Is there anybody out there, who has had any experience, bad or not, with audioequipment having too big a bandwidth, here I'm talking about >1MHz.
The reason for my Q is, that I'm currently building a ballanced preamp w/o global feedback, here we can exclude ordinary oscillation, having a simulated (PSpice) bandwidth of @20MHz.
Too Much Bandwidth
I would think that you would want to limit the bandwidth at the
output. . .you can only hear so high a frequency. Some digital
amps operate in high megahertz. . .but still limit at output.
Why no global feedback? Too flat a response? Like DC offset? Too evil? Just wondering. Global feedback works great for 99.9% of all consumer, professional, commercial, and studio gear. Just wondering what your technical justification is.
global feedback or not,
you will get more problems, than you will get benefits
from using upper limit 1.000 times above audio band
for a shortwave transmitter or receiver circuit, yes
for audio .. sorry .. to me it is a little bit stupid
to avoid troubles and unwanted distortions / interferences
I have no fear of limit upper frequency somewhere in 50-100 kHz region
but I am sometimes a rather brave man!
already this bandwidth, gives almost perfect square waves at 2-5 kHz
if you build for your oscilloscopes and to just get good and impressive figures onto a paper,
you go ahead
personally, I wont be much impressed anyway
and for good sound reproduction such is of very limited positive value
knowing that air can not carry square waves in upper audio
makes it a waste of effort trying to build circuits to feed loudspeakers with this
air, as a medium of sound, acts as a low pass filter in itself,
it is real physical thing, that we can't do much about
... maybe you like to play your square wave music in total vacuum ...
That situation started in the seventies...when people decided to install....
Radio frequency transistors inside audio equipment.
The called that "high speed".....and really high was the speed your neighboor was around you complaining about your Radio Telephone (wireless telephone) that worked in 27 Megahertz, remote control frequencies (receiving beating from nearby frequencies)..... and them passed to 49 Megahertz, 100 Megahertz and now a days working into 900 Megahertz...and complaining because when your telephone was beeing used, he was picking up noises...in special when using turntables and Vinyl...and those things are beeing used till our days... as they sound nice.
Noises were detected and picked by audio appliances....was a problem...more a problem than a solution....and some guys telling that waveform at 20 Kilohertz were better when using 250 Megahertz transistors....my God!.... the Zobel was limiting....no sense at all.
The amplifier will oscilate following radio frequencies that will be captured by component leads and all ferric material you can have inside your equipment..... that enormous polution we have, related magnetic field surrounding us, will be picked and transformed in some microvolts...sometimes milivolts, that will cross the "high speed" amplifier and will produce some output..sometimes will saturate stages, and you can see that those amplifiers works more hot than usual...because working above audibility range too.
Analog AC voltimeters will show some signal without leads connected into the circuit (RF pickup)
A lot of capacitors and compensations will be needed to avoid that... so.... the use of big bandwidth will create a problem that will need solutions...this means to create and to eliminate.....not a good idea...better not to create some problem that you will need to avoid latter.
Wow, that was an old one that just came alive :)
The question was more a theoretical one, since I at the time had very little experience in designing, and wanted to hear of other peoples experiences.
I have since found out, as Leolabs is mentioning, that the PCB-layout is more than critical at such high frequencies, when using global feedback, but actually not as critical when only using local feedback.
Biggest problem was, as Carlos referes to, "radio Moscow", as it's called in Denmark, AND what I'd call load-reflections from what ever power-amp is connected!!!!
I'm actually just running my preamp passively (DC-UV), input-selector and volumn, but no gain, and it works fine :cool:
Thank You every one for Your feedback.
Re: Too Much Bandwidth
You should really look up how they work before including them in an answer where they have little or no relevance.
I would reduce the bandwith to somewhere around 1 - 4 MHz. 20 MHZ could give problems, since it will then be able to boost some digital signals from your CD-player etc...
And for DCPreamp:
There are MANY reasons for using non-feedback design. Best reason is, that feedback is absolutely not usefull for audio equipement!! ;) ;) ;)
The idea of negative feedback is, that you compare input and output signal, and then returns a correction signal to get correct the non-linear output signal. But.... you can't fix the problem before it appears at the output. And when it does appear at the output, it's already in your speaker. The corrected signal is therefor ALWAYS behind. Your THD analyser will not notice this, but your ears will!!
I do not agree that feedback works fine in 99% of all equipement on the market. You should try listening to some high quality non feedback gear. Then you will newer be able to listen to feedback again. But off cause... If feedback is all that you have tried, you would think that this is the best availeble.
Non feedback is VERY popular here in scandinavia. Try listening to amplifiers from one of these companys:
Non feedback technology is in a class of its own. Problem is that you have to be VERY careful in the designfase, since you have no feedback loop to correct you bad engineering. Also non feedback is more critical about components... You can't use cheap components, again because of no loop to correct the bad components. That some of the reasons that most products use lots off feedback.
limit output vs filter output
On previous post, I used limit. . .sorry for confusion. . . I do mean
filter of some sort before being output to the transducer. I believe
most 'digital amps' almost require some sort of passive filter to
work correctly anyway. Complex impedances an all. . . I think digital
amps used to suffer from high output impedance. . .no longer such
a problem now? again, I think. . . Haven't designed one. . .
Audio bandwidth is not the same as oscillation frequency. . .agreed.
The only reason I mentioned 'digital amps' is because they do use
high frequency oscillation but ultimately 'filter it out.' The original post
was worried about the large audio bandwidth. . . sorry about mixing
apples and berries. At some point you don't want to amplify those
high frequencies and waste energy.. . . I guess that would be filtering
before and during amplification. . . output filtering would be to prevent
it (by it, I mean the high frequency + 100kHz) from going on to the
next stage or speaker. . .it would be a waste. Simple cap to ground
should take care of high frequency at the input.
Hope that clarifies my point.
|All times are GMT. The time now is 08:15 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2013 DragonByte Technologies Ltd.
Copyright ©1999-2013 diyAudio