Hi all,
Thank you for your comments in past. I had made few amps and realized that there is no solution using only one amp to have purely three dimensional sound effect. I just came out with my own new ideas. I like class-A mid and treble sound very much but the problem is bass. I had decided why not use two amps to get nice and quality sound effect.
I have posted my ideas pic. I want to use two amp.
Mid and treble = Class A 50w
For BASS = class D 150w.
Hope for more ideas and discussion from you all.
Best regards
Michael
Thank you for your comments in past. I had made few amps and realized that there is no solution using only one amp to have purely three dimensional sound effect. I just came out with my own new ideas. I like class-A mid and treble sound very much but the problem is bass. I had decided why not use two amps to get nice and quality sound effect.
I have posted my ideas pic. I want to use two amp.
Mid and treble = Class A 50w
For BASS = class D 150w.
Hope for more ideas and discussion from you all.
Best regards
Michael
Attachments
Old idea....
hi space2000,
Ideally, the "crossovers" should be replaced by active filters before the amps.
See ESP:
http://sound.westhost.com/bi-amp.htm
regards
hi space2000,
Ideally, the "crossovers" should be replaced by active filters before the amps.
See ESP:
http://sound.westhost.com/bi-amp.htm
regards
I disagree with ESP (and most other authors/builders) on the power requirements of the various frequency band drivers.
I contend that using drivers of the same sensitivity require the same voltage supplied to the driver, irrespective of the frequency range.
If the treble, mid and bass drivers are all 8ohm, 90dB/W/m, then they all need 40Vpk to reproduce a transient peak of 110dB @ 1m.
When one mixes drivers of different sensitivity and/or impedance, only then should an adjustment be made to the supplied signal voltage to each driver.
I contend that using drivers of the same sensitivity require the same voltage supplied to the driver, irrespective of the frequency range.
If the treble, mid and bass drivers are all 8ohm, 90dB/W/m, then they all need 40Vpk to reproduce a transient peak of 110dB @ 1m.
When one mixes drivers of different sensitivity and/or impedance, only then should an adjustment be made to the supplied signal voltage to each driver.
AndrewT said:
Hi,
Well the implication you are making is transient peaks in music can
occur anywhere in the frequency range, which is true is terms of CD
red book standards, but ignores typical high quality music spectra.
The latter is the more pragmatic approach. The former is rigourous
but over-engineering. I am not suggesting amplifiers should be
chosen on average spectral power either, the difference between
say bass / mid and treble amplifiers should be less than this implies.
The argument regarding voltage is a non argument. It applies whether
the c/o point is 100Hz, 1Khz or 10Khz, i.e. it misses the point completely.
100W for below / 100w for above 10Khz is obviously nonsense.
/sreten.I don't agree on nonsense.sreten said:
I believe the transient peaks are very similar when looking/listening at/to a whole range of music.
That's where discos fall down. The treble SPL levels are way down on the rest of the frequency range. This leads to gross overloading of the treble and removes all semblance of realism, particularly on female vocals and similar HF content.
sreten said:The argument regarding voltage is a non argument. It applies whether
the c/o point is 100Hz, 1Khz or 10Khz, i.e. it misses the point completely.
100W for below / 100w for above 10Khz is obviously nonsense.
I agree with AndrewT, the output voltage of the amps should be the same. You can get the same amplitude HF and LF in music.
Hi,
Well you are both missing the point completely, some engineers .......
Your saying two 100W amplifiers can be used split at 100Hz,
1kHz or 10kHz and it makes no difference, it is "technically correct"
because of voltage capability, you cannot see the wood for the trees.
At 10kHz it only makes sense for white noise and fried tweeters.
FWIW :
Take 100w / 400w amplifiers , the 400w has x2 voltage capability.
Take 4 x 100w amplifiers optimally split, you can argue about this but
lets use pink noise, so they all need ~ the same width in octaves.
Say 115Hz, 640Hz, 3.5khz c/o points.
(edit :
do you seriously think 4 x 100w is correct for 5khz, 10kHz, 15kHz ?)
What theoretically is the maximum equivalent voltage swing of these
4 amplifiers into a 4-way speaker, ie. what single amplifier offers
the same maximum voltage capability in each band simultaneously ?
The answer is 1.6kW ....... even though max average = 4 x 100W,
(edit) wideband and only 100w continuous sine wave power.
So should the input to the 4-way active crossover be limited to be
the equivalent voltage capability of each amplifier ? of course not,
unless the only input allowed is a sine wave, where it makes sense.
/sreten.
Well you are both missing the point completely, some engineers .......
Your saying two 100W amplifiers can be used split at 100Hz,
1kHz or 10kHz and it makes no difference, it is "technically correct"
because of voltage capability, you cannot see the wood for the trees.
At 10kHz it only makes sense for white noise and fried tweeters.
FWIW :
Take 100w / 400w amplifiers , the 400w has x2 voltage capability.
Take 4 x 100w amplifiers optimally split, you can argue about this but
lets use pink noise, so they all need ~ the same width in octaves.
Say 115Hz, 640Hz, 3.5khz c/o points.
(edit :
do you seriously think 4 x 100w is correct for 5khz, 10kHz, 15kHz ?)
What theoretically is the maximum equivalent voltage swing of these
4 amplifiers into a 4-way speaker, ie. what single amplifier offers
the same maximum voltage capability in each band simultaneously ?
The answer is 1.6kW ....... even though max average = 4 x 100W,
(edit) wideband and only 100w continuous sine wave power.
So should the input to the 4-way active crossover be limited to be
the equivalent voltage capability of each amplifier ? of course not,
unless the only input allowed is a sine wave, where it makes sense.
/sreten.sreten, sorry I can't get the point you try to make, let me explain my way of thinking, perhaps we can figure out where we differ and why.
When looking at a musical signal from a CD Player on an oscilloscope you can get high and low freq signals of the same peak voltage. If you amplify this signal, both high and low frequencies will be amplified by the same factor. If you use say a 100W amp (40V, 0, 40V supply) for low freq and 50W (20V, 0, 20V supply) for the high freq, you will run into clipping on the high freq amp long before the low freq amp.
André
When looking at a musical signal from a CD Player on an oscilloscope you can get high and low freq signals of the same peak voltage. If you amplify this signal, both high and low frequencies will be amplified by the same factor. If you use say a 100W amp (40V, 0, 40V supply) for low freq and 50W (20V, 0, 20V supply) for the high freq, you will run into clipping on the high freq amp long before the low freq amp.
André
Hi,
An oscilloscope is not a realtime frequency analyser,
you cannot see the peak levels in each frequency band.
A complex signal consists of many frequencies, higher frequencies
ride on top of lower frequencies, most of the time. Simply consider :
An amplifier capable of 20Vrms.
It can do 10V @ 100Hz + 10V @ 10kHz
(or any combination adding to 20V)
It can do 7V @ 100Hz + 7V @ 1kHz + 7V @10kHz
(or any combination adding to 20V)
For a multitone signal or random noise they can only add up to 20V.
An amplifier can only produce a full power transient @ say
"15kHz" in the abscence of any other signal, this is clearly
ludicrous, nothing worth listening to has that sort of spectrum.
For an amplifier playing back music with mild clipping :
Nowhere at any point is the amplifier producing a particular
"frequency" at anywhere near full power at that "frequency."
Further thought will indicate power at a frequency is fairly
meaningless without defining frequency bandwidth, also
voltage swing at a frequency the same, fairly meaningless.
100W / 50W split at ~ 1kHz would be fine, treble end is generous.
On music programme the 100w bass end would clip first nearly all
the time, and would set the maximum level clipping is tolerable.
the treble end basically would not normally clip.
/sreten.
An oscilloscope is not a realtime frequency analyser,
you cannot see the peak levels in each frequency band.
A complex signal consists of many frequencies, higher frequencies
ride on top of lower frequencies, most of the time. Simply consider :
An amplifier capable of 20Vrms.
It can do 10V @ 100Hz + 10V @ 10kHz
(or any combination adding to 20V)
It can do 7V @ 100Hz + 7V @ 1kHz + 7V @10kHz
(or any combination adding to 20V)
For a multitone signal or random noise they can only add up to 20V.
An amplifier can only produce a full power transient @ say
"15kHz" in the abscence of any other signal, this is clearly
ludicrous, nothing worth listening to has that sort of spectrum.
For an amplifier playing back music with mild clipping :
Nowhere at any point is the amplifier producing a particular
"frequency" at anywhere near full power at that "frequency."
Further thought will indicate power at a frequency is fairly
meaningless without defining frequency bandwidth, also
voltage swing at a frequency the same, fairly meaningless.
100W / 50W split at ~ 1kHz would be fine, treble end is generous.
On music programme the 100w bass end would clip first nearly all
the time, and would set the maximum level clipping is tolerable.
the treble end basically would not normally clip.
/sreten.OK sreten, I see what you are getting at. I believe you are talking about active crossovers or else both amps wil still amplify all frequencies, driving the smaller amp into clipping.
The signal I was refering to wasn't that complex, it was a drum and a cymbal and the cymbal was producing the same voltage levels as the drum, even while playing on it's own.
André
The signal I was refering to wasn't that complex, it was a drum and a cymbal and the cymbal was producing the same voltage levels as the drum, even while playing on it's own.
André
Hang on here...
By this theory as proposed an amplifier presented with any two or more frequencies within its bandpass (crossed over or not) is incapable of producing a signal = to "full output" except to the extent that said signal's input is reduced in amplitude - that amplitude being lower and lower as more harmonics are added?
I don't think that this is exactly how it works.
Perhaps I am in error on this, but if the harmonics are constructive, if you add in enough harmonics over a period of time, you end up with a replication of a square wave. All signals starting at time 0...
Otoh, if two signals (without regard to frequency?) are reproduced by separate amplifiers, and separate speakers, then what is the acoustic sum of two speakers with identical input levels? ( the same case as originally mentioned)
So we have two separate cases: within a passband, where different frequencies are interacting; and between two passbands where there is only acoustic summing.
It seems to me that in practice one tends to run any given amplifier 6-10dB or more below clipping for just these reasons - since within any passband, there will always be more than one "signal" presented that wants to "add" above the average level.
It seems to me that if you did a test with only pure signals, run to an amplifier's clipping then the case of needed "1.6kw" to equal the output of 4 x 100w amps in a multiway system might work out properly? But no one runs their system like that? (I hope).
Let's consider it this way... if one ran flat amplitude noise into a system that was not clipped at the equivalent power level that a 100w amp could produce (some voltage into the load), would the ouput level of the system be any different if the system was multi-way amp'd with active xovers & 4 amps (to use our example), vs. run at the same input level with just a single 100w amp? If the SPL produced was different, then why?
Curiouser and curiouser said Alice...
_-_-bear
By this theory as proposed an amplifier presented with any two or more frequencies within its bandpass (crossed over or not) is incapable of producing a signal = to "full output" except to the extent that said signal's input is reduced in amplitude - that amplitude being lower and lower as more harmonics are added?
I don't think that this is exactly how it works.
Perhaps I am in error on this, but if the harmonics are constructive, if you add in enough harmonics over a period of time, you end up with a replication of a square wave. All signals starting at time 0...
Otoh, if two signals (without regard to frequency?) are reproduced by separate amplifiers, and separate speakers, then what is the acoustic sum of two speakers with identical input levels? ( the same case as originally mentioned)
So we have two separate cases: within a passband, where different frequencies are interacting; and between two passbands where there is only acoustic summing.
It seems to me that in practice one tends to run any given amplifier 6-10dB or more below clipping for just these reasons - since within any passband, there will always be more than one "signal" presented that wants to "add" above the average level.
It seems to me that if you did a test with only pure signals, run to an amplifier's clipping then the case of needed "1.6kw" to equal the output of 4 x 100w amps in a multiway system might work out properly? But no one runs their system like that? (I hope).
Let's consider it this way... if one ran flat amplitude noise into a system that was not clipped at the equivalent power level that a 100w amp could produce (some voltage into the load), would the ouput level of the system be any different if the system was multi-way amp'd with active xovers & 4 amps (to use our example), vs. run at the same input level with just a single 100w amp? If the SPL produced was different, then why?
Curiouser and curiouser said Alice...
_-_-bear
Re: Old idea....
Hi,
Refering to Fig. 2 of Elliots site, it is clearly shown that equal voltage is required.
I do believe though that bass notes require more voltage swing to hit target spl's. Just not to the point of Andrew's disco example.
My current bi-amp project will have equal voltage rails.
Why? Just because.
Greg Erskine said:
Hi,
Refering to Fig. 2 of Elliots site, it is clearly shown that equal voltage is required.
I do believe though that bass notes require more voltage swing to hit target spl's. Just not to the point of Andrew's disco example.
My current bi-amp project will have equal voltage rails.
Why? Just because.
Redesign amp and speaker system.
hi,
i just read from some web that woofer don't need crossover so i just showing here how this design. i know the costing will be high. as for sound quality, i would spend that much.
anyone have any ideas. i need more ideas pls.
class A amp is 50w and for class AB will be 100 to 150w. last time i was thinking to make class D for woofer but now i change this.
Treble and mid= CLASS A amp 50W
for woofer = CLASS AB 100 to 150w
is this nice combination? or have to be change?
thank you.
michael
hi,
i just read from some web that woofer don't need crossover so i just showing here how this design. i know the costing will be high. as for sound quality, i would spend that much.
anyone have any ideas. i need more ideas pls.
class A amp is 50w and for class AB will be 100 to 150w. last time i was thinking to make class D for woofer but now i change this.
Treble and mid= CLASS A amp 50W
for woofer = CLASS AB 100 to 150w
is this nice combination? or have to be change?
thank you.
michael
Attachments
..some more complicated bass boxes might have a natural filtering effect i guess but imho not putting a linefilter in there is just cheap..
(so if what I've read here is true, I should use three amps for my three elements with the possability that the tweeter might be classA and with lower voltage.. hehhe I think I'll stick to ONE class AB amp..)
(so if what I've read here is true, I should use three amps for my three elements with the possability that the tweeter might be classA and with lower voltage.. hehhe I think I'll stick to ONE class AB amp..)
bear said:
Hi,
True. Its not necessarily that simple. But it can be.
Consider 4 x100w active c/o's properly split.
Max sine wave power at any frequency is 100W. Seems pointless.
Why not go passive c/o with 400W at have 400W at any frequency ?
The point if the active set up is it can on a theoretical wideband
transient produce the equivalent of 1.6kW unclipped, i.e. do 6dB
higher than 400W under certain circumstances.
Which is better / cleaner ? you decide .....
Properly split active in my opinion.
rob3262 said:
Hi,
No that is not shown. It shows what I stated in post #12.
It can do any combination of voltage that adds to max voltage.
/sreten.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.