Re: Affect of Power Supply on Soundstage Width
Deals with most considerations like single or dual transformators, rectifier, capacitors and more.
Part 2 even mentions Nelson Pass's opinion on diode type selection in rectifier ( see my previous post here ).
Pass recommends at least 2-3 times the max Watt output for the VA rating in a trafo driving one channel.
I know some designers use as much as 4 times.
http://www.tnt-audio.com/clinica/solidstate.html
Good reading, foxyb!
Regarding capacitance across rectifier diodes I havent found much on the web.
But I know I have seen ceramic capacitors in several power supply circuits.
I think they usually are in 100-220nF value range, if remembering correctly.
Also seen sometimes small value resistors in series with these capacitors, that sits across each of 4 diodes in a bridge.
There is a very good 3 Parts Article at tnt-audio.com Solid State Power Amplifier Supply.foxyb said:
Deals with most considerations like single or dual transformators, rectifier, capacitors and more.
Part 2 even mentions Nelson Pass's opinion on diode type selection in rectifier ( see my previous post here ).
Pass recommends at least 2-3 times the max Watt output for the VA rating in a trafo driving one channel.
I know some designers use as much as 4 times.
http://www.tnt-audio.com/clinica/solidstate.html
Good reading, foxyb!Regarding capacitance across rectifier diodes I havent found much on the web.
But I know I have seen ceramic capacitors in several power supply circuits.
I think they usually are in 100-220nF value range, if remembering correctly.
Also seen sometimes small value resistors in series with these capacitors, that sits across each of 4 diodes in a bridge.
Now I mean class AB case : if is amplifier sensitive to disturbing at the rails, in case of common supply will be sensitivity higher ( + crosschannel intermodulation ). Commonly, distortion must be low as possible, if we can good ( and natural ) stereo imagination, 'cos all distortions mask it. I don't like so called " musical " amps, 'cos it is only different word for distortion...
Upupa,
Yes, I thought you meant that. I missed this effect in my first post, where I only thought of distorsion products on the signals in the amp, sneaking through the PSU. However, in the case of class AB, we will get high-level distorsion products on the rails, so I agree this might cause problems through PSU coupling. I guess this answers my original question, unless somebody has further ideas about this.
I also agree with you that an amp should have low distorsion, although I guess not only the figures matter, but also what the spectrum looks like. However, one should of course not deliberately increse certaian harmonics to try masking others. The distorsion must be low enough for a start, before it is meaningful to discuss the spectra.
Yes, I thought you meant that. I missed this effect in my first post, where I only thought of distorsion products on the signals in the amp, sneaking through the PSU. However, in the case of class AB, we will get high-level distorsion products on the rails, so I agree this might cause problems through PSU coupling. I guess this answers my original question, unless somebody has further ideas about this.
I also agree with you that an amp should have low distorsion, although I guess not only the figures matter, but also what the spectrum looks like. However, one should of course not deliberately increse certaian harmonics to try masking others. The distorsion must be low enough for a start, before it is meaningful to discuss the spectra.
This is a factor to take into consideration when designing power supply.Upupa Epops said:
How sensitive is amp to power supply.
There are amplifiers of many types, whether they are class AB or A.
Some are sensitive and some are not so sensitive to supply garbage.
Power Supply Rejection Ratio, PSRR, is a value to discribe this.
If low PSRR of amp, we need to spend more efforts into protect from influence from supply.
Otherwise we can use a slightly more simple solution ( save some money ).
How can we know an amp is sensitive to disturbance from supply?
If we can not measure.
I am not sure how to know this, whilst an experienced amplifier constructor knows it by heart.
How about this JLH 80 Watt MOSFET amp of foxyb?
What would you say?
What does it need?
I couldnt find some schematic at the web.Upupa Epops said:
JLH 80 is mentioned in this topic, too:
http://www.diyaudio.com/forums/showthread.php?threadid=24054
Is sold as a kit by
WILLIAMS HART ELECTRONICS
audio kits supplied by Shaun Williams former Director of Wilmslow Audio.
http://www.williamshart.com/jlhampli.htm
hi,
does the amount of feedback applied to the amp or a regulated power supply make any difference to the percieved soundstage depth and width?
I don't have too much experience but what I have heard myself is that, amplifiers that have higher bandwidth like 250khz and use more feedback, present a flatter soundstage, with more detail and tightness. But the 3d depth seems more realistic when the bandwidth of an amplifier is around 50khz or so.
Thanks
Raja
does the amount of feedback applied to the amp or a regulated power supply make any difference to the percieved soundstage depth and width?
I don't have too much experience but what I have heard myself is that, amplifiers that have higher bandwidth like 250khz and use more feedback, present a flatter soundstage, with more detail and tightness. But the 3d depth seems more realistic when the bandwidth of an amplifier is around 50khz or so.
Thanks
Raja
Hi,
JLH 80 is a mosfet output with both regulated line and output PSUs. It has NO miller comp caps. It has about 4 or 5 stability component networks spread around , one is adjustable.
Automatic shut down on detection of DC on output and other sensors. Both rails monitor each other and shut down in tandem.
I cannot remember much more since it is about 15 or 20 years since I read the construction article which appeared in Electronics Today if my memory is right.
Can anyone find/post a schematic?
JLH 80 is a mosfet output with both regulated line and output PSUs. It has NO miller comp caps. It has about 4 or 5 stability component networks spread around , one is adjustable.
Automatic shut down on detection of DC on output and other sensors. Both rails monitor each other and shut down in tandem.
I cannot remember much more since it is about 15 or 20 years since I read the construction article which appeared in Electronics Today if my memory is right.
Can anyone find/post a schematic?
Ever heard of Edeko's works and of Richard Brice's Book "Multimedia and virtual reality engineering" ?
As "high frequency intensity-derived images are perceived as wider than low frequency delayed images", Edeko, then Brice proposed ways to stabilise the apparent stereo sources.
Brice at page 239 of his book has a chapter called "The improvement of stereophonic sharpness by means of aperiodic, interchannel crosstalk." He shows a very simple circuit, that any tweaker can easily try, which introduces a slight diaphony at high frequenices.
It uses a stereo source of low impedance : in each channel, there is a 10 kOhm resistor in series. An RC series circuit (43 kOhm and 4n7) is connected between the two other ends of the 10 kOhm resistors. A high impedance circuit follows, input and output buffers can be useful.
This does more to improve stereophonic image stability than any fancy and expensive trick using 'special" components.
~~~~~~~~~Forr
§§§
As "high frequency intensity-derived images are perceived as wider than low frequency delayed images", Edeko, then Brice proposed ways to stabilise the apparent stereo sources.
Brice at page 239 of his book has a chapter called "The improvement of stereophonic sharpness by means of aperiodic, interchannel crosstalk." He shows a very simple circuit, that any tweaker can easily try, which introduces a slight diaphony at high frequenices.
It uses a stereo source of low impedance : in each channel, there is a 10 kOhm resistor in series. An RC series circuit (43 kOhm and 4n7) is connected between the two other ends of the 10 kOhm resistors. A high impedance circuit follows, input and output buffers can be useful.
This does more to improve stereophonic image stability than any fancy and expensive trick using 'special" components.
~~~~~~~~~Forr
§§§
AndrewT said:
Funny, this topic comes now with exactly what we wantedUpupa Epops said:
a JLH 80 Watt schematic!
JLH 80w mosfet power amplifier - modifying it
Link to schematic:
CORTEZ
"I dont understand. Why would this improve anything ?"
I think this was clear enough :
FORR
"high frequency intensity-derived images are perceived as wider than low frequency delayed images"
Having a sound not centered in the middle of the stereo stage, the low frequencies do not appear to emanate from the same location as the high frequencies. This is due to human physiology.
To correct this, for a listener at 2.3 m, Edeko used an horizontal array of bass/middle/treble devices separated by 15 cm, the treble towards the inside. Brice said his circuit does the trick. Have a try.
~~~~~ Forr
§§§
"I dont understand. Why would this improve anything ?"
I think this was clear enough :
FORR
"high frequency intensity-derived images are perceived as wider than low frequency delayed images"
Having a sound not centered in the middle of the stereo stage, the low frequencies do not appear to emanate from the same location as the high frequencies. This is due to human physiology.
To correct this, for a listener at 2.3 m, Edeko used an horizontal array of bass/middle/treble devices separated by 15 cm, the treble towards the inside. Brice said his circuit does the trick. Have a try.
~~~~~ Forr
§§§
Forr,
that is interesting, but since this is a psychoacoutical phenomenon, having nothing per se to do with audio electronics, it seems we might want the effect the filter i supposed to reduce. If I am sitting in a concert hall, listening to an orchestra, then I would get a somwhat skewed image with higher frequency instruments appearing further to the sides than they are, if that theory is correct. That is, this would be the expected image, so why should I want to change that when listening to a recording?
that is interesting, but since this is a psychoacoutical phenomenon, having nothing per se to do with audio electronics, it seems we might want the effect the filter i supposed to reduce. If I am sitting in a concert hall, listening to an orchestra, then I would get a somwhat skewed image with higher frequency instruments appearing further to the sides than they are, if that theory is correct. That is, this would be the expected image, so why should I want to change that when listening to a recording?
Hi Christer,
In concert, a sound location is real. At home, it is intended that stereo virtually reconstructs its exact location using two loudspeakers. The Brice's circuit should diminish ambiguity between bass and treble different locations for a same sound due to the psychoacoutical reaction to the two channels / two loudpseakers technique.
~~~~~~~ Forr
§§§
In concert, a sound location is real. At home, it is intended that stereo virtually reconstructs its exact location using two loudspeakers. The Brice's circuit should diminish ambiguity between bass and treble different locations for a same sound due to the psychoacoutical reaction to the two channels / two loudpseakers technique.
~~~~~~~ Forr
§§§
feedbackUpupa Epops said:JLH 80 : This schematic is obviously uncorrect : Too high value of feedback resistor, changed voltage at VAS and output stage, too high value of input pot...
yes, what I can see gain is more than 100, which is a terribly high value
can not explain this
Strange!
voltage
Uses regulated 50 Volt for input/VAS
Unregulated most likely for output stage
input impedance
100-67 kOhm (100//150k)
This is an old circuit. Was usual with higher input impedance those days.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.