I could have been more explicit as to my comments. My post contained the quote "proper absolute phase" from the BOZ line stage introduction. In the original Zen article NP was explicit about the speaker polarity connections in regard to preserving phase. The De-Lite article does not explicitly discuss speaker polarity. Put a constant current source on the De-Lite and, for other than a depletion mode device, I see a close cousin to the Zen.
StefanoBilliani appeared apologetic for connecting the speaker negative to ground when using capacitive coupling. I imagined he was ginger because the common source stage may indicate putting the positive speaker terminal to ground. NP waxed on in the BOZ discussion about the aesthetic, if not sonic, merit of preserving phase.
StefanoBilliani appeared apologetic for connecting the speaker negative to ground when using capacitive coupling. I imagined he was ginger because the common source stage may indicate putting the positive speaker terminal to ground. NP waxed on in the BOZ discussion about the aesthetic, if not sonic, merit of preserving phase.
It's interesting how people often have a preference for a
transformer over an output cap. The cap measures better.
A lot better.... What do you suppose it is?
Well, since you ask.
I think it is a phase issue. I don't have a lot of experience with xfrmr's but, I'll bet the "spectra" having phase shift through a cap (usually not big enough) is much worse, maybe covers more octives of important area's of the music, than the phase shift through a xfrmr. Would not the cap have phase shift at the low end and the transformer at the high end? There is alot of music in the octave from say 1k to 2k. Isn't that the opposite in the octave from 10k to 20k?THD tests would not necessarily show this. I suppose IM would though...
Well, since you ask.
THD tests would not necessarily show this. I suppose IM would though...
The phase shift by caps in series to the signal is through the typical 6db/oct high pass character - also dependend from the load resistance. Therefore the cut-off frequency of the phase shift start (and not at the frequency response diagram) is relevant and must be below the audible range, i. e. 20-25 Hz. Thus the capacity value must be much more higher as assumed in general.
But what about equivalent circuits for output transformers? How I do create such circuit for checking behaviour and results on the simulation aera?
Well, as you say, by the amplitude response it is below our, so called, audible range. Many designs set roll-off at 2-3 octaves below the 20Hz minimum. But, phase wise, is the music shifted by phase in an area of our sensitivity or not? I believe the tambre of the instruments, vocals etc., the "spectra" is well known to our brain, "sensitivity", and we're very sensitive to phase too. As N.P. describes regarding "absolute phase". Maybe it's more than THD measurements can resolve? Most THD measuring instruments only measure residual after nulling the fundamental. With only one tone, no less, not the summation of many like intruments and vocals. Phase is irrelavent to them. It is only total (true)RMS voltage that was not at the input, in a relative percentage, that they measure. (No math or formulas here). Even a spectrum analyzer only see's frequency vs. level, regardless of phase? Correct me if I'm wrong, or out in space, or something... Put a 20Hz square wave through your system and see what comes out...
As to simming?? I haven't thought of that yet. Unless you modify the parameters, most sim models are pretty ideal... Try it the way it is for starters???
I just know, if you don't sum all the natural hamonics of an instument the way nature does, in natural phase, it won't sound right. A cap will ^&*mess*^$ that up with pahse shift. But, they may all still be there in relative amplitude accuracy, just delayed a little. and it seems a cap is closer to the important frequency range of the higher level stuff... What does the xfrmr do??? I'm thinkin it's phase error is higher up where it might make less difference to our sensitivity???
Hey, I'm only a dog
I am no english man and therefore I don't unterstand exactly all this. However, in generall a not correct phase characteristic (e. g. through reversed polarity only by high frequency speakers or too strong pronounced high pass function) always results in change of a spatial placement of instruments and performers on the imaginary stage (compared to the real stage set of the recording). But there is no influence to the distortion behaviour and sound coloration in the true sense.Well, as you say, by the amplitude response it is below our, so called, audible range. Many designs set roll-off at 2-3 octaves below the 20Hz minimum. But, phase wise, is the music shifted by phase in an area of our sensitivity or not? I believe the tambre of the instruments, vocals etc., the "spectra" is well known to our brain, "sensitivity", and we're very sensitive to phase too. As N.P. describes regarding "absolute phase". Maybe it's more than THD measurements can resolve? Most THD measuring instruments only measure residual after nulling the fundamental. With only one tone, no less, not the summation of many like intruments and vocals. Phase is irrelavent to them. It is only total (true)RMS voltage that was not at the input, in a relative percentage, that they measure. (No math or formulas here). Even a spectrum analyzer only see's frequency vs. level, regardless of phase? Correct me if I'm wrong, or out in space, or something... Put a 20Hz square wave through your system and see what comes out...
As to simming?? I haven't thought of that yet. Unless you modify the parameters, most sim models are pretty ideal... Try it the way it is for starters???
I just know, if you don't sum all the natural hamonics of an instument the way nature does, in natural phase, it won't sound right. A cap will ^&*mess*^$ that up with pahse shift. But, they may all still be there in relative amplitude accuracy, just delayed a little. and it seems a cap is closer to the important frequency range of the higher level stuff... What does the xfrmr do??? I'm thinkin it's phase error is higher up where it might make less difference to our sensitivity???
Hey, I'm only a dog
Additional some users forget, that they don't hear the phase characteristic of one component like a couple capacitor.
It is always to hear the phase characteristic of the whole audio system include the complete recording set up.
Last edited:
"proper absolute phase" redux
I am just a cowboy so I don't really understand this stuff either. Let me ask about the role of phase in monaural reproduction. Suppose that while seated around the campfire I plucked a string on the gueetar. Imagine that the first movement of my tympanic membrane was inward due to a leading compression wave. To playback a recording of the event, should the speaker cone first move outward to create a leading compression wave before rarefaction? Thus, the first movement of my tympanic membrane would be inward. The speaker is in phase with the original event as described. Reversing the speaker leads would cause my eardrum to first move outward on a leading rarefaction wave. The tone would ultimately be there put the transient impact may soften. If I prefer to listen to reproductions 180 degrees out of phase that is one thing, but it ain't aesthetic.
I am just a cowboy so I don't really understand this stuff either. Let me ask about the role of phase in monaural reproduction. Suppose that while seated around the campfire I plucked a string on the gueetar. Imagine that the first movement of my tympanic membrane was inward due to a leading compression wave. To playback a recording of the event, should the speaker cone first move outward to create a leading compression wave before rarefaction? Thus, the first movement of my tympanic membrane would be inward. The speaker is in phase with the original event as described. Reversing the speaker leads would cause my eardrum to first move outward on a leading rarefaction wave. The tone would ultimately be there put the transient impact may soften. If I prefer to listen to reproductions 180 degrees out of phase that is one thing, but it ain't aesthetic.
Lonesome on the trail... Sorry dude,,, haven't heard that in years... And turn me around... Roll me over... Sorry bud most of the cowboy's I've met ride quads???
There are to many not following me... remember "Put a 20Hz square wave through your system and see what comes out..." What happened???
Did anyone try that???
Granted, I expect Neslon to straiten me out, or others to contradict, but that's not happenening... I don't have the best understanding or ability to teach, but, I'll give it another try.
It's Not that an amplifier inverts or not. That would be 180degrees or 0 degrees. It's that, anytime the signal goes through a cap or inductor it will be delayed or whatever a small amount(probabably the oposite with an inductor, depending on the topology?). At 20Hz, the descent coupling cap does not cause to much phase shift but it does cuase a little. What is to much??? At 2.5Hz it causes alot of phase shift, but that's unhearable right??? What happens at 50 Hz? What happens at 100 Hz. There is a slow progression to being back on phase. Even at 200 or 400Hz, and 800Hz it's got phase shift, it's just a minute amount. When does it become infetesumly small to the point the human ear(brain) can not detect it??? We are moving in octaves here(in my explaination). That was a substantial range I just supposed we would have phase shift.
If I were explaining the high end I would say something like... What happens at 5kHz, 10 kHz, 20kHz and we're done... Or are we???
There is this logrithmic pahse shift associated with these components that, granted is small, with the typical design criteria, but is is small enough? Do we need 100uF input caps and 100,000uF output caps to avoid their phase shift and "sound"???
Nelson???
Last edited:
Oh, I did not finish my rambling explaination... An artififact of this evening, no doubt... Any "sound" that is not a pure sine wave, and most music isn't, has many harmonics(multiples of the fundemental frequency). If the harmonic content is not in absolute phase and amplitude to the original, it could be considered to be distorted. If the signal changes phase over the frequency band from say 20Hz to 800Hz it's likely distorted, amongst other things... Would we notice the phase difference of these few octaves where most of the music really is vs. the harmonics in the 5K, 10K or 20K range. And all this might be Caused by the 10uF coupling cap that is supposed to roll sheet off even below that???
So, my question is; does the xfrmer do the opposite???
GiGi want's to know
So, my question is; does the xfrmer do the opposite???
GiGi want's to know
Oh, I did not finish my rambling explaination... An artififact of this evening, no doubt... Any "sound" that is not a pure sine wave, and most music isn't, has many harmonics(multiples of the fundemental frequency). If the harmonic content is not in absolute phase and amplitude to the original, it could be considered to be distorted. If the signal changes phase over the frequency band from say 20Hz to 800Hz it's likely distorted, amongst other things... Would we notice the phase difference of these few octaves where most of the music really is vs. the harmonics in the 5K, 10K or 20K range. And all this might be Caused by the 10uF coupling cap that is supposed to roll sheet off even below that???
So, my question is; does the xfrmer do the opposite???
GiGi want's to know
Believe it or not; I understand fully what you are asking. I just hope that someone can chime in and respond with thoughts commensurate to your efforts. I simply do not have the knowledge to do so. There is one thing I would like to make perfectly clear nonetheless; my dreams are filled with images of cowgirls. Seriously, honestly, I swear this to be true.
6N50D2 little amp
Here is my afternoon working amp. I got rid of some of the jumpers.
100uf high pass's the output.
I still have not figured out the bias with the light bulb.
The resistance is 4R cold standalone.
I measure 1.25A with MultiMeter with 1 lightbulb, and 2.6A with 2 of them in parallel. Sounds better with two.
Here is my afternoon working amp. I got rid of some of the jumpers.
100uf high pass's the output.
I still have not figured out the bias with the light bulb.
The resistance is 4R cold standalone.
I measure 1.25A with MultiMeter with 1 lightbulb, and 2.6A with 2 of them in parallel. Sounds better with two.
Attachments
E-Lite with SemiSouth R100 JFET
I tried this time using the SemiSouth JFET instead of the depletion JFET, using the advice of nelson on the De-Lux De-Lite page 6. Wired up again on a RS perf board.
We can do much the same for the IXTH6N50D2, and Figure 10 shows the circuit of Figure 7 at 100V DC and 2.2 A. This requires a Gate voltage in between the Source and Ground voltage, so we adjust the input bias with a potentiometer as shown. Typically a 10K pot will end up being adjusted for about 25V DC at the Drain of the transistor. If you want to use an enhancement mode part (an IRFP240 for example) you can obtain the positive Gate voltage needed by tapping the potentiometer to the Drain of the transistor instead of the Source.
I gave it 14v from Drain to ground using a 50V supply.
I also got some Lowe's 300w lightbulbs (which measure 2.2R cold) now so I can get around 1.6A with one bulb.
The sound is good, and different than the D-Fet's. The SS JFet's, similiar to what I hear with an F2Jfet clone is similiar, very clean upper mids and highs. I think the pot here helps, as I was previously getting more distortion in the circuit w/o it.
Sound wise, if I compare it to tubes, the D-Fet sounds more like a 2a3, the Jfet more like a 45. Of course, those are based on fading memories.
Both have a certain punch and flow that make the circuit sound single ended and dynamic.
I tried this time using the SemiSouth JFET instead of the depletion JFET, using the advice of nelson on the De-Lux De-Lite page 6. Wired up again on a RS perf board.
We can do much the same for the IXTH6N50D2, and Figure 10 shows the circuit of Figure 7 at 100V DC and 2.2 A. This requires a Gate voltage in between the Source and Ground voltage, so we adjust the input bias with a potentiometer as shown. Typically a 10K pot will end up being adjusted for about 25V DC at the Drain of the transistor. If you want to use an enhancement mode part (an IRFP240 for example) you can obtain the positive Gate voltage needed by tapping the potentiometer to the Drain of the transistor instead of the Source.
I gave it 14v from Drain to ground using a 50V supply.
I also got some Lowe's 300w lightbulbs (which measure 2.2R cold) now so I can get around 1.6A with one bulb.
The sound is good, and different than the D-Fet's. The SS JFet's, similiar to what I hear with an F2Jfet clone is similiar, very clean upper mids and highs. I think the pot here helps, as I was previously getting more distortion in the circuit w/o it.
Sound wise, if I compare it to tubes, the D-Fet sounds more like a 2a3, the Jfet more like a 45. Of course, those are based on fading memories.
Both have a certain punch and flow that make the circuit sound single ended and dynamic.
Attachments
I am not familiar with tubes and therefore I don't understand this. Is a "2a3" and "45" a certainly tube number or amp model number?
Otherwise great work
