I think we can explore a tradeoff and still be friends. I in general would prefer not to have to talk about distortion that might be a sonic coloration without putting 40 transistors in the signal path.
Also it seems that whenever there is an internal node that needs bypassing someone will stick a $100 Teflon capactor there and insist it is necessary sonicly.
I think this is a point worth exploring, no capacitors (save Ccomp), single bias source and a fairly small number of active devices to get sub-ppm distortions.
Right, look at some of Nelson Pass' latest stuff. Exploring new FET's that give lower measured distortions in simple circuits. The operating principle (IMO) is that the distortions is these very simple circuits are enough to actually be detectible so some useful reduction is good.
How far are you from Frankfurt? We have friends in Heppenheim and visit every few years.
Last edited:
i admire Pass a lot but why does he limit bandwidth so much ? we worked close to researchers at Würzburg university and found that high speed circuits are essential for preserving the first wave shape for better precission in lokalisation. we found the limit of perception at 1.5Mhz (-3dB). fouriers analysis does not capture noise components that are constituted of "half sines".
I really doubt this - people not understanding correlation keep confusing group delay resolution with bandwidth - they are not the same or even simply related
S/N dynamic range and observation/integration time are also required to estimate group delay resolution in addition to bandwidth - greater S/N and observation time can extend group delay resolution to far below 1/bandwidth
other factors:
no microphone used in recording has > 100 KHz bandwidth
>>100 KHz sound waves are heavily attenuated in the air
only a few tweeters have >20KHz response
so claims of MHz amplifer bandwidth requirements in audio reproduction are suspect
Last edited:
We once built two versions of the same circuit (a variation of a Nelson Pass 100W Class A design), one of which has a bandwidth (-3dB) of 100kHZ, and the other 350kHz. The topology and component choice were otherwise identical. We connected them via a specially built switch to AB test on the same speakers and from the same source. The result was immediately obvious. You can clearly hear a difference. Which one one prefers is a matter of taste, I guess. And I cannot offer you a scientific explanation.
1.5MHz power amp may be a bit much, but I would certainly not object to 800kHz, like Nelson's F5.
Patrick
1.5MHz power amp may be a bit much, but I would certainly not object to 800kHz, like Nelson's F5.
Patrick
dear jcx, i got my take on time issues. as a speaker builder with 40 years experience and 100.000th of objects sold to very happy customers i am a noise expert. one of my teachers is Mr.Manger with 40 patents on the subject. i talk about everything kept the same. ONLY changing the bandwidth and using time coherent speakers. each filter type has it´s own issues with time. sub critical beeing the best (Q less the 0.5) but who has heard about that when many say bessel (0.606) has the best groupdelay performance. they just repeat what the "experts" say and do not make any own fundammental research. i do not expect that you trust any of my words. i got so tired about arguing for extended bandwidth that i gave up. it is posible to build extremely high speed poweramplfiers that are perfectly stable and i have microphones that go up to 100kHz.
first try it and then condamn it.
first try it and then condamn it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.