Actually, micro-modulation of the rail by the entire AC-Transformer-rectification system is a critical point. Since the ear hears via transients, those seeming tiny micro-draw/charge issues via a fully connected standard linear power supply become CRITICAL when attempting 'high-end' reproduction. We aren't concerned with the 99.99% of the PS rail that we have, we are only concerned with rail modulation noise that passes through the entire chain when the diode either switches or is conducting.
That's it.
So removing the linear PS system and going to Battery Supply is ALSO >not< the better solution, as one should remember that Jack Bybee invented similar gear as the Bybee devices for the military to remover BATTERY noise in sensitive sonar systems and other similar gear.
There is less noise with a (pre-charged disconnected from AC charge mode-and switched) pure capacitor set-up vs that of any system that has active devices or is battery powered and direct coupled.
That's it.
So getting this done 'as good as possible' -is the key here.
After getting rid of the AC linear PS generated pass through noise... We are only concerned with the micro-molecular (active device generated and battery originated) noise that rides along with micro level peak V and I value changes.
That's it.
This is HIGH-END as solutions go. IF you see no benefit to this sort of idea, as a 'base thought and consideration', then you might consider that you are not really all that concerned with the best that might be made by man and stick to mediocre and average when it come to your concerns about 'audio' and associated thinking.
As well, no-one ever said that a bank of HQ capacitors 'done right' cannot be charged LESS than 1/60th of a second.
If done right, and switched correctly, one might consider with a low power consumption circuit needing a boosting of the rial maybe once a second -- or less.
It is not without reason that many folks who have heard it, thought that the Cary 300B monoblocks with the optional capacitor bank in use, thought that THE best sonics they had EVER heard in their lives.....was in the 10-15 minute time period it took for the capacitor bank to discharge--after amplifier turn-off.
That's it.
So removing the linear PS system and going to Battery Supply is ALSO >not< the better solution, as one should remember that Jack Bybee invented similar gear as the Bybee devices for the military to remover BATTERY noise in sensitive sonar systems and other similar gear.
There is less noise with a (pre-charged disconnected from AC charge mode-and switched) pure capacitor set-up vs that of any system that has active devices or is battery powered and direct coupled.
That's it.
So getting this done 'as good as possible' -is the key here.
After getting rid of the AC linear PS generated pass through noise... We are only concerned with the micro-molecular (active device generated and battery originated) noise that rides along with micro level peak V and I value changes.
That's it.
This is HIGH-END as solutions go. IF you see no benefit to this sort of idea, as a 'base thought and consideration', then you might consider that you are not really all that concerned with the best that might be made by man and stick to mediocre and average when it come to your concerns about 'audio' and associated thinking.
As well, no-one ever said that a bank of HQ capacitors 'done right' cannot be charged LESS than 1/60th of a second.
If done right, and switched correctly, one might consider with a low power consumption circuit needing a boosting of the rial maybe once a second -- or less.
It is not without reason that many folks who have heard it, thought that the Cary 300B monoblocks with the optional capacitor bank in use, thought that THE best sonics they had EVER heard in their lives.....was in the 10-15 minute time period it took for the capacitor bank to discharge--after amplifier turn-off.
KBK said:Actually, micro-modulation of the rail by the entire AC-Transformer-rectification system is a critical point. Since the ear hears via transients, those seeming tiny micro-draw/charge issues via a fully connected standard linear power supply become CRITICAL when attempting 'high-end' reproduction. We aren't concerned with the 99.99% of the PS rail that we have, we are only concerned with rail modulation noise that passes through the entire chain when the diode either switches or is conducting.
That's it.
So removing the linear PS system and going to Battery Supply is ALSO >not< the better solution, as one should remember that Jack Bybee invented similar gear as the Bybee devices for the military to remover BATTERY noise in sensitive sonar systems and other similar gear.
There is less noise with a (pre-charged disconnected from AC charge mode-and switched) pure capacitor set-up vs that of any system that has active devices or is battery powered and direct coupled.
That's it.
So getting this done 'as good as possible' -is the key here.
After getting rid of the AC linear PS generated pass through noise... We are only concerned with the micro-molecular (active device generated and battery originated) noise that rides along with micro level peak V and I value changes.
That's it.
This is HIGH-END as solutions go. IF you see no benefit to this sort of idea, as a 'base thought and consideration', then you might consider that you are not really all that concerned with the best that might be made by man and stick to mediocre and average when it come to your concerns about 'audio' and associated thinking.
The AC line and the mains transformer are both highly inductive. They already work as current pulse sources when there is a capacitor to charge.
Myths, myths, myths...
Ignorance, Ignorance, Ignorance.
Have it your way. For yourself, that is. That's fine. But don't expect me to pay attention. I'm too busy making things work better.
50% of audio design is understanding human hearing function.
I suggest you look into it. Personal experience and tests are the preferred workable methodology, when it comes to constructing a personal understanding of the situations involved.
Book learnin' don't do squat, here.
Good luck to you.
Have it your way. For yourself, that is. That's fine. But don't expect me to pay attention. I'm too busy making things work better.
50% of audio design is understanding human hearing function.
I suggest you look into it. Personal experience and tests are the preferred workable methodology, when it comes to constructing a personal understanding of the situations involved.
Book learnin' don't do squat, here.
Good luck to you.
Nico Ras said:
Not much, only confirming that such an arrangemet will most likely not have any positive impact on the overall performance, but only add complexity and cost.
Unless you "are only concerned with the micro-molecular (active device generated and battery originated) noise that rides along with micro level peak V and I value changes", a fan of the Bybee devices and have a problem with Shannon theorems on noise.
For an introduction to switched capacitor filters, Google is your friend. For a more in depth understanding, you need to go back to school; switched capacitors circuitry theory and practice are covered in one semester and needs a solid math and electronic devices foundation to follow.
Just stumbled at this while broswing the forum...
http://www.diyaudio.com/forums/showthread.php?postid=1698213#post1698213
This post contains a link to a paper on EMC considerations. Science rather than myth.
http://www.diyaudio.com/forums/showthread.php?postid=1698213#post1698213
This post contains a link to a paper on EMC considerations. Science rather than myth.
syn08 said:
Of course it is not as good - switching ripple (especially with half wave rectification and switching at line frequency
)If you want to reject mains born high frequency interference, then you do that right where the mains lead enters the equiptment case (to prevent internal radiation just for a start).
These are a good and relatively cost effective place to begin, and do a good job over a reasonable spectrum. (have worked well for me over the years in sensitive (much more than audio) instrumentation in some very noisy environments).
http://www.farnell.com/datasheets/76075.pdf
An externally hosted image should be here but it was not working when we last tested it.
Cheers,
Glen
Eva said:
Hi Eva, you just don't get the point do you?
Roender wants to try something new, radically, different, a paradigm shift, lateral thinking, R&D call it what you will.
Whether it will work or not it just does not matter now, it is a concept. If it does work you will probably want to claim some stake in it because of some totally off mark comment, but hey I guess you are welcome to it.
We all know about EMC, RFI and EMI, we know about LISN and all the rest, we know the VDO, CISPR an MIL461 & 462 standards of measurement, heck I made some equipment that you probably use to meausre it.
Now give us a break, go back to sleep and let Roender do his thing.
And yes I know about switched capacitors.
If some people here lay claim that they just listen to things instead of measuring and calculating, well they are lying.
Nico
G.Kleinschmidt said:
Of course it is not as good - switching ripple (especially with half wave rectification and switching at line frequency
If you want to reject mains born high frequency interference, then you do that right where the mains lead enters the equiptment case (to prevent internal radiation just for a start).
These are a good and relatively cost effective place to begin, and do a good job over a reasonable spectrum. (have worked well for me over the years in sensitive (much more than audio) instrumentation in some very noisy environments).
http://www.farnell.com/datasheets/76075.pdf
An externally hosted image should be here but it was not working when we last tested it.
Cheers,
Glen
Hi Glen,
thank you for your input. I guess Roender will probably include that in his power supply application for good measure.
I do not know where this thread is going, I am just curios, it makes the gray matter work a bit (not that there is much left)
There are some interesting points made by some to consider. Besides I had nothing better to do today because I am on well earned leave.
Tomorrow I will utilize my day much better walking to the beach and swallowing beers while watching the holidaymakers infest our town.
As a matter of interest our beaches are allowed topless which makes for real fun watching these +30 year old hags with over-sized udders thinking they still have it, plopping along while jumping and diving after a volley ball.
Of course there are exceptions entering the daily miss toti and wet or no T-shirt competitions. But they skinny and firm and not nearly as much fun to watch.
Cheers
Nico
Nico Ras said:
If I would be Mihai I would certainly be embarassed by your kind of advocacy. But I am not, so all I will do is to sit back and have fun
I've learned that an insult coming from a GEB member should be treated as a trophy, so keep them coming BTW, if you would have the foggiest idea about switched capacitors filters you would recognize the concept up front and figure out yourself, before posting, what would be the impact of the switching ripple (among others), either with four switches or with a single SPDT switch (as in the original schematic) flipping at mains frequency.
Meanwhile, back at the ranch, I think the Linear Technology LT1043 (forgive me if I've got the wrong number) is/was a switched cap building block. It's too low powered for this application, but IMO it would be useful to look at all the LT application notes for things you can do with these switched cap designs. IMO again, the noise issues are manageable and you can get the isolation you want, but the loss of energy on charging, and discharging limits this technique to fairly low powered devices. Like preamps and such. I'd certainly give it a try and see how far you can take it.
Conrad Hoffman said:
Hi Conrad,
it certainly will. I have looked at an application since I have four LMF100 switch capacitor filters from some time ago, but they are useless for transferring enough energy for any audio power application as power source.
I think what Roender wanted to achieve was to simulate or create a battery like power source eliminating the noise generated by the chemical processesinside. (I am guessing.)
I have done a dual power supply on prototype board last night to test the theory and it certainly has some potential.
I used relatively small caps in the trail and loaded it by only a few hundred ohm. Unloaded it is "noiseless" on the scope down to 2mV/div, but the ripple is horrible under even minuscule loads.
One interesting thing is that the ripple is sinusoidal, not what I expected (one would think saw tooth) and there are a good few harmonics present at only a few dB down.
One other thing that I noticed is that the transistors we use in audio applications don't like switching highly capacitive loads.
I wiped a pretty hefty FETs last night when manually switching between two 22 000 uF low ESR caps. It died immediately, in fact the plastic case popped. So some form of soft start would be necessary to make it work reliably.
My interest in this circuit does not necessarily lie with audio but one of my companies specializes in battery charging and I am thinking transformerless "all-type" battery charging.
Particularly along the lines that the switched capacitor does isolate the mains from the battery and the amount of energy transferred can be determined exactly using a microprocessor looking at the battery dv/dt, temperature rise and regulating the current flow into the battery by means of PWM and zero crossing detection.
Obviously a soft start would be implemented from zero crossing but the type of semi-conductors we use is slightly beefier than what is used in audio.
One could probably call this an SMPS of some sort but what I am interested in is that capacitors are much cheaper than copper and steel and you can probably realize any voltage without having to wind different transformers for different chargers and use the exact same controller.
I know this may sound dangerous, but what is between you and a faulty mains transformer anyway?
So my conclusion is that there is a practical application but whether it is mains noise free is of no consequence for the application that I am became aware of two days ago.
I raised some excitement among some of my engineers and we will be brainstorming on the beech with several beers later this morning (while admiring floppy udders.)
But thanks for your contribution.
Kindest regards
Nico
When someone is ...
- concerned about low-level noise riding on the PSU rails
- willing to use additional active parts
- accepting some loss in efficiency
... then why not use a series regulator in the first place ?
OTOH maybe the use of active rectification might be a solution to generate less switching noise.
Regards
Charles
- concerned about low-level noise riding on the PSU rails
- willing to use additional active parts
- accepting some loss in efficiency
... then why not use a series regulator in the first place ?
OTOH maybe the use of active rectification might be a solution to generate less switching noise.
Regards
Charles
Nico,
You must refrain from attacking our other members in this manner. I have removed the posts that add zero useful content to this thread. Any repeat of this will not be greeted with positive actions from the moderating group.
Besides, posts like the ones I have trimmed out do not reflect well on you. Not a great way to start off a new year if you ask me.
One thing I would like to stress is that Eva is a well regarded member here. She truly tries to be helpful and is normally correct. You would do well to listen when she talks. Your sexist remarks went over like a lead balloon, not funny.
-Chris
Nico Ras wrote
Technically you could actually isolate mains by the use of flying capacitors but such a contraption would never get acceptance by any certification authority. It is TOO DANGEROUS !!!!
Discussing such things is not even allowed on this board mybe except from why one shouldn't do it !
A proper SMPS wouldn't be much larger and would even include line regulation.
The use of flying caps can be interesting for voltage conversion when there are no security issues (like doubling of a car's 12 V rails in order to power a PC).
Regards
Charles
Technically you could actually isolate mains by the use of flying capacitors but such a contraption would never get acceptance by any certification authority. It is TOO DANGEROUS !!!!
Discussing such things is not even allowed on this board mybe except from why one shouldn't do it !
A proper SMPS wouldn't be much larger and would even include line regulation.
The use of flying caps can be interesting for voltage conversion when there are no security issues (like doubling of a car's 12 V rails in order to power a PC).
Regards
Charles
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.