From a preamp to power amp supply caps. 
My take:
Esr means slew in the opposite direction regards to the output that the amp must slew against (eating bandwidth). Esr must be kept low, by use of good caps or several not so good. Several caps may mean additional inductance, and we don’t want that either.
From my experience and belief, the implementation and component selection has the greatest influence on sound quality, more so than the capacitance value alone.
My take:
Esr means slew in the opposite direction regards to the output that the amp must slew against (eating bandwidth). Esr must be kept low, by use of good caps or several not so good. Several caps may mean additional inductance, and we don’t want that either.
From my experience and belief, the implementation and component selection has the greatest influence on sound quality, more so than the capacitance value alone.
I don't see why you're scratching your head. All the principles are the same, regardless. There's only the question of how much current is involved.
Grey
Grey
Hi Grey,
I've worked on those as well. The entire reason they had large capacitance is because it was in vogue at the time. I don't know if they still do that, but market pressures probably force them to do this.
I rebuilt a PV-5 preamp once. Very illuminating for me as CJ recommended I replace the older filter caps with film caps. I talked this over with the customer and ended up buying the parts from CJ and performing the change. The film caps had much lower capacitance than the original parts.
So, what did I measure / see? I saw more ripple, but a complete absence of a peak on the leading edge of the rectified wave form. The output of the regulator was cleaner (trace had less fuzz) and it sounded great. I learned at that point in time that capacitance did not need to be excessive. Surprisingly, the value of capacitance we ended up with was very close to what you would see in vintage tube equipment. I also learned that lower DCR in a transformer leads to more noise as the damping is reduced. I install resistance these days in low level circuits because it reduces the amount of overall noise I have to deal with. 120 Hz with lower harmonics is a goal to strive for in my opinion. Voltage regulators can deal with these lower frequencies much better.
I happen to disagree with you on this point and happen to have a fair amount of experience in that area. I'm simply stating things as I have seen them. I have had to change my point of view in the course of my education. Believe it or not, the "old guys" did get it right when designing amps and power supplies.
-Chris
I've worked on those as well. The entire reason they had large capacitance is because it was in vogue at the time. I don't know if they still do that, but market pressures probably force them to do this.
I rebuilt a PV-5 preamp once. Very illuminating for me as CJ recommended I replace the older filter caps with film caps. I talked this over with the customer and ended up buying the parts from CJ and performing the change. The film caps had much lower capacitance than the original parts.
So, what did I measure / see? I saw more ripple, but a complete absence of a peak on the leading edge of the rectified wave form. The output of the regulator was cleaner (trace had less fuzz) and it sounded great. I learned at that point in time that capacitance did not need to be excessive. Surprisingly, the value of capacitance we ended up with was very close to what you would see in vintage tube equipment. I also learned that lower DCR in a transformer leads to more noise as the damping is reduced. I install resistance these days in low level circuits because it reduces the amount of overall noise I have to deal with. 120 Hz with lower harmonics is a goal to strive for in my opinion. Voltage regulators can deal with these lower frequencies much better.
Grey, what are you talking about now? This isn't a contest. There are no prizes for being on one side or the other.
I happen to disagree with you on this point and happen to have a fair amount of experience in that area. I'm simply stating things as I have seen them. I have had to change my point of view in the course of my education. Believe it or not, the "old guys" did get it right when designing amps and power supplies.
-Chris
anatech said:
Ah, but you see, you missed the moral of the story entirely...you didn't increase the film capacitance beyond the original amount, which was dictated by price and by the physical constraints of the chassis.
I know whereof I speak. I owned a Conrad Johnson Premier Two preamp, which had electrolytic caps, and then bought a Premier Three which used much less capacitance, but as all film caps, just as you describe for the PV-5 (another preamp I am intimately familiar with). However, I didn't stop there, I quadrupled the film capacitance.
What happened?
The bass and dynamics improved. Golly, what a surprise!
You stopped too soon and drew the wrong conclusion.
Grey
Hi Grey,
I did some further experimenting later with other CJ pre's and even other brands. I did not stop too soon. I also discovered that less capacitance worked well. The new film caps where horribly expensive compared to an electrolytic. Even at cost (CJ's offer to try out their theory). Price didn't come into the equation, except to get clearance from the customer.
Don't forget, the CJ product (and many others) use a linear voltage regulator which shields the circuitry from the nasty filter side. That is unless there is HF hash that tends to get through.
You know, I once believed as you do now. I was a much younger technician.
Another area to look at is the heater filter / regulator. Here you need to have a large capacitor to prevent the regulator from dropping out. I've had success in reducing noise form this power supply by using a little series resistance before the cap. Many tube preamps have this error (huge capacitor with no series resistance). This drops the average raw DC a little, but the troughs are not as bad. Overall, less noise.
For my next trick, I'll be using switching regulators for the heater supply. 😉
-Chris
I did some further experimenting later with other CJ pre's and even other brands. I did not stop too soon. I also discovered that less capacitance worked well. The new film caps where horribly expensive compared to an electrolytic. Even at cost (CJ's offer to try out their theory). Price didn't come into the equation, except to get clearance from the customer.
Don't forget, the CJ product (and many others) use a linear voltage regulator which shields the circuitry from the nasty filter side. That is unless there is HF hash that tends to get through.
You know, I once believed as you do now. I was a much younger technician.
I am mystified by this comment. What on earth could cause better bass response? The B+ is regulated and doesn't drop even at very high signal levels. There is zero reason for the bass to improve. In fact, there is no reason for any change unless you provoke ringing in the rectifiers. Dynamics improve? Again, no B+ change = no performance change. If anything, you may have increased the noise level due to the creation of hash.
Another area to look at is the heater filter / regulator. Here you need to have a large capacitor to prevent the regulator from dropping out. I've had success in reducing noise form this power supply by using a little series resistance before the cap. Many tube preamps have this error (huge capacitor with no series resistance). This drops the average raw DC a little, but the troughs are not as bad. Overall, less noise.
For my next trick, I'll be using switching regulators for the heater supply. 😉
-Chris
To add some more raw, mostly unusable data based on buiding and modifiying a large number of tube/SS amps/preamps.
- adding capacitance above a certain value always makes an audible difference, both in preamps and power amps. Usually, but not always, this sounds worse to my ears. Depending on taste, choice of speakers and music i can imagine some listeners would prefer the sound.
- strangely, the same is true for dc heater supplies (indirectly heated tubes).
- changing tranformer type or power rating is also clearly audible. Usually, but not always higher power rating is preferable. Some very large transformers mysteriously "sound" bad, both in pre and power amp applications.
In conclusion i have come to believe that an optimal power supply exists for each application, depending on current draw and type of circuit (class A?) supplied. Whether this "best sounding" PS configuration is one which suppresses or excites resonance i really don't know.
- adding capacitance above a certain value always makes an audible difference, both in preamps and power amps. Usually, but not always, this sounds worse to my ears. Depending on taste, choice of speakers and music i can imagine some listeners would prefer the sound.
- strangely, the same is true for dc heater supplies (indirectly heated tubes).
- changing tranformer type or power rating is also clearly audible. Usually, but not always higher power rating is preferable. Some very large transformers mysteriously "sound" bad, both in pre and power amp applications.
In conclusion i have come to believe that an optimal power supply exists for each application, depending on current draw and type of circuit (class A?) supplied. Whether this "best sounding" PS configuration is one which suppresses or excites resonance i really don't know.
GRollins said:
I have a habit of scratching my head, hard to get rid of habits 😉
Yes, principles are the same but currents differ a lot (my scratch)
I think JC is right with the implementation of a follower supplying each stage, this must be one of the least reactive supplies.
anatech said:
You know, for someone who takes every possible opportunity to tell people that you are/were a service tech in an attempt to establish "cred," you show some astonishing lapses in practical knowledge.
Bill and Lew are real people using real parts in the real world, Real regulators are imperfect. Your faith in the regulator's behavior being perfect is touching, but misplaced.
4fun said:
Yes, given that a follower has lower Zout than other connections, I agree that it's a good choice when you're trying to build a voltage source.
Some folks use shunt regulators wherein the collector/Drain is presented to the load. A good case can be made for the use of a shunt regulator in the case of a circuit whose load sums to DC. It is not necessarily the best choice for other circuits.
I've read the arguments and I find them interesting. I have not tried shunt regulators yet so all I can say is that I have no opinion. When I get time, I'll take a whack at it.
Note that there's no reason not to use series and shunt regulators together. That opens up all sorts of interesting possibilities.
Grey
MikeBettinger said:
I'm not sure if they can be heard but I'm sure as hell they can be measured, 'cause I did it myself. At least two effects: eddy currents and magnetic core imbalances. Both translates into mains frequency harmonics, well into the KHz range. Both depend more on the transformer construction rather than the nominal power.
Wires positioning can dramatically impact the sound, due to magnetic field issues (nicely explained in D. Self's book).
It's not easy to qunatitatively separate these contributions, and it is even easier to attribute a certain (audible or measurable) effect to a wrong cause.
Hi syn08,
which order and which magnitude had the mains freq harmonics, and which tranformers were the best/worst performers?
Rüdiger
which order and which magnitude had the mains freq harmonics, and which tranformers were the best/worst performers?
Rüdiger
Onvinyl said:
Odd harmonics, take a look here: http://synaesthesia.ca/distortion.html the IMD measurements. Replacing the toroid with another model made these harmonics vanish in the noise floor level. This was a surplus toroid purchased from a local junk store, the good replacement was an original Plitron.
Unfortunately this thread which has been very interesting and educating in the past has lost its enchantment due to OT postings.
Hope those discussing amplifier PSU choose other threads which would be more appropriate or open a new thread for that matter.
Hope those discussing amplifier PSU choose other threads which would be more appropriate or open a new thread for that matter.
Unfortunately, many here think only of the forward filtering and not so much about the return path back to the power line.
What is 'ground'? How is is formed? How does the ground current that is sent there, return to the power line? I am serious with my questions. Hint: 'Ground' is not the earth.
hurrah,john curl said:
someone that uses different names for different purposes.
Would it help us if the word "ground" was banned from all electronics postings?
'Ground' routing is really hard for me to understand, let alone get right.
Especially when I try to use multiple regulators for a multiple stage preamp.
Noro states in his patent, that shunt regs separate the ground returns in a way that you have a separate loop transformer -> bridge -> caps -> regulator and a separate loop regulator -> load (your actual circuit)
If this is true, one would indeed have solved some serious issues.
I build a phone pre with shunt regs, but I have line hum coupled in and haven't found it's origin until now.
Rüdiger
Especially when I try to use multiple regulators for a multiple stage preamp.
Noro states in his patent, that shunt regs separate the ground returns in a way that you have a separate loop transformer -> bridge -> caps -> regulator and a separate loop regulator -> load (your actual circuit)
If this is true, one would indeed have solved some serious issues.
I build a phone pre with shunt regs, but I have line hum coupled in and haven't found it's origin until now.
Rüdiger
One way of looking at ground return is that the power supply filter caps double as caps in series with the loudspeaker load, much like a crossover component. Remember, a center tap has finite resistance, and potential nonlinearity also.
john curl said:
With respect to the AC line, there is no "ground" per se. You have "hot" and "neutral" but even those assignments are for all intents and purposes arbitrary and really only for safety purposes.
Ultimately what you have is a closed loop circuit formed by the secondary of the AC power line transformer and the primary of the power transformer in the component's power supply.
And since it's AC, the notion of "return" is rather ambiguous as currents alternately flow in both directions.
Which again leaves me wondering just what you're talking about when you refer to the "return path back to the power line."
se
- Status
- Not open for further replies.