• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

KYW's linestage (split from RIAA for life time)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Re: Circuit

Konnichiwa,

Positron said:
Using a 100k with 100k resistor will severly lessen the high frequency response. The feedback circuit won't work on this problem. Look below at some figures of high frequency loss.

Consider a stage of gain of only 10, with 4pf grid to plate capacitance (two tubes in parallel so twice the gp capacitance) and no stray capacitance (very optimistic). I have computed (computer program) the -1db and -3db responses.

I presume you are refering to my suggested parallel 6SN7 Circuit?

I think you will need to check your computer programs a little.... They seem to be written by sciencce fiction writers.

Alternatively you need to change your logic, WRT NFB.

Positron said:

1) with 100k volume control set to midpoint
-1db = approx 6khz
-3db = approx 14khz

Not sure how you arrive at that, discounting stray the REAL results are actually with the volume control at midpoint (worst case) -1db @ 43KHz & -3db @ 84KHz, something I feel is tolerable for a Linestage to be used with CD Sources.

The variation of HF Response from worst to best case is minimal, with the volume set fully open (best case) the -1db point is @ 44KHz and the -3db point @ 86KHz, virtually no change from the "worst case" position.

I actually choose the circuit and recommended precisely because of the way it bahaves WRT to capacitive loading on the volume control, plus it's comparably low output imedance without using cathode followers outside a feedback loop. And of course because keeping the feedback loop around one stage only minimises the blatant sonic problems caused by using loop feedback around muktiple stages.

It makes a pretty good linestage, both subjectively and objectively, though I personaly would prefer an S&B TX-102, against which it can be easily compared BTW....

Sayonara
 
And this for the linestage. Just curious, cause as of yet I'm drawn toward Thorsten Sahib’s schematics. I really want to use those 6SL7’s and 6SN7’s
 

Attachments

  • line_stage_sch.jpg
    line_stage_sch.jpg
    29.5 KB · Views: 1,056
Konnichiwa,

corbato said:
And this for the linestage. Just curious, cause as of yet I'm drawn toward Thorsten Sahib’s schematics. I really want to use those 6SL7’s and 6SN7’s

This linesatge will have a gain of over 20db. Combined with that Phonostage we are taliking overall around 100db gain. That is 30db too much gain for the usual 0.2-0.5mV MC Pickups and 50db too much gain for the usual 2-5mV output MM Pickups. It will also be almost 20db too much gain for any CD Player.

This would mean that your volume control will always be near the left endstop, the preamp will be xcessively noisy and suffer from poor overload behaviour.

For Preamps I'd recommend to consider the following issues in planning gain:

1) A CD player will output eaks of 2.8V with an average level around 14db below this, giving around 0.5V RMS for program material with 14db headroom.

LP's will be cut in the "energy" range of the music with up to 14db above the rated 5cm/s @ 0dB and a "bogey" MM Pickup outputs 2.5mV @ 5cm/s with a "bogey" low output MC pickup at 0.25mV.

This means we can have levels with a similar peak/average ratio than CD and similar average loudness if our MM Phonostage has a gain of around 46db and our MC stage of around 66db.

Most modern amplifiers will have an input sensitivity of around 1V RMS for nominal output. Older Valve Amp's and some rather badly designed modern Valve Amp's have a lower input Voltage, very few Amp's (such as my preferred 2-Stage Amp's with medium to low Mu triode drivers and DHT outputs) require a little more input.

This suggests that our Linestage will most of the time be required to attenuate, to have a littlke extra gain can help in marginal combinations, hence my preferred 6db Linestage gain.

So, the overall for a full function preamp with MC Input and the volume full up should be around 72db (the CAT matches that exactly) and for the MM input 52db (the EAR 834P which is a MM Phonostage aimed at directly driving integrated Amplifiers at CD level BTW has pretty much exactly this level of gain), in practice even these values tend to be 6db too high, leaving 6db gain reserve in the Phonostage and 6db gainreserve in the linestage, based on 1V Amplifier sensitivity. Any higher gain levels will cause serious system matching problems.

Sayonara
 
Wow

This design reminds me of the 15,000pf IC that was going to be marketed. "Hey it sounds good". Of course it didn't and the frequency response of the preamp it was connected was -3db about 7khz. "But it sounded good".


"I presume you are refering to my suggested parallel 6SN7 Circuit?

I think you will need to check your computer programs a little.... They seem to be written by sciencce fiction writers.

(It is an engineering program, you are way off, either in calculating or the thevenin circuit itself.)

Alternatively you need to change your logic, WRT NFB.

(You obviously haven't thought things through. More on this below. Last string I mentioned there is a whole world you don't understand. You just demonstrated it again.)

quote:
Originally posted by Positron

1) with 100k volume control set to midpoint
-1db = approx 6khz
-3db = approx 14khz



Not sure how you arrive at that, discounting stray the REAL results are actually with the volume control at midpoint (worst case) -1db @ 43KHz & -3db @ 84KHz, something I feel is tolerable for a Linestage to be used with CD Sources.

(You better check your math again, or get the thevenin circuit correct. I also didn't include the capacitance of the volume control itself either.)

The variation of HF Response from worst to best case is minimal, with the volume set fully open (best case) the -1db point is @ 44KHz and the -3db point @ 86KHz, virtually no change from the "worst case" position.

(In your dreams, and wrong as usual. You obviously don't have the correct thevenin circuit. Did you even bother to measure it? With the 100k grid resistor, the pot does make less difference, but 6 to 10k isn't much difference.

There are major differences between midpoint and control set to "max".
It depends on getting rid of the 100k grid resistor as mentioned indicated above, and then the frequency response is determined by the source Z (when when the pot is at max volume, wiper arm 100k from ground) and the capacitance. This raises the frequency response well above that of midresistance point, which is stated in my post. And remember, the FR deviation accumulates from circuit to circuit.)

I actually choose the circuit and recommended precisely because of the way it bahaves WRT to capacitive loading on the volume control, plus it's comparably low output imedance without using cathode followers outside a feedback loop. And of course because keeping the feedback loop around one stage only minimises the blatant sonic problems caused by using loop feedback around muktiple stages.

(Compromise as usual instead of using a good design to begin with.
1st, capacitance does nothing good in this area.

2ndly, one doesn't need cathode followers, another stage, or the circuit you showed. There are ways around both.

3rdly, Using feedback around one stage. The problems of feedback depends on what follows, ICs capacitances, input Z of amp etc.
Do you even understand how these relate???

It makes a pretty good linestage, both subjectively and objectively, though I personaly would prefer an S&B TX-102, against which it can be easily compared BTW....

(Hope you don't give him a crappy PS design. This one measures and sounds crappy, if you like the highs dissappearing.)

_______________________________________

Yes, the gain does change the miller capacitance, but not enough to raise the poor frequency response of the network out of the audio band. Halving the gain to 5 only doubles the FR; thus still very poor.

And time delays in the feedback. This time delay depends on the Capacitance of ICs, and tube input capacitance of the amp etc. Changing Z input of the amp, changes feedback ratio etc., which changes the miller capacitance etc.

Mediocre design.
 
Re: Wow

Konnichiwa.

Positron said:
(It is an engineering program, you are way off, either in calculating or the thevenin circuit itself.)

Hmmm. My results come from P-Spice and cloesely match those I observed on a real version of this stage.

Positron said:
Last string I mentioned there is a whole world you don't understand. You just demonstrated it again.

Last time I mentioned that you deliberatly claim unrealistic results and consitently fail to show an appreciation of basic electronics. I again observe the same thing.

Positron said:
You better check your math again, or get the thevenin circuit correct.

My math for the Circuit I presented is correct. Please build teh circuit if you do not believe basic electronics and physical laws hold sway, you will be thought that still do.

Positron said:
I also didn't include the capacitance of the volume control itself either.

Nor did I, however most modern volume controls show very little local capacitance (mostly because the are largely made from plastic.

Positron said:
In your dreams, and wrong as usual.

How would you know, as you obviously fail to even appreciate how the circuit works.

Positron said:
There are major differences between midpoint and control set to "max".

Sorry, NOT IN MY CIRCUIT.

Positron said:
It depends on getting rid of the 100k grid resistor as mentioned indicated above, and then the frequency response is determined by the source Z (when when the pot is at max volume, wiper arm 100k from ground) and the capacitance. This raises the frequency response well above that of midresistance point, which is stated in my post.

Non of which has anything to do with the circuit I presented. So we are again at the game of using numbers that while having some relation to the context are completly unrelated to the topic at hand. Well, nothing new.

Positron said:
Compromise as usual instead of using a good design to begin with.

The compromise was set by the basic requirements stated. We may argue if my circuit works better in reality, within the stated limitation than yours - oopps, I forgot, you don't suggest circuits, you only try to put other peoples suggestions down without even comprehending what is being suggested.....

Positron said:
Using feedback around one stage. The problems of feedback depends on what follows, ICs capacitances, input Z of amp etc.
Do you even understand how these relate???

Absolutely.

Positron said:
Hope you don't give him a crappy PS design. This one measures and sounds crappy, if you like the highs dissappearing.

So, between my posting the circuit (which was previously build with ECC88 and a slightly higher gain but also tested with the 6CG7 which is a noval 6SN7) you build the circuit, managed to have an exhaustive listen, to test the crcuit on the bench and posted your results. All within 5 Hours. I am mightly impressed. I cannot compete with that. However, when you build the ciorcuit you must have made a few wiring mistakes (not to worry, even happens to me when I'm rushed) and as a result your tests seem to have gone a little astray.

Or maybe you did not test the circuit? Maybe you just used what little of electronics to try to understand and abysimally failed to do so?

Positron said:
Mediocre design.

Well, why don't you suggest one that is demonstrably better? I don't mind. This is a free for all. But please ensure to provide sufficient wide bandwidth, low output impedance and distortion. It's not like I'm stopping you to present your designs, which are so superior to uthose from the other great unwashed.

Sayonara
 
It makes a pretty good linestage, both subjectively and objectively, though I personaly would prefer an S&B TX-102, against which it can be easily compared BTW....


Can you expand a bit on the comparative sonic advantages of each relative to the other? The TX-102 seems to be a bit pricy, maybe justifying a little tweaking of the non-magnetic option.

TIA -
 
Wow

"Konnichiwa.

quote:
Originally posted by Positron
(It is an engineering program, you are way off, either in calculating or the thevenin circuit itself.)

Hmmm. My results come from P-Spice and cloesely match those I observed on a real version of this stage.

<Really, well you didn't use the proper thevenin circuit did you.>


quote:
Originally posted by Positron
Last string I mentioned there is a whole world you don't understand. You just demonstrated it again.


Last time I mentioned that you deliberatly claim unrealistic results and consitently fail to show an appreciation of basic electronics. I again observe the same thing.

<Interesting as I am the one who graduated number 1 in my Electronics Engineering class and you don't have a degree, do you???? Of course, you skipped over that one last string, didn't you. And you lost pitifully last string, remember.
Caught you stating you had Way more than two years into tubes, but your own bio stated you had only approx 2, remember? You've have been caught numerous times posting inaccurate positions I made, remember. Isn't that called lying?
By the way, do you know some more of the other weaknesses of "your" circuit? If you can't answer this, then who is the novice?>

quote:
Originally posted by Positron
You better check your math again, or get the thevenin circuit correct.

My math for the Circuit I presented is correct. Please build teh circuit if you do not believe basic electronics and physical laws hold sway, you will be thought that still do.

<Well, obviously, then you don't know the BASICs of how to thevenize a circuit. And I am the novice? >

quote:
Originally posted by Positron
I also didn't include the capacitance of the volume control itself either.

Nor did I, however most modern volume controls show very little local capacitance (mostly because the are largely made from plastic.

<Where did you come up with that reply? If you had measured some famous controls, the capacitance is fairly large do to the shielding of the metal case. Caught yourself again.>

quote:
Originally posted by Positron
In your dreams, and wrong as usual.

How would you know, as you obviously fail to even appreciate how the circuit works.

<Caught yourself again. And who pointed out the major weaknesses of the circuit? Me. Can you explain the ones I haven't brought up yet? This will tell if you really understand your circuit.>

quote:
Originally posted by Positron
There are major differences between midpoint and control set to "max".

Sorry, NOT IN MY CIRCUIT.

<Unfortunately, there is.>

quote:
Originally posted by Positron
It depends on getting rid of the 100k grid resistor as mentioned indicated above, and then the frequency response is determined by the source Z (when the pot is at max volume, wiper arm 100k from ground) and the capacitance. This raises the frequency response well above that of midresistance point, which is stated in my post.


Non of which has anything to do with the circuit I presented. So we are again at the game of using numbers that while having some relation to the context are completly unrelated to the topic at hand. Well, nothing new.

<If you understood the circuit as you claim, you would have answered differently.
At midpoint of the pot, phase shifting occurs well below 20khz. It is easily heard on a system.>


quote:
Originally posted by Positron
Compromise as usual instead of using a good design to begin with.

The compromise was set by the basic requirements stated. We may argue if my circuit works better in reality, within the stated limitation than yours - oopps, I forgot, you don't suggest circuits, you only try to put other peoples suggestions down without even comprehending what is being suggested.....

<No, there are many circuits I don't comment on here. But when I see a completely idiotic circuit, and the person wants something really good, well.... Let's see if you can find some more problems with this circuit. By the way, one can build much better with his list of parts, wider bandwidth, very low distortion, low output Z.>

quote:
Originally posted by Positron
Using feedback around one stage. The problems of feedback depends on what follows, ICs capacitances, input Z of amp etc.
Do you even understand how these relate???

Absolutely.

<When were you going to tell him about the problems I have mentioned and others I haven't yet?> Can you explain more other than "absolutely"?>

quote:
Originally posted by Positron
Hope you don't give him a crappy PS design. This one measures and sounds crappy, if you like the highs dissappearing.


So, between my posting the circuit (which was previously build with ECC88 and a slightly higher gain but also tested with the 6CG7 which is a noval 6SN7) you build the circuit, managed to have an exhaustive listen, to test the crcuit on the bench and posted your results. All within 5 Hours. I am mightly impressed. I cannot compete with that. However, when you build the ciorcuit you must have made a few wiring mistakes (not to worry, even happens to me when I'm rushed) and as a result your tests seem to have gone a little astray.

<Typical response. I have been designing for decades. I have used 6sn7a etc in all sorts of designs, including the one you presented, and have some in my stash right now.>

Or maybe you did not test the circuit? Maybe you just used what little of electronics to try to understand and abysimally failed to do so?

<Well, if you understood the circuit, and how it sounds, as you claim, you would have known all the weaknesses and flaws and could have recommended a better design. (Again, if you understand the basics, you explain the other problems this circuit has, that I have not yet mentioned. If you don't, everyone will know your novice status.)>

quote:
Originally posted by Positron
Mediocre design.

Well, why don't you suggest one that is demonstrably better? I don't mind. This is a free for all. But please ensure to provide sufficient wide bandwidth, low output impedance and distortion. It's not like I'm stopping you to present your designs, which are so superior to uthose from the other great unwashed."

<Sure, I am just going to give you my designs, which took years to develop. typical shill comment. :whazzat: >

<Ok, keep away from CFs, work much more on the power supplies, and use low output Z, high bandwidth, low distortion tubes, the best brands sonically too.>
 
Valves, Millercapacitance and feedback circuits

Konnichiwa,

Well, it seems I made my usual error and actually assume that those who claim to understand electronics are sufficiently competent not talk complete loblocks (as they say on Planet anagramia) and hence will make some sense and the rest doesn't want to know anyway. Seems I'm wrong.

So hence, let's evaluate, strictly on paper and in a simplified form the 6SN7 parallel linestage.

First, open loop we have an anode impedance of around 8k. Paralleled this gives an internal impedance of 4k and with a 10k Load a stage gain of 12.6, this comes from comparing the Load (22k//10k) Rl and the internal impedance of the stage (4k) Ri as voltage divider and reducing the Valves Mu by this factor as:

Gain = Mu/((Rl+Ri)/Rl)

Now we can calculate the Miller capacitance and using this also the input capacitance. For average Valves my Datasheets list 3.3 & 3.5pF as Anode-Grid Capacitance, so both systems in parallel show 6.8pF which will be "amplified" by the stage gain, 12.6 to give a virtual 85pF input Capacitance between Grid and Andode. Adding the other parasitic valve capacitances we get a rounded 100pF capacitance if good quality (ceramic) sockets are used and the layout is sensible.

Now we can calculate the relative impedance of the gridnode to obtain the OPEN LOOP bandwidth (note that this a looped feedback amplifirer stage).

At the worst case (volume control at -6db setting) the 100k Volume control will present around 25k output impedance, to this is added the 100k Input resistor of the inverting, looped feedback amplifier. In parallel to this overall 125k impedance is the feedback loop and internal impedance of the circuit, so basically 244k. In parallel this gives a impedance of 82.6k.

If we combine the 82.6k and the 100pF we get a timeconstant of 8.26uS or a -3db point of 159115/8.26=20KHz.

So, OPEN LOOP we have a -3db point of 20KHz in the worst case scenario. This frequency is shifted upwards in the usual manner as discussed by Harold Black in the 1930's by the diference between the Open Loop gan and the closed loop gain.

In my case the closed loop gain is 2, so we have around 6.3 or 16db feedback factor, so in theory and not accounting for some of the parasitics I include in my P-Spice simulation (such as the 1nF output load of the 10k/1nF IEC Load) we have theoretically a closed loop -3db point of 126KHz under worst case conditions, but only if there is no capacitive load. My simulations assume 1nF capacitive load and hence are less favourable.

Let's also work out the apparent output impedance of the Linestage. WE have a 22k Load in parallel with a 4k internal impedance, giving under open loop conditions 3.4KOhm output impednace. This will be reduced (again) by the feedback factor giving us theoretically around 530 Ohm output impedance.

A last check would be how much the difference between fully open and worst case setting on the volume control is. Using above Methode we now have 100k//244k to give 71k effective impedance at the grid node. This gives a 7.1uS timeconstant or 22.4KHz open loop -3db point and closed loop theoretically a 141KHz -3db closed loop.

This concludes the current EE101 lesson for those who wanted to know and those who obviously slept through EE101. We thank you for your attention and suggest that everyone her or himself evaluate designs themselves, using sensible methodes instead of listening to people who obviously lack basic analytic skills (or deliberatly distort the truth - I leave it to the audience to decide which takes place).

Sayonara
 
Wow.

Too bad you have to resort to such negative tatics. But the important thing is that the truth is being exposed.

First off, he Again refuses answer the question of education. Not surprising since, as we shall see below, he only has a couple of years experience with tubes and the design he presented shows it. Let's take a look.

<"If we combine the 82.6k and the 100pF we get a timeconstant of 8.26uS or a -3db point of 159115/8.26=20KHz.">

Actually 17.5k, using the electronics computer program, and the
-1db point is only around 8.7k (we will use your own figures), Very poor open loop response indeed. Similar to open loop SS designs with feedback, esp in the old days.

The only reason the open loop FR is even this good is because of his feedback scheme and the need for the 220k in the circuit. Of course Thorsten needs to increase the open loop FR response. Very messy circuit indeed. Remember how delicate music is and how easy it is to upset things sonically.

You finally got around to mentioning the 16db of negative feedback to try to bring up the response and lower the output Z. Real brilliant design there. But there are more problems indeed using this design. Can you figure them out?

By the way, my design has an Open loop FR of over 100khz (pot midpoint) with 100pf/100k load, extremely low distortion, and output Z pretty low. All without 16db of feedback to screw the sound up!

<"This concludes the current EE101 lesson for those who wanted to know and those who obviously slept through EE101. We thank you for your attention and suggest that everyone her or himself evaluate designs themselves, using sensible methodes instead of listening to people who obviously lack basic analytic skills (or deliberatly distort the truth - I leave it to the audience to decide which takes place).">


Not a smart comment Thorsten. You already got caught in the last string trying to hide the perks/compensation you receive under the table (I had to ask you twice before you would spill the truth, or partial truth, remember; while I volunteered mine). You also distorted my positions too. And now the comment above. I have always found that deception and unethical behaviour doesn't lead to being trusted?

By the way, you only mentioned the basics, including the 16db of feedback. Now explain the rest of the problems with the circuit and the feedback.
 
Re: Wow.

Konnichiwa,

Positron said:
<"If we combine the 82.6k and the 100pF we get a timeconstant of 8.26uS or a -3db point of 159115/8.26=20KHz.">

Actually 17.5k, using the electronics computer program,

A computer program that arrives at 17500 when dividing 159115/8.26 seems rather inacurate.

And if you mean to say that your program arrives at a different analisys, you should be able to point out the ommisions in my calculations (which where done primarily for breviy and clarity), that assuming of course that you actually understand what is going on, which based on your comments so far I am forced to seriously doubt.

Positron said:
You finally got around to mentioning the 16db of negative feedback to try to bring up the response and lower the output Z.

Well, number one, the use of Feedback is an inherent design feature. Negative feedback is a design tool, just like any other. It needs to be understood to be used correctly. Only those who fail to understand it take refuge to comments of "feedback is bad" as an absolute.

In my case negative feedback is the primary tool to achieve low gain while using an active stage and avoiding followers. The frequency response and low output impedance are the redsult of doing things that way. One could have used an output transformer to achieve low gain, low output impedance and wide bandwidth, the original request was not to use Transformers, so non are used.

Positron said:
Real brilliant design there.

Thank you for your compliment.

Positron said:
But there are more problems indeed using this design. Can you figure them out?

There are no PROBLEMS if the circuit is implemented as suggested and operated within it's design parameter limits. We may argue about the way this circuit sounds or not, which would require after all experience with the circuit and it's sonics, which you seem to still lack.

Positron said:
By the way, my design has an Open loop FR of over 100khz (pot midpoint) with 100pf/100k load, extremely low distortion, and output Z pretty low.

A design, for which you make any number of claims which cannot be verified (unlike the ones for mine) as you fail to share it. I will simply put these parameters down to your usual baseless claims as we have seen in past discussions.

Positron said:
All without 16db of feedback to screw the sound up!

If the use of negative feedback "screws up the sound" the implementation of negative feedback was incompetently done. So you may wish to re-evaluate your views on negative feedback as a design tool.

Positron said:
I have always found that deception and unethical behaviour doesn't lead to being trusted?

I would consider making deliberatly misleading statements (eg claiming a circuit that has illustrably a frequency response of -1db @ > 40KHz has one of 1db @ 6KHz) both unethical and deceptive.

Of course, it seems you call that behaviour ethical and truthfull. I must insist that this is where we part company.

Positron said:
Now explain the rest of the problems with the circuit and the feedback.

How about you do so, making sure your math is actually up to scratch and your computer program can at least divide correctly and you math is based on reality not wishful thinking.

Sayonara
 
Wow.

<"quote:
Originally posted by Positron
<"If we combine the 82.6k and the 100pF we get a timeconstant of 8.26uS or a -3db point of 159115/8.26=20KHz.">

Actually 17.5k, using the electronics computer program,

A computer program that arrives at 17500 when dividing 159115/8.26 seems rather inacurate.

And if you mean to say that your program arrives at a different analisys, you should be able to point out the ommisions in my calculations (which where done primarily for breviy and clarity), that assuming of course that you actually understand what is going on, which based on your comments so far I am forced to seriously doubt.">

Your figure is only approx. I actually measured them, but of course measured stray capacitance too, which we both didn't include when starting. But let's give you the benefit of the doubt. The open loop response is still -1db at 10khz and -0,5db at approx 5khz. Still very poor.

<"Well, number one, the use of Feedback is an inherent design feature. Negative feedback is a design tool, just like any other. It needs to be understood to be used correctly. Only those who fail to understand it take refuge to comments of "feedback is bad" as an absolute.

In my case negative feedback is the primary tool to achieve low gain while using an active stage and avoiding followers. The frequency response and low output impedance are the redsult of doing things that way. One could have used an output transformer to achieve low gain, low output impedance and wide bandwidth, the original request was not to use Transformers, so non are used.">

Well, give me a wrong way to use the feedback in a single stage preamp. It isn't hard to do correctly, or at least as best as is possible.

<"There are no PROBLEMS if the circuit is implemented as suggested and operated within it's design parameter limits. We may argue about the way this circuit sounds or not, which would require after all experience with the circuit and it's sonics, which you seem to still lack.">

Well, that sounds suspicious to me. You don't know of any more problems?
And I built the circuit decades ago, so another inaccurate claim. The circuit is only ordinary at best.

<"Originally posted by Positron
By the way, my design has an Open loop FR of over 100khz (pot midpoint) with 100pf/100k load, extremely low distortion, and output Z pretty low.


A design, for which you make any number of claims which cannot be verified (unlike the ones for mine) as you fail to share it. I will simply put these parameters down to your usual baseless claims as we have seen in past discussions">

Actual measurements Thorsten. I am certainly not going to give you a schematic to copy and waste decades of private research.
By the way, I just recently got the latest review from BFS, who accepts no kickbacks or advertising, on my preamp. What reviews do you have on yours??

<"Originally posted by Positron
All without 16db of feedback to screw the sound up!

If the use of negative feedback "screws up the sound" the implementation of negative feedback was incompetently done. So you may wish to re-evaluate your views on negative feedback as a design tool.">

So tell me how to do it wrong?? What you have done is very elementary, my dear watson.

<"Originally posted by Positron
I have always found that deception and unethical behaviour doesn't lead to being trusted?


I would consider making deliberatly misleading statements (eg claiming a circuit that has illustrably a frequency response of -1db @ > 40KHz has one of 1db @ 6KHz) both unethical and deceptive.

Of course, it seems you call that behaviour ethical and truthfull. I must insist that this is where we part company. ">

Interesting since you were the one who refused to devulge your unethical behaviour until I asked twice.

I probably could have brought up the point directly, that the open loop response was very poor. But I wanted to find out if you actually understood the circuit, and for you to mention this problem. If I mentioned it upfront, you could have easily slid by. Frankly, I didn't think you even understood the basics. You do show some knowledge of design though, I will give you credit for that.

<"Originally posted by Positron
Now explain the rest of the problems with the circuit and the feedback.

How about you do so, making sure your math is actually up to scratch and your computer program can at least divide correctly and you math is based on reality not wishful thinking.">

Remember I am finding out what you know; as you claim, from your above post, to be teaching me.
I wouldn't have graduated No. 1 in my class if I didn't know my stuff. I am not going to give you a free education. It took me alot of work and time.

If you wish to exit this conversation, go ahead. No problem.

ps. Go fighting Illini
 
Re: Wow.

Konnichiwa,

Positron said:
Your figure is only approx.

Do not distract from the issue.

Positron said:
I actually measured them, but of course measured stray capacitance too, which we both didn't include when starting. But let's give you the benefit of the doubt. The open loop response is still -1db at 10khz and -0,5db at approx 5khz. Still very poor.

You measured them. Right. May I ask when you build the circuit? And could you then illustrate the change in response with volume control settings you calimed earlier? Any comments on the sound?

On another line, would you mind illustrating in what sense an open loop response of -3db @ 20KHz in a looped feedback Amp (single stage) is "poor". It is more than adequate for the purpose. The open loop response is a direct result of a number of design choices.

It would be easy to change the open loop response to a much wider one by lowering the impedance values in the feedback setting circuitry and by using a lower resistance volume control. If we (for arguments sake) lower all values by a factor 10 we can easly achive a ridiculously wide bandwidth, which has no advantage whatsoever (sonically or other) and introduces a number of other problems.

It is my experience that a well implemented and sensibly set bandpass behaviour is preferable in linestages, arguably a sentiment open to discussion, however I usually follow my own design requirements when designing something.

Positron said:
Well, give me a wrong way to use the feedback in a single stage preamp. It isn't hard to do correctly, or at least as best as is possible.

Is the above your opinion or are you stating a fact? If you are stating a fact, would you mind illustrating WHY applying loop feedback to achieve low gain and to take advantage of the stages excess gain to improve linearity and lower the outputimpedance is "a wrong way to use feedback"? Maybe you would like suggest the "right" way to use feedback?

At any extent, you may notice that we are now outside the realm of factual performance (unless you can illustrate that the laws of physics have recently gone on annual leave) and in the realm of design preferences and how they relate to sonical performance.

Positron said:
And I built the circuit decades ago, so another inaccurate claim. The circuit is only ordinary at best.

You build a Circuit I published here a few days ago DECADES AGO, with identical resistor, cpacitor and other values? Interresting feat. I now would also ask for plans to your time machine, I could use one.

I will readily admit that similar circuits have been around for ages and some of the best sounding commercial preamps of the last two decades used variations thereoff. So it surely cannot be THAT BAD....

Positron said:
<A design, for which you make any number of claims which cannot be verified (unlike the ones for mine) as you fail to share it. I will simply put these parameters down to your usual baseless claims as we have seen in past discussions">

Actual measurements Thorsten. I am certainly not going to give you a schematic to copy and waste decades of private research.
By the way, I just recently got the latest review from BFS, who accepts no kickbacks or advertising, on my preamp. What reviews do you have on yours??

Well, first, are your actual measurement independently verified?

Secondly, the design I presented is strictly DIY, not commercial> A design based around my ideas and parts mostly designed by me is the MF Audio Passive Preamp, the Bent Audio Noh and the DIY Hifisupply Django, plus there are other Preamps using the same core (S&B TX-102) in both the DIY and commercial domain.

Many reviews by many individual exist and they tend to agree in the largest part that the S&B TX-102 based passives outperform ANY active preamp these individuals have tried and also most/all passive linestages from other sources. That is good enough for me.

Positron said:
So tell me how to do it wrong?? What you have done is very elementary, my dear watson.

First, I never claimed that what I did was NOT elementary. As to "How to do it wrong"? For starters loop feedback around several stages and in circuits where there is more than one dominant pole and have an open loop response (-3db) that is substantially narrower than the desired signal bandwidth. But you should know that, if you are 1/10th as good as you claim to be.

Positron said:
<"Originally posted by Positron
I have always found that deception and unethical behaviour doesn't lead to being trusted?


I would consider making deliberatly misleading statements (eg claiming a circuit that has illustrably a frequency response of -1db @ > 40KHz has one of 1db @ 6KHz) both unethical and deceptive.

Of course, it seems you call that behaviour ethical and truthfull. I must insist that this is where we part company. ">

Interesting since you were the one who refused to devulge your unethical behaviour until I asked twice.

WHAT unethical behaviour? Giving free advise when asked for it? Providing my designs for all to use who would like to do so? Pointing out the mistakes in basic electronic theory application you make when leveling irrational criticism at designs I suggest?

Positron said:
I probably could have brought up the point directly, that the open loop response was very poor.

Correction. The open loop response is poor in your considered opinion, whereas in my case I consider it adequate and a good compromise between a number of conflicting design requirements. We can argue endlessly about what is the better compromise, the choice is in the ear of the beholder.

Positron said:
But I wanted to find out if you actually understood the circuit, and for you to mention this problem.

Nice excuse for incompetence. It would ring truer if you had less of a history of completly ballsing up on basic electronics. I took the time yesterday to review a decent number of posts by you. I rarely seen as many mistakes in basic things from anyone having at least basic competence in electronics. I had more than a few good laughs on your account.

Positron said:
If I mentioned it upfront, you could have easily slid by. Frankly, I didn't think you even understood the basics. You do show some knowledge of design though, I will give you credit for that.

I'm honoured.

But truth is, what is the point if an illiterate peasant pays backhanded compliments to Newton?

Positron said:
<"Originally posted by Positron
Now explain the rest of the problems with the circuit and the feedback.

How about you do so, making sure your math is actually up to scratch and your computer program can at least divide correctly and you math is based on reality not wishful thinking.">

Remember I am finding out what you know;

I notice. But don't expect much else from here.

Positron said:
I wouldn't have graduated No. 1 in my class if I didn't know my stuff. I am not going to give you a free education. It took me alot of work and time.

Well, it seems that you spend a lot of time and work to fail to know even the basics of electronics. I shall take in the next few days take the time to illustrate this a bit further (I'm actually going to enjoy that) and collect your various psotings of great wisdom over the last few month which demonstrably show an appaling lack of being able to apply basic logic and evaluating basic electronic circuits. And not forget the public (re) posting of some of the olded urban legends in the book, that was priceless.

Sayonara
 
Ex-Moderator
Joined 2003
Gentlemen, would it be possible for one or both of you to post diagrams of what you are discussing? Otherwise, entertaining though it may be, your exchanges are beginning to resemble a pantomime ("Yes, I did" - "No, you didn't").

With the aid of diagrams, a spirited design discussion would be most useful for less experienced members...
 
Konnichiwa,

EC8010 said:
Gentlemen, would it be possible for one or both of you to post diagrams of what you are discussing?

I have posted what I am discussing within this thread, at this post:

http://www.diyaudio.com/forums/showthread.php?postid=353015#post353015

As you are aware from the discussion the other guy (I refuse to use "Gentleman" in the context) has repeatedly refused to share anything, whatsoever, prefering sniping from the sides at those more willing to share (and possibly more able than he is?) to actually contributing anything useful.

EC8010 said:
Otherwise, entertaining though it may be, your exchanges are beginning to resemble a pantomime ("Yes, I did" - "No, you didn't").

I am sorry, but if someone posts utter nonsense IN CRITIQUE to contributions by me I am unwilling to let this go. I don't much care in other contexts.

EC8010 said:
With the aid of diagrams, a spirited design discussion would be most useful for less experienced members...

As pointed out above, the aid of such diagrams is indeed provided, at least for my side of the argument.

Sayonara
 
Konnichiwa,

Peter Daniel said:
Although the RIAA and Linestage might not, this disusion has good chances to be for life time;)

One rather hopes not. I for one am well sick and tired of this whole thing (and very irritable anyway as it's financial year end where I work). I got better things to do. Where the technical side is concerned I have made any possible point (and scored them all and easily) as for ethics and morals, i leave that to gentlemanly reader to decide.

I think however I will not want to miss putting together my little "positronic non-sense" post (as compensation), but that will go into a seperate thread.... ;-)

Sayonara
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.