Say I want to make a three way active speaker. In which band would cross-over distortion be most audible.
To make my question a little clearer would follow: Which bands would be best served by a class A amplifiers and which by a class AB amplifier assuming I have two 10 watt stereo class A amps and one 50 watt class AB amp.
I am guessing that the tweeter and mids should be fed by the class A amps and the woofer by the class AB amps. Or would I be better off just making another stereo class A amp to feed the woofers.
To make my question a little clearer would follow: Which bands would be best served by a class A amplifiers and which by a class AB amplifier assuming I have two 10 watt stereo class A amps and one 50 watt class AB amp.
I am guessing that the tweeter and mids should be fed by the class A amps and the woofer by the class AB amps. Or would I be better off just making another stereo class A amp to feed the woofers.
For one, a well designed class AB amp should have very little crossover distortion, but I think your situation is somewhat restricted by another factor!
you need more power at lower frequencies, unless your bass driver is of 100dB+/w efficiency, then 10W is not going to do.
I think IMHO that you end up using the bigger amp for bass and the 10W for treble.
BTW you mentioned woofer(S) - are you using two bass drivers per side or one?
Are they going to be in a closed box, or port loaded?
I am only asking because the final efficiency of the bass driver(s?) with the cabinet can tell you if you can getaway with smaller amps.
At any rate, since crossover distortion sounds like a buzz, I would use my SET for treble, unless your handover frequency is chosen very high.
you need more power at lower frequencies, unless your bass driver is of 100dB+/w efficiency, then 10W is not going to do.
I think IMHO that you end up using the bigger amp for bass and the 10W for treble.
BTW you mentioned woofer(S) - are you using two bass drivers per side or one?
Are they going to be in a closed box, or port loaded?
I am only asking because the final efficiency of the bass driver(s?) with the cabinet can tell you if you can getaway with smaller amps.
At any rate, since crossover distortion sounds like a buzz, I would use my SET for treble, unless your handover frequency is chosen very high.
Your question just egged my on to read up a bit (just enough to be dangerous ;-) ) on crossover distortion.
From what I gather it is a fixed level anomaly.
So in a Class B amp when it occurs at the zero crossing point it is the most obvious as it is momentarily the only signal coming out of the amp.
I guess the idea with Class AB amps is to put the anomaly at a point where its signal is masked by the music signal. The higher the bias into class A the more masking (or the lower a percentage of the total signal is due to xover distortion).
It also appears to be a treble problem when audible.
I've stumbled across a spectrum analysis of crossover distortion which showed that it consisted of a 10kHz tone and its odd harmonics but I'm not sure if they started off by playing a 10k tone or if it's 10k with any signal.
(Distortion)
With that in mind I'd use the AB amp for bass all else being equal.
Since your AB is the more powerful anyway is of course a bonus.
In a 3way you could drive mid/high via a passive network and the bass active.
Any benefits of active operation is most audible in the bass.
At least to my ears. ;-)
Disclaimer: Everything I said about crossover distortion technicalities could turn out to be BS. It's what I gathered to be the case after 45 minutes of googling.
From what I gather it is a fixed level anomaly.
So in a Class B amp when it occurs at the zero crossing point it is the most obvious as it is momentarily the only signal coming out of the amp.
I guess the idea with Class AB amps is to put the anomaly at a point where its signal is masked by the music signal. The higher the bias into class A the more masking (or the lower a percentage of the total signal is due to xover distortion).
It also appears to be a treble problem when audible.
I've stumbled across a spectrum analysis of crossover distortion which showed that it consisted of a 10kHz tone and its odd harmonics but I'm not sure if they started off by playing a 10k tone or if it's 10k with any signal.
(Distortion)
With that in mind I'd use the AB amp for bass all else being equal.
Since your AB is the more powerful anyway is of course a bonus.
In a 3way you could drive mid/high via a passive network and the bass active.
Any benefits of active operation is most audible in the bass.
At least to my ears. ;-)
Disclaimer: Everything I said about crossover distortion technicalities could turn out to be BS. It's what I gathered to be the case after 45 minutes of googling.
Ken, I have been using a stereo JLH amp at 10w per channel for the past 8 months instead of all the other amps I have, it just sounds too sweet to get rid of. In the mean while I build one more, for no particular reason so I have two.
Previously I did have an active set-up using 3 x 100 watt amps of my own design with a Beringher 3 way cross-over into my set of transmission-line 3 way speakers.
I was happy with the setup for years until one day I felt compelled to build the 1969 version of JLH's 10 watt amp and connect it to my BBC 3/5a speakers in the office. This amp sounded so nice that I decided to take it home one evening and connected it to my regular system and that is where it stayed, everything else became secondary. It played more than load enough and I have a pretty big listening area (11 x 9 meters) so I do not have a power problem.
Before starting another amp, my question was really where is cross-over distortion more noticeable, in the low frequency range or the upper range, has anybody tried using three class a amps in a 3 way active system. I am talking experience not guessing, anyone can guess and find a reason why he should have guessed right.
One of you guys in the best DAC discussion runs an active three way set-up using DSP in the crossover, and it would be interesting to hear his motivation.
Previously I did have an active set-up using 3 x 100 watt amps of my own design with a Beringher 3 way cross-over into my set of transmission-line 3 way speakers.
I was happy with the setup for years until one day I felt compelled to build the 1969 version of JLH's 10 watt amp and connect it to my BBC 3/5a speakers in the office. This amp sounded so nice that I decided to take it home one evening and connected it to my regular system and that is where it stayed, everything else became secondary. It played more than load enough and I have a pretty big listening area (11 x 9 meters) so I do not have a power problem.
Before starting another amp, my question was really where is cross-over distortion more noticeable, in the low frequency range or the upper range, has anybody tried using three class a amps in a 3 way active system. I am talking experience not guessing, anyone can guess and find a reason why he should have guessed right.
One of you guys in the best DAC discussion runs an active three way set-up using DSP in the crossover, and it would be interesting to hear his motivation.
Charles,
I hear you and I think your speculation and reading matches my own, but would like to hear from guys that do run class A rigs to give some indication of what they have found.
Now I do not want to praise class A at all but there exists something very different in the sound character between class A and over biased class B (A/B) and I think it is cross-over distortion, which I think has a hard characteristic sound to it, very much like 5th or 7th order overtones.
Am I dreaming?
I hear you and I think your speculation and reading matches my own, but would like to hear from guys that do run class A rigs to give some indication of what they have found.
Now I do not want to praise class A at all but there exists something very different in the sound character between class A and over biased class B (A/B) and I think it is cross-over distortion, which I think has a hard characteristic sound to it, very much like 5th or 7th order overtones.
Am I dreaming?
Quick answer is No, I have never set up three class A amps in a 3 way.
What I have done is make a three way active speakers, many moons ago in the 80's.
I used a (now vintage) Meridian 105 mono poweramp for the bass drivers (2X Peerless 8"), setup so I could have two little pots adjusting the bias point and DC offset while working.
At low bias (mostly class B) I could hear the crossover distortion as buzz!
increasing the bias would irradiate the distortion both audible and seen on a scope.
My trick was increasing the bias high enough to see no trace of it on the scope, while making sure the amp was not over heating, then add a simple coil after the amp, to filterout any trace of it.
For mid and treble, I just used a healthy bias point, and since the amps were smaller, they wouldn't overheat that much.
I used them as my main speakers for 10 years or so, before giving them to a friend, since I was immigrating, they are still working, sounding good.
that was my experience.
What I have done is make a three way active speakers, many moons ago in the 80's.
I used a (now vintage) Meridian 105 mono poweramp for the bass drivers (2X Peerless 8"), setup so I could have two little pots adjusting the bias point and DC offset while working.
At low bias (mostly class B) I could hear the crossover distortion as buzz!
increasing the bias would irradiate the distortion both audible and seen on a scope.
My trick was increasing the bias high enough to see no trace of it on the scope, while making sure the amp was not over heating, then add a simple coil after the amp, to filterout any trace of it.
For mid and treble, I just used a healthy bias point, and since the amps were smaller, they wouldn't overheat that much.
I used them as my main speakers for 10 years or so, before giving them to a friend, since I was immigrating, they are still working, sounding good.
that was my experience.
Similar to above.
Never used pure class A amps but my current active stereo uses amps which are quite heavily biased so that the first 20W or so are class A before transitioning into B.
So in normal operation they hardly ever get there but they do get quite warm at idle since then I can hear the fans kicking in. I think of them as class A amps with bonus headroom. Wouldn't necessarily want to run 200W Class A amps to get the headroom I rarely use.
When playing I hear nothing annoying whatsoever, not even from the horn tweeter (around 106dB/1W/1m) of the Tannoy part. Just a smooth, clean and non-fatiguing sound. The bass up to 250Hz is covered by 12" Volt radials.
I think Nelson Pass follows a roughly similar philosophy in some of his amp designs which you may well be able to diy.
Never used pure class A amps but my current active stereo uses amps which are quite heavily biased so that the first 20W or so are class A before transitioning into B.
So in normal operation they hardly ever get there but they do get quite warm at idle since then I can hear the fans kicking in. I think of them as class A amps with bonus headroom. Wouldn't necessarily want to run 200W Class A amps to get the headroom I rarely use.
When playing I hear nothing annoying whatsoever, not even from the horn tweeter (around 106dB/1W/1m) of the Tannoy part. Just a smooth, clean and non-fatiguing sound. The bass up to 250Hz is covered by 12" Volt radials.
I think Nelson Pass follows a roughly similar philosophy in some of his amp designs which you may well be able to diy.
Let's not confuse class A with a SET amplifier.
A push-pull amp can be class A too or class AB in case of my vintage amp, and I guess with Charles.
I hazard a guess that what you are hearing is even order harmonic distortions, which makes SET amps sound warm and rich, perhaps.
on the other hand a push-pull (even class A) amp has little even harmonic distortions (the push-pull cancels them out) leaving odd harmonic distortions being the dominant, which in comparison makes them not as musical, not warm or rich (but more pure).
A well designed class AB push-pull amp should have negligible crossover distortion, certainly not audible.
If you get used to one, the other sounds odd .
My 2 pence worth.
Last edited:
Yes and no. You're hearing a difference but its not due to crossover distortion. I'd say its due to the difference in levels of noise generated on your power supply. In my experience, classAB can sound sweeter when its power supply impedance at HF is significantly reduced. And also when its PSRR is improved. Neither of these mods affect the crossover distortion.
Thanks to all for your advice, I will use the class A/B for woofers and get used to it.
B.T.W Ken I use two Linsley Hood stereo class A amps with 120mF and 27VDC regulated from 300VA transformer, which I could never quite work out whether it is push-pull or set because it can be explained to behave as either.
B.T.W Ken I use two Linsley Hood stereo class A amps with 120mF and 27VDC regulated from 300VA transformer, which I could never quite work out whether it is push-pull or set because it can be explained to behave as either.
"get used to it" ....man!
AB can be manipulated to be even order dominant.
For woofers , you will run cool and quiet.
I've made the slewmaster PSRR on some of the input stages >120db.
Much research and prototyping has suppressed odd order by
fine tuning the driver compensation to the output Cob. AB does
not have to be "sloppy' !
"5th and 7th" - these are barely even measurable with some
of the input stages.
That "first watt" with heavy A bias up to >10W is poppycock.
OS
AB can be manipulated to be even order dominant.
For woofers , you will run cool and quiet.
I've made the slewmaster PSRR on some of the input stages >120db.
Much research and prototyping has suppressed odd order by
fine tuning the driver compensation to the output Cob. AB does
not have to be "sloppy' !
"5th and 7th" - these are barely even measurable with some
of the input stages.
That "first watt" with heavy A bias up to >10W is poppycock.
OS
One may try my distortion test:
http://www.diyaudio.com/forums/ever...-nonlinear-distortion-new-listening-test.html
Signal4 is a simulation of cross-over distortion, so it is possible to try audibility of such distortion profile and level. Best ABX vs. signal2, which is a pure tone.
http://www.diyaudio.com/forums/ever...-nonlinear-distortion-new-listening-test.html
Signal4 is a simulation of cross-over distortion, so it is possible to try audibility of such distortion profile and level. Best ABX vs. signal2, which is a pure tone.
Thanks Pavel, I have been following your thread which lead me to start this one.
OS, I did not measure anything, nor did I poke a pencil in my eyes to perform the proverbial double blind test, it just occurred to me, it may be placebo or the fact that I like things hot! hence the question in this regard of other experienced designers.
OS, I did not measure anything, nor did I poke a pencil in my eyes to perform the proverbial double blind test, it just occurred to me, it may be placebo or the fact that I like things hot! hence the question in this regard of other experienced designers.
Very interesting! I have the same feeling like Nico about class-A amps. I couldn't find anything other than crossover distortion as the variable that makes the difference...
So how possible is it to set up a listening test that will explain the audibility of crossover distortion? Does signal4 sound very different than signal2 in your listening test?
The point is that a class-A amp has less high order distortion and less or no crossover distortion BUT has high 2nd order distortion. I want to prove to myself that my preference of class-A is not because of the availability of 2nd order but because of the unavailability of the other distortions.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.