I did not say they did not exist.Can't hear the difference, huh? I guess they don't exist ..........
If an amp with one type of resistors would give you 0.05% THD and with another type - 0.0500001% THD, who cares.
The difference is so insignificant that nobody can hear it why bother?
RE: input dc blocking cap.
based on what i could find, the f3 rolloff point for the 2.2uF cap would be .72hz. is that too low to do any good? i also found another formula for picking this cap's value (1/4pi*r). that resulted in a cap of 1.25uF. will it still work with that 2.2uF cap as i have some on-hand? or should i find a 1uF?
based on what i could find, the f3 rolloff point for the 2.2uF cap would be .72hz. is that too low to do any good? i also found another formula for picking this cap's value (1/4pi*r). that resulted in a cap of 1.25uF. will it still work with that 2.2uF cap as i have some on-hand? or should i find a 1uF?
I prefer to use a power amp bandwidth of a decade either side of the nominal Audio bandwidth, i.e about F-3dB of 2Hz to 200kHz. This roughly equates to F-1dB of 4Hz to 100kHz.
The DC blocking cap interacts with the following resistance to return (Rin) to create a passive high pass filter.
F-3dB of that passive filter is given by F= 1 / 2 / Pi / R / C where R=ohms C=Farads, Pi=3.14159
If you know Rin you can calculate a capacitor value to give you the filter frequency you want/need.
Some say F-1dB @ 20Hz is OK for passing bass frequencies, using speaker passband and ear/hearing passband as evidence that down at 20Hz you're not going to hear a droop of 1dB.
Some say the Power amp LF passband must extend down to DC to allow the music to be reproduced correctly.
My recommendation is in between and some agree that passive F-3dB @ 1Hz to 2Hz works for them.
Last edited:
thanks andrew! the .72hz f3 rolloff on the 2.2uF cap should be close enough to that range. if i can dig up a 1uF, i'll use that instead. but the 2.2uF seems like it should do in a pinch.
thanks for the additional info. i had found the formulae, but wasn't sure what a good practice was for choosing the rolloff.
cheers!
thanks for the additional info. i had found the formulae, but wasn't sure what a good practice was for choosing the rolloff.
cheers!
Have you confirmed that Rin ~ 100k?
Once you have the the best quality (you can afford) DC blocking cap on your input, you can alter Rin over quite a wide range to let you hear what that passive filter does to your LF passband. 1% metal film resistors are cheap.
Once you have identified the RC time constant you can live with, keep that RC value in blood and use it as the starting value for any other project you begin.
BTW,
I prefer that the pre-amps and sources pass an extra octave at both ends of the frequency passband. This helps avoid the -6dB and the droopy Q=0.5 that would happen if the two passives were set to the same frequency
Once you have the the best quality (you can afford) DC blocking cap on your input, you can alter Rin over quite a wide range to let you hear what that passive filter does to your LF passband. 1% metal film resistors are cheap.
Once you have identified the RC time constant you can live with, keep that RC value in blood and use it as the starting value for any other project you begin.
BTW,
I prefer that the pre-amps and sources pass an extra octave at both ends of the frequency passband. This helps avoid the -6dB and the droopy Q=0.5 that would happen if the two passives were set to the same frequency
I think that it is an imagination. People believe if they have, for example, better wires the amp and speaker sound better, but with a double blind test they don't hear any difference. Same about resistors, capacitors etc.
Someone even believes in direction of wires and resistors. Someone thinks power cord does matter.
Frags, if you do not believe or have not found that the changing of power resistors in the F5 makes an audible difference, fine. Others, including myself have deluded ourselves otherwise... there is another thread on listening impressions of the F5 relating to changes in various aspects of the implementation... you might want to check that one and see what people said.
Of course people also think that the different brands of output Mosfets make a difference too, also the biasing point and IF they are matched or not... but they're probably wasting their time and effort on this too?
If ur a DBT guy, I suggest that there is little point in DIY or for that matter the F5...
unless of course you are also on a tight budget and are trying to save a few $$, but look, a decent receiver is both cheaper and has much lower distortion than an F5, right??
This is not the thread to debate "DBT" - so let's just agree to differ on this aspect of the F5?? Ok?
Regards,
_-_-bear
Peter Daniel had an interesting post about DC offset and speakers.
I think he (maybe accidentally) ran a speaker at close to full rail voltage on a chipamp for a while without any damage.
My personal experience is that anything above a few volts is asking for major trouble, and you should not be measuring anything more than 100mV at speaker output terminals.
Your 44mV reading should be safe. I got about 60mV because my DAC has no output capacitors at all and the F5 was directly coupled.
I think he (maybe accidentally) ran a speaker at close to full rail voltage on a chipamp for a while without any damage.
My personal experience is that anything above a few volts is asking for major trouble, and you should not be measuring anything more than 100mV at speaker output terminals.
Your 44mV reading should be safe. I got about 60mV because my DAC has no output capacitors at all and the F5 was directly coupled.
BTW, 15 V DC killed two pair of disposable PC speakers
D.
I dont worry about running anything under 200mv to good full range drivers. of course it is not ideal, but sometimes I work with less than ideally matched outputs.
Yes, I have had full rail voltage smoke a couple of pair of cheap speakers, that's why we use them in the shop phase.
