So, what does crossover distortion sound like?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Yikes! Seems difficult at best to look through every single 50 + page thread with the word 'crossover' and 'distortion' in it.

I have seen that crossover distortion is a big argument against Class B in favour of class A.

Is crossover distortion actually audible?
If so, what does crossover distortion sound like?


Are there simple experiements that demonstrate this?
 
But it also depends on how much of it there is. You generally won't notice a little unless you are looking for it. It is different in your hifi than in your table radio. Not that the distortion would be different, but you would listen different.

A god class B aml might well sound a lot better than a crappy class A amp. You can't make overall generalizations that hold up.

As to what it sounds like? Geez, what do eggs taste like?
 
I have always found the crossover distortion sounds like kind of 'buzzy'', almost like the speaker voice coil is loose and buzzing with the music. The effect is stronger with quiet music and disappears as you turn up the volume.

I really dont think crossover distortion is audible at high listening levels unless its absolutely horrendus. Even then, it will be always be worse at low volumes than at high volumes.

So if you amp sounds 'buzzy' or 'grainy' for very quiet passages but ok for loud passages, then you might have crossover distortion.

You might have to put your ear up to the tweeter to hear it during the quiet passages. You cant make it more audible by turning up the amp, the sound of the music will mask it.

Doug
 
music soothes the savage beast
Joined 2004
Paid Member
i have been tweeking one solid state class AB amp lately, where i was changing bias current in output power transistors, from almost nothing to realy high
so i believe that i was actualy changeing crossover distortion
i did comparison how it sounded on my stax earphones on a couple of cds
sound with very low bias (almost class B amp) was a little flatter, had less depth, more grain, less injoyable
with a current set at recommended level, sounded fine, just like you would expected with well designed class AB amp
however, when I turned the bias over the recommended level, I believe i was running it in low output in almost class A like mode (offcourse its not possible as fully class A, because inherently it has two output devices), the sound was much smoother, simply more musical, one could not stop listening and wanted to try one cd after another
this is just my personal view, i might be wrong, how crossover distortion influences sound
 
music soothes the savage beast
Joined 2004
Paid Member
Even if it is not fully audible is seems to grate on the ear and makes listening tiring.

i agree
with the bias set too low, in class B like mode, high crossover distortion, sound did not pulled me in
no matter what level
it might be more audible on lower level, but even at high volume its allways there, for the signal has to cross thrue zero no matter what level

this might not necassarily has anything to do with crossover distortion, but i have noticed many times that in bad sounding mass produced receiver, the sound was not messed up by power section, but by preamp section
for instance in one fisher receiver, spec'd at 2x100 watts, amp built from separate parts of pretty high quality with four power transistors, when preamp section was bypassed, sounded excelent
like many remote controled receivers, pream section had many IC's to control volume/balance/bass/trebles by swithing circuitry
somehow this was messing up the sound terribly
when I instaled dual pot for volume and put it directly to power amp section, it sounded dynamic and clean in contrast to original flat boring sound
i guess a phase issue here
 
If designed correctly class B amplifiers can sound fine.

McIntosh is an example of this.

If designed poorly they will not sound good even when the bias is increased into class AB.

A Phase Linear 400 and Crown DC 300A are examples of this.
 
When the outputs are biased in cutoff, it takes a small step in Vbe for the output to turn on and output above 0V. The amount of the audio signal that is below this turn-on step is what is lost. If the signal is smaller, more will be lost. The slew rate comes to mind here. I suppose if the slew rate of a class B was good enough, crossover could be minimized?
 
cunningham said:
...The slew rate comes to mind here. I suppose if the slew rate of a class B was good enough, crossover could be minimized?
Theoretically, crossover distortion would be completely eliminated by an amplifier with both infinite open-loop gain and with a stage preceding the output stage with infinite slew-rate.

Of course no such amplifier exists, but it's surprising how little distortion is left with just an op-amp + unbiased emitter follower.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.