Of course 1% resistors can replace the pot. My surplus wirewound pots cost me $1 each. Now they are $25 or more, so a fixed resistor is best, but you need a whole range of values. What is best is to parallel a resistor to either side that needs it, with plug in sockets. The next generation will be made that way.
I don’t remember when it was the last posting of such a spectrum but I remember that I had posted some FFTs in support of an earlier comment of Dan over how detrimental to the sound across the audible bandwidth are the low frequency components due to IM products
http://www.diyaudio.com/forums/lounge/146693-john-curls-blowtorch-preamplifier-part-ii-4483.html#post3689830
There is not a single word about dither in Masterlink 9600 user’s manual!
Ed
The 60Hz peak is about 2uV, the rest of the peaks are around 0.2uV. What’s the recipe of success?
Very kind of you remembering us Steven
George
Engineers are practical first, so they get along well - when it suits them.
I also get the feeling that a lot of readers are interpreting the exchanges here with more emotionality than the writers actually feel. I get the impression that for those who have been arguing for decades, it just doesn't get to them anymore since they've been at it so long.
Maybe ... what pot's do IME is "dirty" up the sound - a pristine recording may come across just as well as normal, but a recording that is marginal, that is on the point of constantly irritating you, will be pushed over the edge - will be impossible to listen to. IOW, the damage is done not to superb recordings but the dodgy ones, the extra layer of dirt added by the non-perfect pot is reducing by a great deal one's ability to enjoy "difficult" recordings ...
George
You realize the output is the voltage across a resistor that is not connected to anything else. I wanted to raise the issue of critical distance.
If the resistor were placed close to the power supply the noise would as expected increase. As you move the resistor farther away the level will begin to drop following the inverse square law until you reach the critical distance. No matter how far you are from the power supply the noise will bounce around inside the case this field once you get a bit of distance away from the power supply is pretty uniform. The critical distance is where the residual level is equal to the directly radiating field.
What this means is that every signal wire or trace will be picking up this level of noise. The higher impedances will of course have a greater voltage.
Now as this level is constant as the music drops the noise is no longer masked.
What is interesting is that it is mostly odd order harmonics and shows two peaks. The higher band is probably from the internal inductance of the filter capacitor. The other peak is from the effect that the harmonics roll off with increasing frequency while the coupling increases with frequency.
Note the lack of diode switching noise or even harmonics.
To me the important issue is that PSSR is nice in a circuit but the limit for power supply noise may be elsewhere.
Now since this is the blowtorch thread it is worth noting that it uses an outboard power supply with additional filtering in the preamp itself.
ES
You realize the output is the voltage across a resistor that is not connected to anything else. I wanted to raise the issue of critical distance.
If the resistor were placed close to the power supply the noise would as expected increase. As you move the resistor farther away the level will begin to drop following the inverse square law until you reach the critical distance. No matter how far you are from the power supply the noise will bounce around inside the case this field once you get a bit of distance away from the power supply is pretty uniform. The critical distance is where the residual level is equal to the directly radiating field.
What this means is that every signal wire or trace will be picking up this level of noise. The higher impedances will of course have a greater voltage.
Now as this level is constant as the music drops the noise is no longer masked.
What is interesting is that it is mostly odd order harmonics and shows two peaks. The higher band is probably from the internal inductance of the filter capacitor. The other peak is from the effect that the harmonics roll off with increasing frequency while the coupling increases with frequency.
Note the lack of diode switching noise or even harmonics.
To me the important issue is that PSSR is nice in a circuit but the limit for power supply noise may be elsewhere.
Now since this is the blowtorch thread it is worth noting that it uses an outboard power supply with additional filtering in the preamp itself.
ES
Last edited:
There is such a thing as 'professional courtesy'. I learned it when I worked for large companies in the '60's and the '80's. It is a little like military courtesy, it isn't easily completely forgotten.
SY has been attacking my professional reputation for years now. However, he needed professional courtesy from me to make his task easier. However, I really don't need to be second guessed as to how and why I design the circuits I design, especially from someone who does not believe that it is necessary to make designs to the extent that I prefer to take them. Bipolars vs jfets? You choose. I have already made my choice, and I don't have to be 'second guessed' why.
It is a little like why I used a 10 ohm 10 turn wirewound resistor to balance out the 2'nd harmonic distortion. There are other ways, BUT 30 years ago, we had lots of military surplus stores in this area, AND the 10 ohm pots only cost $1 ea or so. Why not use them?
Today, an alternative would be used, because 10 turn wirewound pots cost 25 times more, than they did back then. I stand behind my original choice 30 years ago. It is the same with using jfets. They were CHEAP and plentiful back then, with many companies supplying them. If a jfet costs $0.015 each, as they did then, why bother with bipolars? Are bipolars essentially as good? Sometimes. There is no controversy, or even room for debate. Let's keep this in mind.
SY has been attacking my professional reputation for years now. However, he needed professional courtesy from me to make his task easier. However, I really don't need to be second guessed as to how and why I design the circuits I design, especially from someone who does not believe that it is necessary to make designs to the extent that I prefer to take them. Bipolars vs jfets? You choose. I have already made my choice, and I don't have to be 'second guessed' why.
It is a little like why I used a 10 ohm 10 turn wirewound resistor to balance out the 2'nd harmonic distortion. There are other ways, BUT 30 years ago, we had lots of military surplus stores in this area, AND the 10 ohm pots only cost $1 ea or so. Why not use them?
Today, an alternative would be used, because 10 turn wirewound pots cost 25 times more, than they did back then. I stand behind my original choice 30 years ago. It is the same with using jfets. They were CHEAP and plentiful back then, with many companies supplying them. If a jfet costs $0.015 each, as they did then, why bother with bipolars? Are bipolars essentially as good? Sometimes. There is no controversy, or even room for debate. Let's keep this in mind.
I'm trying to determine what the circuit IS. If you're aware of things on the drawings you posted not being correct, it would help the discussion to point them out. If you don't want to discuss your circuits, just say so and I'll stop asking questions and trace everything out for myself.
That would be too bad, since there's been a lot of useful and pertinent discussion over the past few days.
SY, you are not the first person to find fault with the schematic. Just look at what we discussed about it previously.
Repair, does not mandate that you completely understand what I designed, if you can get it to work, with extra details. I DO hope that the pchannel jfets that you replaced were reasonably matched in Idss and over 30ma Idss each. This could become a problem if you did not use similar Idss jfets.
Now, IF you have your voltages, and they are essentially correct, then the circuit should work again. If you still have problems, it will not be in the power supply portion, and you should look elsewhere on the circuit board. Replacing any other jfets will be a TRUE challenge for you and you might have to send it to me.
Repair, does not mandate that you completely understand what I designed, if you can get it to work, with extra details. I DO hope that the pchannel jfets that you replaced were reasonably matched in Idss and over 30ma Idss each. This could become a problem if you did not use similar Idss jfets.
Now, IF you have your voltages, and they are essentially correct, then the circuit should work again. If you still have problems, it will not be in the power supply portion, and you should look elsewhere on the circuit board. Replacing any other jfets will be a TRUE challenge for you and you might have to send it to me.
Where the hell do you get this stuff from Frank? Either you honestly believe it, or you are just having fun rattling everyone's cage.
I suppose you gleaned this listening to AC/DC at 115dB SPL? On your computer speakers by any chance?
LOL
Try, "honestly believe it" ...
I don't just come up with an idea just for the hell of it - I do, gulp, experiments ... if after replacing the pot mechanism say, 10 times in various pieces of equipment the sound improves each time, then I get the sense that there might be some sort of trend going on ....
Actually, just the process of refreshing the pot contact when the sound lost some of its quality, and hearing the better quality being restored by that action gave me the answer: the wiper is a point of weakness in the implementation. So, get rid of the wiper!
Why do people have a hangup about loudness? A solo violin piece, say, makes this sort of misbehaviour so obvious ...
I don't just come up with an idea just for the hell of it - I do, gulp, experiments ... if after replacing the pot mechanism say, 10 times in various pieces of equipment the sound improves each time, then I get the sense that there might be some sort of trend going on ...
.Actually, just the process of refreshing the pot contact when the sound lost some of its quality, and hearing the better quality being restored by that action gave me the answer: the wiper is a point of weakness in the implementation. So, get rid of the wiper!
Why do people have a hangup about loudness? A solo violin piece, say, makes this sort of misbehaviour so obvious ...
The trouble is, in spite of people rationalising like crazy, this, https://www.youtube.com/watch?v=aP_Z3s8Zas0, probably underlies a lot of people's thinking ...
The function of the very-low-ohm potentiometer in JC's machine is a trim of gain of the complementary sections for the primary purpose of nulling 2nd harmonic. It was a convenience and I think a judicious choice, given the care lavished on the product. Otherwise in general I eschew production line adjustments if they can be avoided (litany of horror stories reserved for the moment).Try, "honestly believe it" ...
I don't just come up with an idea just for the hell of it - I do, gulp, experiments ... if after replacing the pot mechanism say, 10 times in various pieces of equipment the sound improves each time, then I get the sense that there might be some sort of trend going on ...
Actually, just the process of refreshing the pot contact when the sound lost some of its quality, and hearing the better quality being restored by that action gave me the answer: the wiper is a point of weakness in the implementation. So, get rid of the wiper!
Why do people have a hangup about loudness? A solo violin piece, say, makes this sort of misbehaviour so obvious ...
Would a step attentuator with lots of wiping action and silver (?) contacts have worked better? Perhaps, but (if I understand correctly) as he has stated future things will use fixed resistors whose values have been determined before final installation---he is not advocating the use of potentiometers in general, or if used is fully aware of their potential drawbacks.
One friend swears by a tapped transformer level control. His data on its performance reveals some remarkably suspect peaking in the circuit he evaluated with an analyzer. But he is adamant about its virtues. I don't know. In any event it has little relevance to the Blowtorch stepup stage, and that's why I mentioned it.
That's why, while I've been a big fan of transformers, I've never been a fan of transformer-based attenuators.
A given transformer will give its best performance into a given load. But with a transformer-based attenuator, you effectively have n number of transformers depending on how many taps you use, but they're all driving the same load. So you'll never be able to optimize for all taps unless you also switch in different load resistors, which is just crazy.
Plus having all those taps just messes up your windings compared to a straight transformer.
So I've always preferred a fixed transformer loaded by a resistive attenuator.
se
Every area of a design will have ways of including intelligent solutions which are inaudible in impact - a good enough pot may be fine in many situations. I was frustrated in my early days of audio experimenting with hearing artifacts from volume and tone controls - and "solved" it by using silicon attenuators, or digital volume.
True but when a transformer involved, it must be the best one possible. Silver wire?
That's one application where silver wire might make some sense, but it would only result in a rather small (about 6% or so) reduction in parasitic resistance. And that's all that silver brings to the table.
se
- Status
- Not open for further replies.