Im having trouble determining a set in stone answer for this with the search when it comes to power amps, and ive already made some poor decisions through making assumptions.
If im running signal into a single ended line stage, then from that into a single ended power stage, is my assumption that that phase will invert in each stage, and hit the drivers correctly with no need for follower stages or phase inverters accurate?
If im running signal into a single ended line stage, then from that into a single ended power stage, is my assumption that that phase will invert in each stage, and hit the drivers correctly with no need for follower stages or phase inverters accurate?
I suppose you are assuming that "absolute polarity" has been maintained through the entire recording chain from the microphone, recording console, amplifiers, and every other step in the chain including ADC and DAC. In fact, there is a 50/50 chance that a compression wave recorded by the microphone will result in a compression wave emerging from your speaker.
In a simple system like this, if both channels are fed from the same source, then there is no incorrect phase output to the speakers. They don't care what the change may be from the original input because they can't see it to get all upset about.... But if you are building an amp and need the correct phase at the output transformer secondary to run feedback to the driver section, then you need the OPT to be connected in the correct phase to send NFB forward. The OPT can also be a phase flipper.
I have a sealed 1 cu ft subwoofer enclosure, and on certain tracks i listen to i can absolutely tell the difference when i wire my speakers the other way around. Mainly in the sub. Not on every track, but a good number of them. That actually has nothing to do with my question though even if part of the curiosity. I tend to type, a lot, and was trying to be as simple as possible. I think my question was slightly misunderstood. It's from a perspective of just trying to learn about something im planning on fooling around with, not fix anything i feel is wrong or something like that.
My question in a better worded form would be:
Will a single ended power stage feeding an output transformer invert the phase of the incoming signal at it's output the same as a single gain stage in a preamp?
As far as a single ended output stage of amplification goes, i know it will, but i wasn't sure how the transformer would affect it. And thanks for that tidbit because i wasn't even considering how global feedback or something would factor into this
And i would like the correct phasing to be hitting my drivers. Particularly my sealed sub. i listen to a good bit of electronic music and can absolutely 100% tell when my* system is out of phase. Although this build is going to be feeding bookshelf speakers for a turntable. no sub, so that part does become pretty irrelevant.
My question in a better worded form would be:
Will a single ended power stage feeding an output transformer invert the phase of the incoming signal at it's output the same as a single gain stage in a preamp?
As far as a single ended output stage of amplification goes, i know it will, but i wasn't sure how the transformer would affect it. And thanks for that tidbit because i wasn't even considering how global feedback or something would factor into this
And i would like the correct phasing to be hitting my drivers. Particularly my sealed sub. i listen to a good bit of electronic music and can absolutely 100% tell when my* system is out of phase. Although this build is going to be feeding bookshelf speakers for a turntable. no sub, so that part does become pretty irrelevant.
Last edited:
Yes, most single ended amplifiers are inverting (except for buffers).
If your circuits have voltage gain, they are likely inverting, so two in series would be non-inverting.
However, a single ended stage with a transformer load could be either way, depending on how
the transformer is wired. I would check it with a scope to be sure.
If your circuits have voltage gain, they are likely inverting, so two in series would be non-inverting.
However, a single ended stage with a transformer load could be either way, depending on how
the transformer is wired. I would check it with a scope to be sure.
If you are sending signals to the sub ahead of the amp and using full range main speakers then you could very well have some bass phase cancellation issues. Sho'nuff.
The particular system i was mentioning is actually currently fed by a 2.1 3255 chip amp with a tube preamp. Has 2 stereo chips, one bridged and low passed for the sub. And audibly i can only really detect phase cancellation right up with my head practically splitting the difference between them and the sub. Honestly and truly its the best integrating sub into a sounstage i've heard or been able to put together personally. My particular phase issues that need resolving on that setup come where my mid range driver crosses with the woofer. But that is being sorted out soon (when i can afford the crossover components i need).
I chose woofers that naturally rolled off at about 100hz, and cross the sub over at the same. Same brand/series of drivers being fed by same chip in parallel. The integration is on point as ive ever heard it. And i tweeked the preamps negative feedback to soften the sharp high end on the class D setup. It's actually incredibly musical. And listenable for very long periods.
As far as this build goes the phase at speakers wont really matter, will just be feeding a set of bookshelf speakers. although i did just obtain my first working scope this past weekend. But yea im a complete novice with amplifier design and implementation. 100%
i will likely be implementing global feedback with this build though. And have realized thats where my phasing concerns should be lol.
I chose woofers that naturally rolled off at about 100hz, and cross the sub over at the same. Same brand/series of drivers being fed by same chip in parallel. The integration is on point as ive ever heard it. And i tweeked the preamps negative feedback to soften the sharp high end on the class D setup. It's actually incredibly musical. And listenable for very long periods.
As far as this build goes the phase at speakers wont really matter, will just be feeding a set of bookshelf speakers. although i did just obtain my first working scope this past weekend. But yea im a complete novice with amplifier design and implementation. 100%
i will likely be implementing global feedback with this build though. And have realized thats where my phasing concerns should be lol.
Last edited:
A two stage single ended amplifier:
Start with the mid frequencies (there may, or may not, be phase shifts at low frequencies, high frequencies, or both. To be discussed later in this post [***]).
Many 2 stage SE amps have this topology . . .
Common Cathode input stage, input connector to volume control to input tube grid, to input tube plate - that is inverted at mid frequencies.
Input tube plate to RC coupling to output tube grid - the input tube's plate phase is preserved at mid frequencies (sum total is inverted).
One inversion (1)
Output tube grid to output tube plate is inverted at mid frequencies.
So far, the sum total of the input signal at the volume control all the way to the output tube plate is non-inverted at mid frequencies.
Now, we have two 'serial' inversions (2) That is non-inverting.
If the output tube plate is connected to the primary connection that is in-phase with the secondary output tap [8 Ohm tap for example]
(not the secondary common / ground tap), then the output tube plate to output transformer secondary tap is not further inverted.
No additional inversions,
So, the sum total is non-inverted from input connector/volume control, all the way to the secondary output tap, at mid frequencies.
[***]
With a sum total of non-inverting input connector to output tap, at mid frequencies, what happens at low frequencies, and high frequencies?
The phase at low frequencies is changed mostly by:
The RC coupling network between the input tube and the output tube
The inductance of the output transformer primary
These are all Leading Phase at low frequency, versus at mid frequency
The phase at high frequencies is changed mostly by:
The input stage capacitance, especially the Miller Effect capacitance of the input tube
The output stage capacitance, especially the Miller Effect capacitance of the output tube
The distributed capacitance ot the output transformer primary
The leakage inductance from the output transformer primary to its secondary
These are all Lagging Phase at high frequency, versus at mid frequency
In some cases these leading and lagging phases are very small, and sometimes are fairly large from 20Hz to 20kHz. But for two identical mono-block or identical channels of a stereo amplifier, they are the same from channel to channel.
With the exception of the bass leading phase of the amplifier, versus the phase of the sub woofer, do not worry about the mid frequencies (already non-inverting).
An interstage transformer between input stage and output stage, instead of RC coupling, may be connected inverting or non-inverting; but the phases at low frequencies and phases at high frequencies will be affected, more or less, according to the quality of the interstage transformer.
Input transformers, interstage transformers, and output transformers can be connected to be non-inverting; and they can be connected to be Inverting too. That can be good if the number of stages in the amplifier needs that extra inversion (like 3 stages instead of 2 stages); but if you connect it incorrectly, you created a problem.
Even in a system without a subwoofer, the identical state of inverted or non-inverted mid frequencies, and high and low frequency phases, of the two channels from input connector all the way to the loudspeakers needs to be identical (do not swap the leads of one and only one loudspeaker cable, versus the other loudspeaker cable).
A low pass filter that is used when there is a subwoofer, does cause a lagging phase. The degree of phase shift, and as it varies versus frequency is what is important.
And, if a high pass filter is used to keep low frequencies out of the amplifiers and loudspeakers, that causes a leading phase shift, that varies versus frequency.
One question to ask yourself, is what is the phase of the main loudspeaker woofers at 100Hz, if they naturally roll off there, the phase is different than it is at mid frequencies.
Crossover design time . . .
I am about ready to bring my subwoofer upstairs and use it in my latest stereo system.
Wish me luck in getting it to work well, even if not totally perfect.
I sympathize with the difficulties of getting your system right.
Have Fun!
Start with the mid frequencies (there may, or may not, be phase shifts at low frequencies, high frequencies, or both. To be discussed later in this post [***]).
Many 2 stage SE amps have this topology . . .
Common Cathode input stage, input connector to volume control to input tube grid, to input tube plate - that is inverted at mid frequencies.
Input tube plate to RC coupling to output tube grid - the input tube's plate phase is preserved at mid frequencies (sum total is inverted).
One inversion (1)
Output tube grid to output tube plate is inverted at mid frequencies.
So far, the sum total of the input signal at the volume control all the way to the output tube plate is non-inverted at mid frequencies.
Now, we have two 'serial' inversions (2) That is non-inverting.
If the output tube plate is connected to the primary connection that is in-phase with the secondary output tap [8 Ohm tap for example]
(not the secondary common / ground tap), then the output tube plate to output transformer secondary tap is not further inverted.
No additional inversions,
So, the sum total is non-inverted from input connector/volume control, all the way to the secondary output tap, at mid frequencies.
[***]
With a sum total of non-inverting input connector to output tap, at mid frequencies, what happens at low frequencies, and high frequencies?
The phase at low frequencies is changed mostly by:
The RC coupling network between the input tube and the output tube
The inductance of the output transformer primary
These are all Leading Phase at low frequency, versus at mid frequency
The phase at high frequencies is changed mostly by:
The input stage capacitance, especially the Miller Effect capacitance of the input tube
The output stage capacitance, especially the Miller Effect capacitance of the output tube
The distributed capacitance ot the output transformer primary
The leakage inductance from the output transformer primary to its secondary
These are all Lagging Phase at high frequency, versus at mid frequency
In some cases these leading and lagging phases are very small, and sometimes are fairly large from 20Hz to 20kHz. But for two identical mono-block or identical channels of a stereo amplifier, they are the same from channel to channel.
With the exception of the bass leading phase of the amplifier, versus the phase of the sub woofer, do not worry about the mid frequencies (already non-inverting).
An interstage transformer between input stage and output stage, instead of RC coupling, may be connected inverting or non-inverting; but the phases at low frequencies and phases at high frequencies will be affected, more or less, according to the quality of the interstage transformer.
Input transformers, interstage transformers, and output transformers can be connected to be non-inverting; and they can be connected to be Inverting too. That can be good if the number of stages in the amplifier needs that extra inversion (like 3 stages instead of 2 stages); but if you connect it incorrectly, you created a problem.
Even in a system without a subwoofer, the identical state of inverted or non-inverted mid frequencies, and high and low frequency phases, of the two channels from input connector all the way to the loudspeakers needs to be identical (do not swap the leads of one and only one loudspeaker cable, versus the other loudspeaker cable).
A low pass filter that is used when there is a subwoofer, does cause a lagging phase. The degree of phase shift, and as it varies versus frequency is what is important.
And, if a high pass filter is used to keep low frequencies out of the amplifiers and loudspeakers, that causes a leading phase shift, that varies versus frequency.
One question to ask yourself, is what is the phase of the main loudspeaker woofers at 100Hz, if they naturally roll off there, the phase is different than it is at mid frequencies.
Crossover design time . . .
I am about ready to bring my subwoofer upstairs and use it in my latest stereo system.
Wish me luck in getting it to work well, even if not totally perfect.
I sympathize with the difficulties of getting your system right.
Have Fun!
Last edited:
Absolute polarity of a recording system plus a playback system is a completely more complex and different discussion.
My favorite example is an Orchestral Thunder Drum [***].
If the microphone is placed on the struck side of the drum, the initial polarity is negative (rarefaction of air).
If the microphone is placed on the other side (not the struck side) the initial polarity is positive (compression of air).
Compare that to any microphone placed anywhere near enough to pick up a Trombone . . . the initial polarity is always positive (compression of air).
(At least I am not aware of any recordings of a Trombone where the microphone was placed inside the Trombone player's mouth).
The book ("The Wood Effect") discusses absolute phase in great detail.
But it might, or might not, take into account, or mentioned the example above [***].
My favorite example is an Orchestral Thunder Drum [***].
If the microphone is placed on the struck side of the drum, the initial polarity is negative (rarefaction of air).
If the microphone is placed on the other side (not the struck side) the initial polarity is positive (compression of air).
Compare that to any microphone placed anywhere near enough to pick up a Trombone . . . the initial polarity is always positive (compression of air).
(At least I am not aware of any recordings of a Trombone where the microphone was placed inside the Trombone player's mouth).
The book ("The Wood Effect") discusses absolute phase in great detail.
But it might, or might not, take into account, or mentioned the example above [***].
Yet again thank you for the knowledge dump that i will likely be referencing for a bit to come, @6A3sUMMER
i have been spending a good bit of time in xsim now. As far as i know this chip amp, unlike the fosi ones doesn't seem to have a high pass filter on the main chip, just a low pass on the bridged chip. Which in most cases i wouldnt prefer but the inexpensive drivers i was looking at started to roll off around 100hz anyway. I just made frd and zma files from the impedance and freq response charts for the drivers and used those in xsim for the current crossover design.
I havent got into simulating my boxes yet, but i have already significantly improved my crossovers. as well as others lol, and more improvements to come when im done tinkering with amps for a bit. Im a pretty one track obsession mind.
As far as the sub integrating into the soundstage of the system goes, in person i haven't heard one that integrates better. However it seems i am the only music lover in my extended personal life that has even taken notice that a stereo soundstage is even a possibility. Subs that aren't mine tend to just be bouncing around all boomy and in your face out of a corner somehwere in my personal life lol. To my ears this is pretty spot as i can really get it there, and mainly just there. You dont even notice there's a sub at all unless you turn it off, how it should be imho. I'd like to think i have decent ears having played with the mixing and mastering side of music for quite a few years now. Not legitimate full scale productions or anything. More like DI into an interface. But i've got pretty damn good at delivering solid mixes that translate well between systems. As soon as i heard the first iteration of these speakers i knew it was off and went about trying to rectify that. And so far it's been a year long journey into crossover and then amplifier design. And the more i learn the more i realize i have to learn.
I definitely prefer running a sub from an identical parallel bridged chip vs running it from high level outs on my old yamaha to the high level ins on a plate amp sub. And i'm very excited about the future of these chip amps as power amplifiers. The subwoofer integration with this kind of amp is truly fantastic imo. In my limited experience, i've neve heard anything like it.
The weak point in the main system is my mid range driver choice but i purchased for these boxes not having any idea what i was doing. Much like my father before me hahaha. I just got all drivers (minus tweets, a mistake since rectified) from same brand and series. And threw an off the shelf crossover that was the right ohm rating and had xover points that seemed like they'd play nice. I've wittled down to a few small full range drivers i'll likely end up choosing between. I just havent been playing with the speaker side of things much recently. And im referencing results of tweaks against a set of active studio monitors because as of yet, i do not have a way to test my response in the room lol.
And i will absolutely look into this book.
i have been spending a good bit of time in xsim now. As far as i know this chip amp, unlike the fosi ones doesn't seem to have a high pass filter on the main chip, just a low pass on the bridged chip. Which in most cases i wouldnt prefer but the inexpensive drivers i was looking at started to roll off around 100hz anyway. I just made frd and zma files from the impedance and freq response charts for the drivers and used those in xsim for the current crossover design.
I havent got into simulating my boxes yet, but i have already significantly improved my crossovers. as well as others lol, and more improvements to come when im done tinkering with amps for a bit. Im a pretty one track obsession mind.
As far as the sub integrating into the soundstage of the system goes, in person i haven't heard one that integrates better. However it seems i am the only music lover in my extended personal life that has even taken notice that a stereo soundstage is even a possibility. Subs that aren't mine tend to just be bouncing around all boomy and in your face out of a corner somehwere in my personal life lol. To my ears this is pretty spot as i can really get it there, and mainly just there. You dont even notice there's a sub at all unless you turn it off, how it should be imho. I'd like to think i have decent ears having played with the mixing and mastering side of music for quite a few years now. Not legitimate full scale productions or anything. More like DI into an interface. But i've got pretty damn good at delivering solid mixes that translate well between systems. As soon as i heard the first iteration of these speakers i knew it was off and went about trying to rectify that. And so far it's been a year long journey into crossover and then amplifier design. And the more i learn the more i realize i have to learn.
I definitely prefer running a sub from an identical parallel bridged chip vs running it from high level outs on my old yamaha to the high level ins on a plate amp sub. And i'm very excited about the future of these chip amps as power amplifiers. The subwoofer integration with this kind of amp is truly fantastic imo. In my limited experience, i've neve heard anything like it.
The weak point in the main system is my mid range driver choice but i purchased for these boxes not having any idea what i was doing. Much like my father before me hahaha. I just got all drivers (minus tweets, a mistake since rectified) from same brand and series. And threw an off the shelf crossover that was the right ohm rating and had xover points that seemed like they'd play nice. I've wittled down to a few small full range drivers i'll likely end up choosing between. I just havent been playing with the speaker side of things much recently. And im referencing results of tweaks against a set of active studio monitors because as of yet, i do not have a way to test my response in the room lol.
And i will absolutely look into this book.
Last edited:
it is in certain contexts that pretty much exclusively include the few electronic artists i listen to frequently, and really only when a sealed subwoofer enclosure is involved can i notice a difference at all. And even then, only on my own system with tracks ive become incredibly familiar with. Parts of the low end dont sit in with the soundstage the way i've com to expect it to on this setup. 99% of the time if you flipped the wires on me i wouldn't notice anything at all. And even if i did, my brain wouldn't assume phasing first thing i dont think.
Phasing won't actually be much concern with this build though, beyond its effects on global feedback i dont think. It's just going to feed a couple small ported bookshelf speakers by a turntable. Probably a parts express kit build. i highly doubt even once familiar for a time, and if i played the bassy electronic tracks in question would i even be able to tell its inverted or not. Id assume it impossible.
Phasing won't actually be much concern with this build though, beyond its effects on global feedback i dont think. It's just going to feed a couple small ported bookshelf speakers by a turntable. Probably a parts express kit build. i highly doubt even once familiar for a time, and if i played the bassy electronic tracks in question would i even be able to tell its inverted or not. Id assume it impossible.
Last edited:
As far as the creation of music side of things goes, as an armchair bedroom "producer" phasing doesn't matter a whole lot to me beyond using phase inversion to help multiple goings ons in the mid bands feel more distinct and separate from one another. If i like how an acoustic guitar is EQ'd for instance, but want it to pop and feel a little bit forward of the mix itself, ill invert it's phase, and more often then not i get what im after. stereo Bass synth and messing with phase alignment can be real interesting lol.
Fun to play with partial phase shifts on stereo panned instruments to get feelings of movement outside of the soundstage.
Im a big fan of in-daw phase null tests for lots of reasons. Until recently thats all that phase mattered to me. (Although, now having had the experience of polyester and polystyrene feedback caps testing exactly the same in audio analyzer suite. But sounding different to more then just my ears demonstrably, and basically my entire experience so far with tubes, i am tending to lend less credit to a wav vs flac phase null test then i once did)
Fun to play with partial phase shifts on stereo panned instruments to get feelings of movement outside of the soundstage.
Im a big fan of in-daw phase null tests for lots of reasons. Until recently thats all that phase mattered to me. (Although, now having had the experience of polyester and polystyrene feedback caps testing exactly the same in audio analyzer suite. But sounding different to more then just my ears demonstrably, and basically my entire experience so far with tubes, i am tending to lend less credit to a wav vs flac phase null test then i once did)
Last edited:
This discussion is continually conflating (and confusing) phase and polarity, and in the frequency region and application where it might actually matter. To progress, we must begin by separating these similar sounding but unrelated terms. Lots and lots of confusion seems to arise from this issue.
All good fortune,
Chris
All good fortune,
Chris
Take a pair of two way speakers, and either a stereo amplifier, or a pair of mono-block amplifiers.
Find a CD recording that has the Bass player in the Left Channel.
Use an RCA plug splitter, and send the Left Channel signal from the CD player to Both the Left channel amplifier input, And Right channel amplifier input.
Put the two way speakers side by side. Listen to the recording, especially to the Bass instrument.
The Bass is present.
Swap the speaker leads on one end, and only one end . . . of one, and only one, loudspeaker cable. Listen to the recording again.
The Bass is Gone!
In-phase with bass
Out-of-Phase without Bass
a Simple explanation of phase.
a Simple way to hear the results of phase similarity, and of phase dis-similarity.
Have Fun!
Find a CD recording that has the Bass player in the Left Channel.
Use an RCA plug splitter, and send the Left Channel signal from the CD player to Both the Left channel amplifier input, And Right channel amplifier input.
Put the two way speakers side by side. Listen to the recording, especially to the Bass instrument.
The Bass is present.
Swap the speaker leads on one end, and only one end . . . of one, and only one, loudspeaker cable. Listen to the recording again.
The Bass is Gone!
In-phase with bass
Out-of-Phase without Bass
a Simple explanation of phase.
a Simple way to hear the results of phase similarity, and of phase dis-similarity.
Have Fun!
I've mentioned this before. Android has an app that injects a signal into your system via normal inputs and then "hears" the click as being in-phase or out. It has different frequency ranges for checking LF/HF and mid-range. makes it easy to check the whole system at once. Called Polarity Checker
Start with a simple stereo system, 2 speaker systems.
Reverse the polarity of one speaker (swap the speaker cable leads at one channel of the amplifier output).
Play a stereo recording.
You will hear sound that appears to be at or outside of the speaker positions; and there will be very little sound in the center.
If that does not work very well, try playing a mono recording.
If it sounds better than it was when both speakers were connected 'properly', it sounds like your original setup was out-of phase.
Reverse the polarity of one speaker (swap the speaker cable leads at one channel of the amplifier output).
Play a stereo recording.
You will hear sound that appears to be at or outside of the speaker positions; and there will be very little sound in the center.
If that does not work very well, try playing a mono recording.
If it sounds better than it was when both speakers were connected 'properly', it sounds like your original setup was out-of phase.