lows depends on energetic behavior of power supply. I propose the formula for evaluation the behavior of amps in low frequencies which could be helpful for selecting amp for driving for example subwoofers. It uses only two honest specs which are presented in every manual for each amp: rms power at 8 and 4 Ohms loads.
Kd=1/[1.414*SqR(P8/P4)-1] and which must be more then 20 (the more the better).
Let P8=500 and P4=700, then Kd=5.13.
And P8=500, P4=950, then Kd=38. Second amp is better.
Of coarse losses in cabling must be as minimal as possible,
This formula is proven by my practice.
 
What always concerns me is my inability to assess what comes out of subs. With 50-20K, 'quality' and tone are quite obvious and unwanted amp artefacts can be heard, but only recently have I discovered I really have trouble trying to do this with what comes out of a sub, only fed with the bits at the bottom. I'm not sure I can hear things well enough to even determine if its the amp, speakers or anything? With a dodgy sub, it sounds normal until it suddenly makes the new noise denoting a speaker in distress. Up to that point it MUST be changing - something you'd hear higher up - the roughness or the onset of distortion, but in a speaker that only shifts wind for a lot of the time, what does a poor damping factor actually sound like. I found with a signal generator and tones - I could see things that I couldn't hear. It's taken me years to realise I simply don't have the ability to detect problems with LF in any meaningful way. A scope can detect overshoot and make some sense, but I very often cannot hear what it is seeing?
 
Vladimir Kuzmin Sr. said:
Yes that's right and this is the reason why I took figure 20 in my formula.
Click to expand...
What is your thinking behind needing to know if DF is higher than the audibility limit, and how does this make it important? We can already use the existing amplifier DF data to calculate the effective DF including our specific connectors and cable.
 
  • Like
Reactions: Vladimir Kuzmin Sr.
Carl Klinkenborg said:
What is your thinking behind needing to know if DF is higher than the audibility limit, and how does this make it important? We can already use the existing amplifier DF data to calculate the effective DF including our specific connectors and cable.
Click to expand...
As I state it is important to know this border of 20 and to know out does your amp meet this requirement. You yourself cannot hear the difference but other could and one day you will hear “oh man, something goes wrong with your sub”.
 
Carl - that's my problem - at below 60Hz or so I cannot detect distortion at all, let alone assess quality and hear subtle Damping factor artefacts. Above 60Hz then 'quality' judgements are so much easier.
Carl Klinkenborg said:
According to most sources of information I have found are concerned, anything above a DF of about 20 is sufficient to be inaudible.
Click to expand...
 
  • Like
Reactions: Carl Klinkenborg
I hope you've read the article here https://soundforums.net/articles/loudspeakers/damping-factor-soggy-at-best-190493/ where stated that "Scientifically speaking, Damping Factor is the ratio of the resistance component of the loudspeaker’s impedence to the resistance component of the amplifier’s output impedence. This number is used to determine how well the amplifier is able to control the motion of the loudspeaker. Too low and the amplifier won’t be able to resist back-EMF from the speaker and the cone will “flop around” a little, resulting in a less clear reproduction of the signal being sent to it." I can add that this back-EMF is really big in LF drivers with their heavy voice coils and diffusors and they "flop around" just at their resonance frequency (when DF is extremely low).
 
All these comments make perfect sense, and I am comfortable that I know what DF is and its effects, but do not understand why we need another specification when we can already work out DF in our own specific application by including connectors and cabling?
 
Carl Klinkenborg said:
All these comments make perfect sense, and I am comfortable that I know what DF is and its effects, but do not understand why we need another specification when we can already work out DF in our own specific application by including connectors and cabling?
Click to expand...
Not really.That article basically talks about how the cable and connectors affect the DF. The output impedance of the amplifier is considered to be close to 0. And you are right that we can consider the recommendations for cables and connectors. But we don't know the real output impedance of an amplifier with conventional power supply, and it can be high at low frequencies. And LF drivers with their heavy moving system can not be well controlled by the amplifier, so in this case they will give their resonant frequency at any signal in the form of LF squeal even with cabling well done. My recommendation is only for such a case.
 
Last edited:
Vladimir Kuzmin Sr. said:
Concerning damping factor (DF).... I propose the formula for evaluation the behavior of amps in low frequencies which could be helpful for selecting amp for driving for example subwoofers. It uses only two honest specs which are presented in every manual for each amp: rms power at 8 and 4 Ohms loads.
Kd=1/[1.414*SqR(P8/P4)-1] and which must be more then 20 (the more the better).
Let P8=500 and P4=700, then Kd=5.13.
And P8=500, P4=950, then Kd=38. Second amp is better.
Of coarse losses in cabling must be as minimal as possible,
This formula is proven by my practice.
Click to expand...
Vladimir,

I could agree the "Second amp is better" (not exclusively in the low frequencies), but not because of it's damping factor, but because of the progressive power supply current limiting that takes place at the lower impedance.
Low frequency response may exhibit more current limiting "pumping" due to impedance extremes and current demands, a "four ohm" cabinet may drop to 3.5 ohms at Fb (box tuning frequency), and be 20 to 40 ohms (or more) either side in just a 1/2 octave.
An amp with no current limiting above 4 ohms, P4=P8x2 would have a "Perfect Kd" of 6222.
This amp would tend to emit sparks and smoke if driving too low of an impedance when current exceeds the output devices capability.
By your metric, an amp with a P8/P4 ratio lower than 500/906.66 fails the "current limit suck" test, coming in at 19.9.
That type of amp will still sound OK at lower drive levels, just begins to suck when pushed beyond the current limits.

That said, if the speaker cables are too long and thin, damping factor falls below 20, and speaker's transient response is reduced at any level, regardless of the amplifier's damping factor or current limiting.

Art
 

Attachments

  • Current Limiting Suck Factor.png
    Current Limiting Suck Factor.png
    76.9 KB · Views: 3
  • Like
Reactions: Vladimir Kuzmin Sr.
Art Welter said:
Vladimir,

I could agree the "Second amp is better" (not exclusively in the low frequencies), but not because of it's damping factor, but because of the progressive power supply current limiting that takes place at the lower impedance.
Low frequency response may exhibit more current limiting "pumping" due to impedance extremes and current demands, a "four ohm" cabinet may drop to 3.5 ohms at Fb (box tuning frequency), and be 20 to 40 ohms (or more) either side in just a 1/2 octave.
An amp with no current limiting above 4 ohms, P4=P8x2 would have a "Perfect Kd" of 6222.
This amp would tend to emit sparks and smoke if driving too low of an impedance when current exceeds the output devices capability.
By your metric, an amp with a P8/P4 ratio lower than 500/906.66 fails the "current limit suck" test, coming in at 19.9.
That type of amp will still sound OK at lower drive levels, just begins to suck when pushed beyond the current limits.

That said, if the speaker cables are too long and thin, damping factor falls below 20, and speaker's transient response is reduced at any level, regardless of the amplifier's damping factor or current limiting.

Art
Click to expand...
Art, thanks for the nice addition, it shows your understanding of the process in the amplifier. I completely agree with it. I didn't go deep into it, hoping someone would ask deeper questions or make comments like yours. In the post I did not count the value calculated by the formula as damping factor DF, I just called it a 'conditional value df', indicating "We can conclude that this "df" in lows depends on energetic behavior of power supply" which you confirmed with your clarification. I still don't know what to call this value, I just called it Kd, not DF or df. The user has no exact knowledge of how the amplifier will behave at LF, the specifications do not specify anything other than these two values. Of course, the value of 20 is a conditional smooth boundary, not an exact one. The amplifier and its power supply start to behave strangely near maximum power, which is typical for subwoofers. In this mode, the amplifier even may have no negative feedback, and the driver itself may not only have low impedance, but also act as an oscillator. In my work (this is the sounding of large spaces) I have often found that cheap conventionally powered amplifiers do not behave well when powering subwoofers, while labgruppen type amplifiers behave fine not only to power subwoofers, but also as wideband amps, increasing the efficiency at LF. Thank you again.
 
  • Like
Reactions: Carl Klinkenborg
Top Bottom