Standard TX/Rect/Caps Ain't Good Enough....
Dan.
Yes, I bought the Kikusui as a tool for this kind of proving/improving.
Dan.
Not Magical To Me...
No 'floobydust' explanations required....just proving what is already well known, but perhaps not fully realised by some....that standard DC supplies are not fully fit for purpose.
The veil of IM that I am hearing is pretty normal really....changing power cords, ferrite filters etc ought to subtley change the nature of the IM, ditto AC frequency.
So far this is with 'clean' 0.3% THD 240V....I am keen to try adding all manner of junk once I get hold of a GPIB interface.
Dan.
The audibility is real, very real....no need to question it.
No 'floobydust' explanations required....just proving what is already well known, but perhaps not fully realised by some....that standard DC supplies are not fully fit for purpose.
The veil of IM that I am hearing is pretty normal really....changing power cords, ferrite filters etc ought to subtley change the nature of the IM, ditto AC frequency.
So far this is with 'clean' 0.3% THD 240V....I am keen to try adding all manner of junk once I get hold of a GPIB interface.
Dan.
Last edited:
+\-5%! Or 10% min to max on voltage 21% power increase.
A toroid into a single bridge with a single pair of big electrolytic filter capacitors is the combination most sensitive to power line noise.
Then the BJT monolithic amplifier chips are high on the list of not dealing well with EMI.
Its not quite that simple- the phase of the harmonics matters. But powerline harmonics are rarely less than 3%, without exception flattened peaks. The DC voltage on the power supply is proportional to the peak voltage of the AC, something that seems impossible to communicate when an uninformed gets a true RMS meter. . . The true RMS value only matters to the resistive components and how hot they get. Measure the current harmonics into your load to really see how bad a load an audio amp is.
Demian
He isn't using a real power line!
5% is the typical "problem" limit for an AC line. So if you use the 5% distortion and phase shift it from flat to peak you will get the 10%.
The issue was with an AC test source not what is expected on a real line.
So if you keep posting additional info some folks might get a clue
He isn't using a real power line!
5% is the typical "problem" limit for an AC line. So if you use the 5% distortion and phase shift it from flat to peak you will get the 10%.
The issue was with an AC test source not what is expected on a real line.
So if you keep posting additional info some folks might get a clue
Last edited:
We Are Considering Typical Minimalist PSU Arrangements....
I have made a subjective observation, and I now have the tools to measure what is happening, in the near future.
What I find interesting is the magnitude of power supply frequency/musical scale IM relationship/dependency.
All to be expected, I should think.
Dan.
Indeed, that is why my approach is quite the opposite.
I have made a subjective observation, and I now have the tools to measure what is happening, in the near future.
What I find interesting is the magnitude of power supply frequency/musical scale IM relationship/dependency.
All to be expected, I should think.
Dan.
50Hz/60Hz Causes Listening Fatigue....
Agreed the Behringer (standard consumer PSU topology) PSU arrangement is hopelessly transparent to incoming AC borne noise, and the 3886 amps are not particularly immune to DC rail noise also.
So these active speakers reasonably duplicate the supply dependency of any normal (suboptimal) Pioneer, Sony, Yamaha etc amplifier or receiver....DIY stuff too.
More listening results this afternoon on classical fm broadcast.
1 - Wall Socket Power - veiled, lack of fine detail, musically 'wrong'.
WRT to 1 -
2 - Kikusui 50Hz - less veiled, bit more detail, musically slightly 'wrong'.
3 - Kikusui 60Hz - less veiled, bit more detail, musicially slightly 'wrong', but slightly different to 2.
4 - Kikusui 55Hz - much less veiled, bit more detail, musically 'right'.
Bass 'grounded' not 'floating', mids clearer and more 'pleasant', overall 'calmer' and more 'dynamic'.
Next step is to run a selection of blues and rock stuff.
Dan.
I have not FFT'd the Kikusui 'clean' output yet, so I don't yet know how clean it really is, but no doubt much better than standard house power.
Agreed the Behringer (standard consumer PSU topology) PSU arrangement is hopelessly transparent to incoming AC borne noise, and the 3886 amps are not particularly immune to DC rail noise also.
So these active speakers reasonably duplicate the supply dependency of any normal (suboptimal) Pioneer, Sony, Yamaha etc amplifier or receiver....DIY stuff too.
More listening results this afternoon on classical fm broadcast.
1 - Wall Socket Power - veiled, lack of fine detail, musically 'wrong'.
WRT to 1 -
2 - Kikusui 50Hz - less veiled, bit more detail, musically slightly 'wrong'.
3 - Kikusui 60Hz - less veiled, bit more detail, musicially slightly 'wrong', but slightly different to 2.
4 - Kikusui 55Hz - much less veiled, bit more detail, musically 'right'.
Bass 'grounded' not 'floating', mids clearer and more 'pleasant', overall 'calmer' and more 'dynamic'.
Next step is to run a selection of blues and rock stuff.
Dan.
Come to papa ...!!
Very amusing - these are precisely the units I have here, one of my Round Tuit projects, to sort them out. Were the best of the bunch when I did an auditioning tour of small pro actives, but still have pretty severe failings in raw form - especially the usual, collapsing power supplies when asked for decent output levels.
So, IMO just about anything you do within a mile of these will make a difference ... will be interesting ...
No, you just made clear that you've decided what the answer is without doing a real experiment, and have proclaimed that you already know the origin of your observation without bothering to eliminate any other possibility. Congratulations, you have the makings of a fashion audio guru!
Hearing The Truths.....
I am deliberately refraining from further tweaking....yet.
In current configuration on high stands, with a quite decent active sub on a low stand, they go very well up to reasonable levels in a lively, tiled floor, rendered brick main room....clean, clear and musical, and no real colourations to speak of.
Interesting observation today listening to ABC Classic FM service....on all systems I have heard to date, the announcers sound wrongly boomy, and with an odd 'dynamics' characteristic.
With the power feed set to 55Hz, that characteristic mostly disappears and the announcers, male and female sound much more natural, pleasingly so.
I also played Helen Reddy (I Don't Know How To Love Him) this arvo...with 55Hz supply her voice is quite different and pleasingly rather more natural....her solo voice becomes cleaner, clearer, calmer and sits tonally a bit lower.
Next experiments will be playing some bands that I have mixed and know intimately the natural speaking voices of the vocalists.
Dan.
Yeah Frank, have had them for years, did some tweaking (internal cables and damping the plastic front) back then, but have since been in storage...rebooted them only recently.
I am deliberately refraining from further tweaking....yet.
In current configuration on high stands, with a quite decent active sub on a low stand, they go very well up to reasonable levels in a lively, tiled floor, rendered brick main room....clean, clear and musical, and no real colourations to speak of.
Interesting observation today listening to ABC Classic FM service....on all systems I have heard to date, the announcers sound wrongly boomy, and with an odd 'dynamics' characteristic.
With the power feed set to 55Hz, that characteristic mostly disappears and the announcers, male and female sound much more natural, pleasingly so.
I also played Helen Reddy (I Don't Know How To Love Him) this arvo...with 55Hz supply her voice is quite different and pleasingly rather more natural....her solo voice becomes cleaner, clearer, calmer and sits tonally a bit lower.
Next experiments will be playing some bands that I have mixed and know intimately the natural speaking voices of the vocalists.
Dan.
Last edited:
Curious, any sense of whether 55Hz is a magic no. or not? In other words, if you tried, say, 53, or 57 Hz, whether there is a "falling off", on either side, in quality so to speak ...
And, is it possible to dynamically adjust the frequency - while the track is playing to slowly move the frequency up or down?
And, is it possible to dynamically adjust the frequency - while the track is playing to slowly move the frequency up or down?
Last edited:
What would be interesting is some screen shots of the outputs of the PSU ripple etc at the various frequencies. Then some test tones and view the intermodulation from the amp.
Though again, why use a linear with an operating frequency right in the audio range (50-60Hz) when with a good SMPS you can move the noise out of the audio range, and use smaller components to filter the rest of the noise.
Though again, why use a linear with an operating frequency right in the audio range (50-60Hz) when with a good SMPS you can move the noise out of the audio range, and use smaller components to filter the rest of the noise.
Tuneable Hi-Fi...
I have six presets memorised, (50, 55, 60, 100, 110 and 120 Hz) and any of these can be tweaked in 0.01Hz steps in real time....and then memorised to another preset if desired.
I did tweak the 55Hz upwards half a Hz or so whilst listening to classical fm today, and found minor variation, but no nearly so significant as the step from 50Hz.
The effect I am finding is sort of a bit like fine tuning an AM radio....50Hz and 60Hz sound 'off' whilst 55Hz or near abouts sounds the 'rightest', with least 'wrong' artefacts.
Dan.
I have six presets memorised, (50, 55, 60, 100, 110 and 120 Hz) and any of these can be tweaked in 0.01Hz steps in real time....and then memorised to another preset if desired.
I did tweak the 55Hz upwards half a Hz or so whilst listening to classical fm today, and found minor variation, but no nearly so significant as the step from 50Hz.
The effect I am finding is sort of a bit like fine tuning an AM radio....50Hz and 60Hz sound 'off' whilst 55Hz or near abouts sounds the 'rightest', with least 'wrong' artefacts.
Dan.
Musical Frequency Power Supply Makes Sense To Me....
I well agree that properly done SMPS stages ought to be the solution to these dependencies that I am finding/revealing.
I have repaired, heard and used thousands of amps in my time, and not too many have grabbed my attention.
One amp that did grab my attention years ago in a hi-fi dealer show room was a Parasound that just sounded 'right'....JC's insistence/practice of seriously good regulation of input and driver stage supplies is well founded it seems.
This experimentation is really a look/see to verify my suspicions regarding the sonic effects of lousy conventional supplies, ie supply ripple causing beat notes/IM against the musical scale causing a whole cloud of non musically related 'junk', in practice most noticeable in bass up through and past the vocal ranges.
Dan.
Sorry guys, this is strictly armchair subjective testing for the moment...I am about to move house so my workbench stuff is packed down currently.
I well agree that properly done SMPS stages ought to be the solution to these dependencies that I am finding/revealing.
I have repaired, heard and used thousands of amps in my time, and not too many have grabbed my attention.
One amp that did grab my attention years ago in a hi-fi dealer show room was a Parasound that just sounded 'right'....JC's insistence/practice of seriously good regulation of input and driver stage supplies is well founded it seems.
This experimentation is really a look/see to verify my suspicions regarding the sonic effects of lousy conventional supplies, ie supply ripple causing beat notes/IM against the musical scale causing a whole cloud of non musically related 'junk', in practice most noticeable in bass up through and past the vocal ranges.
Dan.
Max, thanks for your courage in reporting what you have found, so far.
It does seem that your audio equipment is probably part of the reason that you can hear this difference, but it is typical of much that is thought here as being as 'good' as anything else in audio reproduction by many.
Looking at the schematics, I see 'design to a price' and this leaves little 'guard-band' to get it 'right'. Parasound makes a lot of products in a similar manner, but not MY designs. That is why my Parasound designs cost more than many other Parasound products. I, personally, don't get that much from my efforts that would raise the price significantly, BUT it is my insistence of a standard of design that often eliminates the very 'cheapest' parts and circuitry to just get the job done, makes these products to sometimes standout when a difficult load or power supply condition is encountered. Unfortunately this costs more money to make the product, and this is multiplied by the usual markups to a much higher end price.
I suspect that a number of other audio products will be sensitive to this line frequency change. I would doubt that more 'hi end, well engineered' products would show the same change with frequency, but it is interesting to note. Who knows, until we try?
It does seem that your audio equipment is probably part of the reason that you can hear this difference, but it is typical of much that is thought here as being as 'good' as anything else in audio reproduction by many.
Looking at the schematics, I see 'design to a price' and this leaves little 'guard-band' to get it 'right'. Parasound makes a lot of products in a similar manner, but not MY designs. That is why my Parasound designs cost more than many other Parasound products. I, personally, don't get that much from my efforts that would raise the price significantly, BUT it is my insistence of a standard of design that often eliminates the very 'cheapest' parts and circuitry to just get the job done, makes these products to sometimes standout when a difficult load or power supply condition is encountered. Unfortunately this costs more money to make the product, and this is multiplied by the usual markups to a much higher end price.
I suspect that a number of other audio products will be sensitive to this line frequency change. I would doubt that more 'hi end, well engineered' products would show the same change with frequency, but it is interesting to note. Who knows, until we try?
Thanks John.
Yes, I fully agree that the Behringer's lack of engineering renders it prone to PS ripple causing IM at any signal level.
On first inspection the schematic looks to be quite ideal, and ought to work reasonably flawlessly......ditto just about any other piece of mid-fi gear, however, in common practice this is not quite the case.
The current state of subjective experimentation has opened my ears to how much damage is done by typical standard power supplies, and the importance of very low noise/ripple supplies and high PSRR circuitry.
It is indeed interesting to note that just changing AC frequency can instantly polish a turd.
Regards, Dan.
Yes, I fully agree that the Behringer's lack of engineering renders it prone to PS ripple causing IM at any signal level.
On first inspection the schematic looks to be quite ideal, and ought to work reasonably flawlessly......ditto just about any other piece of mid-fi gear, however, in common practice this is not quite the case.
The current state of subjective experimentation has opened my ears to how much damage is done by typical standard power supplies, and the importance of very low noise/ripple supplies and high PSRR circuitry.
It is indeed interesting to note that just changing AC frequency can instantly polish a turd.
Regards, Dan.
To truly test the hypothesis of this being power-supply related it's going to be necessary to separate the system aspects as alluded to by Demian from the strictly PS ones.
I've seen some cases where, in the attempt to come up with a power rating for a cost-challenged powered speaker, one could slowly vary the frequency of a test signal near the powerline frequency or harmonics, and wait for a favorable phase relationship to make the reading. Of course this is cheating --- but the difference was not small. But it does suggest a phenomenon that could be related. The numerology of the 55Hz preference is puzzling, but if it holds up with further tests, depending on the source material (a lot of bass guitar?) there may be some technically analyzable basis.
I've seen some cases where, in the attempt to come up with a power rating for a cost-challenged powered speaker, one could slowly vary the frequency of a test signal near the powerline frequency or harmonics, and wait for a favorable phase relationship to make the reading. Of course this is cheating --- but the difference was not small. But it does suggest a phenomenon that could be related. The numerology of the 55Hz preference is puzzling, but if it holds up with further tests, depending on the source material (a lot of bass guitar?) there may be some technically analyzable basis.
- Status
- Not open for further replies.