I've been meaning to do this for a long time. The basic circuit is two (valve) SRPP stages followed by a source follower MOSFET stage.
1. Voltage series (VSFB)feedback from output to the cathode of the 1st SRPP stage.
2. Voltage shunt (VHFB) feedback in first SRPP stage and voltage shunt feedback from output to input of 2nd SRPP stage. This splits the amp into two sections with no global negative feedback like the voltage series configuration.
The listening test: Transients in the VSFB was better than in VHFB . Much tighter .
The bass was also tighter in VSFB vs VHFB. Deep bass was present but better defined in VSFB.
Voice. Voice was noisy in VSFB as compared to VHFB where the voices were VERY good. Treble was a little laid back in VHFB. The overall sound was much cleaner and sounds spaced apart in VHFB. Bass drum was 'meatier' in VHFB. Traansients were not as dramatic in VHFB.
VHFB was easier on the ear and more musical. Voices were very real and upfront. The speakers kind of disappeared. But VSFB had punch and impact in comparison with VHFB ( good for rock?).
So what do I choose? The VSFB for the bass end (sub) and VHFB for the mids and everything above. I do not mind missing some punch for a more musical sound. VHFB sounds "bassier" than the VSFB. That could be because of less control over the speaker due to higher output impedance at LF.
Anyone out there with a similar or a different experience?
Cheers.
1. Voltage series (VSFB)feedback from output to the cathode of the 1st SRPP stage.
2. Voltage shunt (VHFB) feedback in first SRPP stage and voltage shunt feedback from output to input of 2nd SRPP stage. This splits the amp into two sections with no global negative feedback like the voltage series configuration.
The listening test: Transients in the VSFB was better than in VHFB . Much tighter .
The bass was also tighter in VSFB vs VHFB. Deep bass was present but better defined in VSFB.
Voice. Voice was noisy in VSFB as compared to VHFB where the voices were VERY good. Treble was a little laid back in VHFB. The overall sound was much cleaner and sounds spaced apart in VHFB. Bass drum was 'meatier' in VHFB. Traansients were not as dramatic in VHFB.
VHFB was easier on the ear and more musical. Voices were very real and upfront. The speakers kind of disappeared. But VSFB had punch and impact in comparison with VHFB ( good for rock?).
So what do I choose? The VSFB for the bass end (sub) and VHFB for the mids and everything above. I do not mind missing some punch for a more musical sound. VHFB sounds "bassier" than the VSFB. That could be because of less control over the speaker due to higher output impedance at LF.
Anyone out there with a similar or a different experience?
Cheers.
No Feedback test
Hi Peter,
I did try the no feedback scheme with one input SRPP reduced.
My conclusion is that no NFB is certainly the best sounding for 'voice'. There is no two ways about this. But the high output impedance is a problem as it makes the bass flabby and bloated.
The HF is also nice but does not have as much bite as in Voltage Series feedback. If one had to choose "only one circuit" I would go for Voltage shunt as the voice is pretty close to the no feedback version. If you want to be a purist and do not mind the bass because you have a separate sub , then no feedback for 100Hz and up should be good if not very good.
Cheers,
Ashok.
Voltage Series feedback HAS to be the choice of "rockers". The transient attack and taught deep bass is incredible.
Hi Peter,
I did try the no feedback scheme with one input SRPP reduced.
My conclusion is that no NFB is certainly the best sounding for 'voice'. There is no two ways about this. But the high output impedance is a problem as it makes the bass flabby and bloated.
The HF is also nice but does not have as much bite as in Voltage Series feedback. If one had to choose "only one circuit" I would go for Voltage shunt as the voice is pretty close to the no feedback version. If you want to be a purist and do not mind the bass because you have a separate sub , then no feedback for 100Hz and up should be good if not very good.
Cheers,
Ashok.
Voltage Series feedback HAS to be the choice of "rockers". The transient attack and taught deep bass is incredible.
...Interesting...
With solid state Class A amplifiers I ended up with a series regulated supply feeding a shunt regulated supply for best sound.
These amps were IT coupled between input/voltage gain stages and driver/output stages with no global feedback. They were also fully balanced designs.
I've been playing with the two options and combinations with a valve amplifier and haven't reached conclusions yet as my passive supply sounds better than either regulated supply or one reg feeding the other... From which I conclude that I need to do more work on my regulators...
ciao
James
With solid state Class A amplifiers I ended up with a series regulated supply feeding a shunt regulated supply for best sound.
These amps were IT coupled between input/voltage gain stages and driver/output stages with no global feedback. They were also fully balanced designs.
I've been playing with the two options and combinations with a valve amplifier and haven't reached conclusions yet as my passive supply sounds better than either regulated supply or one reg feeding the other... From which I conclude that I need to do more work on my regulators...
ciao
James
Long term burn in.
Hi,
I was beginning to think that the first few minutes / tracks of music that we know well are the most important .
Since we have ears that tend to get tuned to the sound if we listen long enough , we could end up getting used to the 'new' sound - good or bad.
So long term burn in may often be 'burn in for our ears'. So a quick test - no more than a couple of hours should tell us if we hear a difference. If we cannot then I think maybe there is no difference at all. The mind is great at playing tricks , especially if we want it to !
Cheers.
Hi,
I was beginning to think that the first few minutes / tracks of music that we know well are the most important .
Since we have ears that tend to get tuned to the sound if we listen long enough , we could end up getting used to the 'new' sound - good or bad.
So long term burn in may often be 'burn in for our ears'. So a quick test - no more than a couple of hours should tell us if we hear a difference. If we cannot then I think maybe there is no difference at all. The mind is great at playing tricks , especially if we want it to !
Cheers.
FEEDBACK.
Hi,
Thanks, Jam.I'm sure I am not the only one around here though.
With regard to feedback;I try to avoid global NFB as much as possible and try to make each stage as linear as it needs to be before applying any kind of feedback to it.
As you all know I am pretty much ignorant about semi-conductors but with most triode valves you can use local (degenerative FB) at the expense of Zo.
Not many people are aware of it but a triode has FB built in already and if you operate it in its most linear area you often don't need feedback.
Other techniques such as distortion cancellation techniques can be used.
As James has pointed to regulation of the PSU, this too plays a very important role in the sound of a circuit...especially with the SRPP types which are senstive to the PSU.
Quite likely because the don't really shunt regulate all that well in their standard form.
So, as far I can tell, I wouldn't make general conclusions although I do tend to agree with Ashok's findings, generally speaking.
Cheers,
Hi,
Thanks, Jam.I'm sure I am not the only one around here though.
With regard to feedback;I try to avoid global NFB as much as possible and try to make each stage as linear as it needs to be before applying any kind of feedback to it.
As you all know I am pretty much ignorant about semi-conductors but with most triode valves you can use local (degenerative FB) at the expense of Zo.
Not many people are aware of it but a triode has FB built in already and if you operate it in its most linear area you often don't need feedback.
Other techniques such as distortion cancellation techniques can be used.
As James has pointed to regulation of the PSU, this too plays a very important role in the sound of a circuit...especially with the SRPP types which are senstive to the PSU.
Quite likely because the don't really shunt regulate all that well in their standard form.
So, as far I can tell, I wouldn't make general conclusions although I do tend to agree with Ashok's findings, generally speaking.
Cheers,
Series or shunt feedback!
I prefer the sound of series feedback.!
I have expressed the same feelings not so many time ago in the solid state forum...the shunt feedback sound to may ears much more dull.
The explanation of this is that in the shunt (inverting) configuration the EMF from the speaker find a way via the feedback resistor to the cd or preamp that is feeding it!..that is the cause of that dinamic intermodulation! And the dull sound.!
Regards
Jorge
PS. people when talk about topologies and distortion allways forget that the amp is feeding a reactif load!
I prefer the sound of series feedback.!
I have expressed the same feelings not so many time ago in the solid state forum...the shunt feedback sound to may ears much more dull.
The explanation of this is that in the shunt (inverting) configuration the EMF from the speaker find a way via the feedback resistor to the cd or preamp that is feeding it!..that is the cause of that dinamic intermodulation! And the dull sound.!
Regards
Jorge
PS. people when talk about topologies and distortion allways forget that the amp is feeding a reactif load!
Frank,
I guess you are not going to give up your favourite pre-amp. schematic? Must be some family secret.
I agree with Mr.Pass that keeping the number of stages in a design to a minimum is important. I guess if you keep the feedback path short the negative effects of feedback are less apparent.
Jam
I guess you are not going to give up your favourite pre-amp. schematic? Must be some family secret.
I agree with Mr.Pass that keeping the number of stages in a design to a minimum is important. I guess if you keep the feedback path short the negative effects of feedback are less apparent.
Jam
Preamp.
Hi Jam,
Italian restaurants don't hand out recipes either...
Still, if I can help...tell me what you need from a preamp: such as do you need phono RIAA correction, line etc...maybe I can help.
It looks as if I can easily persuade you to use tubes then you'd be convering me to the sand camp he?
So do I, better to use local loops than global ones if you must use FB.
Cheers,
Hi Jam,
Italian restaurants don't hand out recipes either...
Still, if I can help...tell me what you need from a preamp: such as do you need phono RIAA correction, line etc...maybe I can help.
It looks as if I can easily persuade you to use tubes then you'd be convering me to the sand camp he?
So do I, better to use local loops than global ones if you must use FB.
Cheers,
Frank,
Ah! Ha! A response. All I need is a schematic if a line stage to drive a transformer to obtain a balanced out. I suppose a reasonably low output impedence is in order. ( your choice of topology ) .
If you worked for Goldman how come you are in the tube camp?
I also find low feedback or circuits with no global feedback are a lot more critical of the power supply. Care to comment?
Jam
Ah! Ha! A response. All I need is a schematic if a line stage to drive a transformer to obtain a balanced out. I suppose a reasonably low output impedence is in order. ( your choice of topology ) .
If you worked for Goldman how come you are in the tube camp?
I also find low feedback or circuits with no global feedback are a lot more critical of the power supply. Care to comment?
Jam
IN A JAM...
Hi,
That's not too hard to do, a single ended stage driving a trannie is rather easy...tell us the gain you need...hold on, you don't mind starting a thread in the tube section for that, do you?
You'll get much more input that way and some advice from me on regulated PSUs to boot...
Well, that's a long story...we were never employed by Goldmund as such...see it as a take over "avant la lettre".
It is an interesting, yet sad story on how some brilliant people lost a fortune not realising the immense opportunity they had...
A fascinating story in high-end history.
As for the tube part...well if you look at the fact that some contributors such as J.C. Verdier, Jean Hiraga and yours truly were tubes guys already before we entered this mechanical engineering task, you may understand it a little better.
If you ever get the chance to buy the "L'Audiophile" collection of magazines in English (I don't know if you read French) you may find it a treasure throve even to this day.
The only thing in the Anglosaxon world that came close was the British mag "Hifi-Answers" under Jimmy Hughes.
Lucky for me I have met all of them in person and maybe one day I'll write about these people for I feel those were really pioneering days.
Cheers,
Hi,
That's not too hard to do, a single ended stage driving a trannie is rather easy...tell us the gain you need...hold on, you don't mind starting a thread in the tube section for that, do you?
You'll get much more input that way and some advice from me on regulated PSUs to boot...
Well, that's a long story...we were never employed by Goldmund as such...see it as a take over "avant la lettre".
It is an interesting, yet sad story on how some brilliant people lost a fortune not realising the immense opportunity they had...
A fascinating story in high-end history.
As for the tube part...well if you look at the fact that some contributors such as J.C. Verdier, Jean Hiraga and yours truly were tubes guys already before we entered this mechanical engineering task, you may understand it a little better.
If you ever get the chance to buy the "L'Audiophile" collection of magazines in English (I don't know if you read French) you may find it a treasure throve even to this day.
The only thing in the Anglosaxon world that came close was the British mag "Hifi-Answers" under Jimmy Hughes.
Lucky for me I have met all of them in person and maybe one day I'll write about these people for I feel those were really pioneering days.
Cheers,
Frank's the Man!
Most excellent idea ...Frank.
I speak no French so where do I purchase copies of the English version of L'Audiophile?
Tubes?.......6DJ8's or 6SN7's perhaps you choose.
I shall call the preamp Franken Tube in honor of it's designer (that being you).
Maybe you can teach us old (solid state) dogs a few new tricks.
Regards,
Jam
Most excellent idea ...Frank.
I speak no French so where do I purchase copies of the English version of L'Audiophile?
Tubes?.......6DJ8's or 6SN7's perhaps you choose.
I shall call the preamp Franken Tube in honor of it's designer (that being you).
Maybe you can teach us old (solid state) dogs a few new tricks.
Regards,
Jam
Frankenstein.
Hi,
Unfortunately other than a few articles on the Kaneda preamp (remember?) I am not aware of translations...maybe my work should start there?
O.K...so an active preamp with some 15dB of amplification feeding a transformer would be your ticket?
I assume you need to feed a balanced amp?
Cheers,
Hi,
Unfortunately other than a few articles on the Kaneda preamp (remember?) I am not aware of translations...maybe my work should start there?
O.K...so an active preamp with some 15dB of amplification feeding a transformer would be your ticket?
I assume you need to feed a balanced amp?
Cheers,
Dull sound?
Hi Jorge,
Like many others have said , I repeat that the actual implementation of a circuit can affect its 'sound'. To make a general statement that shunt feedback is 'dull' would probably not be correct.
My shunt implementations, do not have 'dull' sound but different sound. HF is present just like the series feedback implementation.But it sounds different . It sounds smoother - not really duller!
About the speaker signals ( back emf?) going back into the previous stage : If this 'did' happen it's effect would depend a lot on the output impedance of the previous stage. Besides I was under the impression that significant back emf in a speaker happened significantly at LF.
Note that zero negative feedback circuits can also sound"dull". There is no way the speaker back emf can cause this ( by going back to the input).
So give "shunt' a chance. maye be you will have to spend more time listening to such circuits. Not all will be good implementations. So give it time and have fun. That's what it is all about !
Cheers.
Hi Jorge,
Like many others have said , I repeat that the actual implementation of a circuit can affect its 'sound'. To make a general statement that shunt feedback is 'dull' would probably not be correct.
My shunt implementations, do not have 'dull' sound but different sound. HF is present just like the series feedback implementation.But it sounds different . It sounds smoother - not really duller!
About the speaker signals ( back emf?) going back into the previous stage : If this 'did' happen it's effect would depend a lot on the output impedance of the previous stage. Besides I was under the impression that significant back emf in a speaker happened significantly at LF.
Note that zero negative feedback circuits can also sound"dull". There is no way the speaker back emf can cause this ( by going back to the input).
So give "shunt' a chance. maye be you will have to spend more time listening to such circuits. Not all will be good implementations. So give it time and have fun. That's what it is all about !
Cheers.
The missing link!
Hi ashok!
I have spent much time studing the speaker interface and his relation to sound quality!Is that interface that explain the dramatic diferences betwen amps!
When i talk about dull sound i don't mean that there are no highs...but more a sensation of intermodulation.
But you don't need to belive in me ...try this:
In a valve amp suitch the ground reference of the output transvormer for the amp became inverting...then conect the feedback resistor to the grid of the first tube.
After this conect another amp via a 8 oHms resistor to the output.Put same music or sinusoide in this second amp and put the volume control at a medium level.
Then you can see at the grid of the first tube the sinusoide or the music that the second amp is feeding.(see with a osciloscope).
That second amp simulate the EMF of the speaker.
Is obvius that that voltage that apears at the input grid will enter in the preamp or cd...even if the ouput impedance of the preamp is low,but the open loop impedance is not! then this parasitic voltage will be processed!
Better than listening to people...is to do some experiences!
Cheers
jorge
Hi ashok!
I have spent much time studing the speaker interface and his relation to sound quality!Is that interface that explain the dramatic diferences betwen amps!
When i talk about dull sound i don't mean that there are no highs...but more a sensation of intermodulation.
But you don't need to belive in me ...try this:
In a valve amp suitch the ground reference of the output transvormer for the amp became inverting...then conect the feedback resistor to the grid of the first tube.
After this conect another amp via a 8 oHms resistor to the output.Put same music or sinusoide in this second amp and put the volume control at a medium level.
Then you can see at the grid of the first tube the sinusoide or the music that the second amp is feeding.(see with a osciloscope).
That second amp simulate the EMF of the speaker.
Is obvius that that voltage that apears at the input grid will enter in the preamp or cd...even if the ouput impedance of the preamp is low,but the open loop impedance is not! then this parasitic voltage will be processed!
Better than listening to people...is to do some experiences!
Cheers
jorge
feedback tests
Hi Jorge,
You are right about the signal being visible at the input. more accurately you will find a current flowing in the feedback resistors caused by the signal being injected at the output of the first amp.
You should also find this happenning in a voltage series feedback circuit - that is current flow in the feedback resistors.
I will think about the effect of this in both cases. It may not be as simple as it looks. Right now I have to run. Lots of work today. I will post again later. Thanks for your views.
Cheers.
Hi Jorge,
You are right about the signal being visible at the input. more accurately you will find a current flowing in the feedback resistors caused by the signal being injected at the output of the first amp.
You should also find this happenning in a voltage series feedback circuit - that is current flow in the feedback resistors.
I will think about the effect of this in both cases. It may not be as simple as it looks. Right now I have to run. Lots of work today. I will post again later. Thanks for your views.
Cheers.
Update.
I must admit that I did not do any thinking.
I did glug , glug , and lots of listening.
I noticed that the voltage shunt feed back had what seemed to be "darker" spaces between sounds ! Know what I mean ?
In addition on track 5 of The best of Four Play, there is a powerful transient drum . This has tremendous attack even on the voltage shunt circuit and in addition it has 'weight'. There is also a long decay that I cannot hear easily on the voltage series feedback circuit. It is not due to the electronics as it is probably more than a second - sounds like a huge leather skinned drum !
Played quite loud I am not missing much of the 'bite ' which I earlier said that shunt feedback did not have.
There is certainly more weight in the sound. It's simply more tuneful!
If this test were conducted with a low level opamp it might not show similar results because the circuit configuration would be different and the effect of the speaker impedance would be missing. Just a thought .
Now I will be listening to Koji Kitaro and Yani and see how the system sounds.
Cheers.
I must admit that I did not do any thinking.
I did
glug , glug , and lots of listening.I noticed that the voltage shunt feed back had what seemed to be "darker" spaces between sounds ! Know what I mean ?
In addition on track 5 of The best of Four Play, there is a powerful transient drum . This has tremendous attack even on the voltage shunt circuit and in addition it has 'weight'. There is also a long decay that I cannot hear easily on the voltage series feedback circuit. It is not due to the electronics as it is probably more than a second - sounds like a huge leather skinned drum !
Played quite loud I am not missing much of the 'bite ' which I earlier said that shunt feedback did not have.
There is certainly more weight in the sound. It's simply more tuneful!
If this test were conducted with a low level opamp it might not show similar results because the circuit configuration would be different and the effect of the speaker impedance would be missing. Just a thought .
Now I will be listening to Koji Kitaro and Yani and see how the system sounds.
Cheers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.