Nuuk said:
Probably is the answer here Lee as I don't have the time to test it myself. But it is the amp that is benefiting from an active stage preceding it so you should hear some advantage. If it also improves the impedance matching between the PS1 and amp, you will get a further improvement.
As always....thanks for your time and considerable knowledge...cheers Nick.
Lee
Preamp or not, that is the question.
If you look at the input to the chip you see The input resistor is connected to the feedback resistor. What this means is that anything more than a zero source impedance right at the resistor is effectively in the feedback path and will have a corresponding effect on the amp and the sound. Having a true low source impedance becomes disproportionately important. Cables and wiring to this point will also have their characteristics exaggerated. This must be why having a volume control very close is important. Also the volume control becomes part of that network if it isn’t buffered.
Roger
If you look at the input to the chip you see The input resistor is connected to the feedback resistor. What this means is that anything more than a zero source impedance right at the resistor is effectively in the feedback path and will have a corresponding effect on the amp and the sound. Having a true low source impedance becomes disproportionately important. Cables and wiring to this point will also have their characteristics exaggerated. This must be why having a volume control very close is important. Also the volume control becomes part of that network if it isn’t buffered.
Roger
Re: Preamp or not, that is the question.
OK silly questions from a novice...here we go....
I always wondered if the volume becomes part of the input network and therefore affects gain? Can you have a variable resistor on the feedback to balance things out depending on volume?
sx881663 said:
OK silly questions from a novice...here we go....
I always wondered if the volume becomes part of the input network and therefore affects gain? Can you have a variable resistor on the feedback to balance things out depending on volume?
Silly question?
No, not silly admitting to your self you don’t know something is the first step to knowledge! The more difficult second step is admitting it to others, like by asking questions. To my mind not knowing something is not a fault but failing to correct this is.
You could adjust the feedback resistor but why would you want to? You set the volume to your ear anyway so you already are compensating for the lost gain. You really wouldn’t want to bring these points of the circuit off board or add to them as they would pick up noise or could even cause the circuit to go unstable. The input node of the opamp is extremely sensitive! The result of having the vc as part of the input network is the lower end of the adjustment range doesn’t get much use. So the result is the log curve just becomes more severe and this gives you more resolution at the low end of the control, not necessarily a bad thing.
The more important consideration is putting some really inferior parts in the feedback path. I think the feedback it’s self may be helping overcome some of the limitations of cheap controls but it isn’t doing much to reduce noise. On the other end of the range, i.e. loud, the control is effectively out of the circuit so is a nonissue. If it isn’t loud enough when the control is wide open and the signal isn’t being clipped then you should adjust the feedback resistor (not input) to a higher value for more gain. Best setting is to reach clipping just before the wide open position. This should be done with a recording that is relatively quiet so you have some margin to play with.
Roger
No, not silly admitting to your self you don’t know something is the first step to knowledge! The more difficult second step is admitting it to others, like by asking questions. To my mind not knowing something is not a fault but failing to correct this is.
You could adjust the feedback resistor but why would you want to? You set the volume to your ear anyway so you already are compensating for the lost gain. You really wouldn’t want to bring these points of the circuit off board or add to them as they would pick up noise or could even cause the circuit to go unstable. The input node of the opamp is extremely sensitive! The result of having the vc as part of the input network is the lower end of the adjustment range doesn’t get much use. So the result is the log curve just becomes more severe and this gives you more resolution at the low end of the control, not necessarily a bad thing.
The more important consideration is putting some really inferior parts in the feedback path. I think the feedback it’s self may be helping overcome some of the limitations of cheap controls but it isn’t doing much to reduce noise. On the other end of the range, i.e. loud, the control is effectively out of the circuit so is a nonissue. If it isn’t loud enough when the control is wide open and the signal isn’t being clipped then you should adjust the feedback resistor (not input) to a higher value for more gain. Best setting is to reach clipping just before the wide open position. This should be done with a recording that is relatively quiet so you have some margin to play with.
Roger
Power supply capacitors
I wanted to share my experience with adding capacitors between the power supply and my Charlize. Thanks to all of you who have posted on this forum; I am constantly learning from your insights and wanted to contribute what I could.
Here's the short version: I think Charlize sounds better with caps across the power supply.
Here's the long version:
As mentioned earlier, I'm using the Skynet 8080 SMPS for its 12 volt output (from Nuuk's Decibel Dungeon). I keep it and my Charlize in separate enclosures. I built my speakers using Steve Deckert's "High Definition Tower" design at www.decware.com using the Fostex FE206E drivers as modified by Steve. In a nutshell: 96dB efficient full range 8" drivers with no crossover and a stainless steel phase plug for high frequencies.
My Charlize is now 6 months old with hundreds of hours on it (I don't know exactly how many). My original plan was to use two 4700 uF Nichicon VX(M) capacitors in parallel across the power supply wires inside Charlize's case. (I know these caps aren't audiophile grade, but they were the biggest and only capacitors sold at my local Radioshack. Nuuk recommended 10,000 uF and I figured 9,400 uF would be close enough.)
As a newbie, my soldering skills are lousy--and I was too eager to hear Charlize sing! I settled for simply wiring it up without the caps. Well, it worked great. I enjoyed it like this for 6 months.
Today, I decided to open up the case and solder the caps in.
Conclusion: I like the sound with the capacitors. Improved "attack" was the first thing I noticed. The sound is even more detailed. Bass is a little tighter (but had been tightly controlled to begin with). The caps deserve more break-in time, but I'll say this: I'm loving Charlize even more
-Peter
I wanted to share my experience with adding capacitors between the power supply and my Charlize. Thanks to all of you who have posted on this forum; I am constantly learning from your insights and wanted to contribute what I could.
Here's the short version: I think Charlize sounds better with caps across the power supply.
Here's the long version:
As mentioned earlier, I'm using the Skynet 8080 SMPS for its 12 volt output (from Nuuk's Decibel Dungeon). I keep it and my Charlize in separate enclosures. I built my speakers using Steve Deckert's "High Definition Tower" design at www.decware.com using the Fostex FE206E drivers as modified by Steve. In a nutshell: 96dB efficient full range 8" drivers with no crossover and a stainless steel phase plug for high frequencies.
My Charlize is now 6 months old with hundreds of hours on it (I don't know exactly how many). My original plan was to use two 4700 uF Nichicon VX(M) capacitors in parallel across the power supply wires inside Charlize's case. (I know these caps aren't audiophile grade, but they were the biggest and only capacitors sold at my local Radioshack. Nuuk recommended 10,000 uF and I figured 9,400 uF would be close enough.)
As a newbie, my soldering skills are lousy--and I was too eager to hear Charlize sing! I settled for simply wiring it up without the caps. Well, it worked great. I enjoyed it like this for 6 months.
Today, I decided to open up the case and solder the caps in.
Conclusion: I like the sound with the capacitors. Improved "attack" was the first thing I noticed. The sound is even more detailed. Bass is a little tighter (but had been tightly controlled to begin with). The caps deserve more break-in time, but I'll say this: I'm loving Charlize even more
-Peter
Charlize, my ramble...
Well, since this thread was one of the web's resources that got me
interested in the Charlize, I thought it might be time to share my
thoughts too. First post on this forum also, so here goes.
Ordered one Charlize from yeo about a month ago. Received it in a week
or so, quite quickly. I inquired at that time whether a 12V 2A smps
would be sufficient. Yeo suggested at least 2.5A.
Charlize sat in her primal wireless state for a while until I finally
swung my soldering iron in her direction. Brute-forced an aluminum
open enclosure, installed binding points, 2.1mm DC jack, input jack,
attached hookup wire in about a couple hours. No on-off switch and run
through a volume pot cum passive pre. Adequate test configuration I
suppose. The only suitable smps I had lying about was that for an
external Seagate usb 3.5" hdd, rated @ 12V 2A. It was summoned,
plugged in, for immediate ad hoc service. Measured output voltage to
preempt any nasty surprises.
Initial listening impressions: Firstly, it worked. No flame, no
smoke. Good start. I could concur - more or less - with a great deal
of the reports on the Tripath sound in general, and Charlize in
particular. For instance, on the following: surprisingly good control
of complex works, include orchestral/symphonic. Great retrieval of
detail. Very low background noise. Unusually good performance for
everything with vibrato - piano, electronic music and the like. Better
performance on such acoustic profiles however, than say on voice,
choral works, strings, and so on. On the Charlize in particular,
especially given its size, I did not detect the bass shyness that one
might have expected a priori from such a small form factor device. Nor
was the amp bright. Nevertheless, there was a slight -- comment peut
dire? -- edginess or sharpness to the overall musical presentation. I
supposed this was part of the Charlize (or even Tripath) signature
sound, one which some have described as holographic, hyperreal and so
on. In sum I was very pleased. A so-called keeper as some are wont to
say. The word "civilized" came to mind. Given the low noise floor,
level of detail and resolution, reasonable imaging and spatiality,
Charlize is rather well-suited to listening, critical listening even,
at low volumes. The amp certainly recommends itself for classical
music.
Charlize DIY.
A perusal of the numerous, but not innumerable URL's on Charlize
modifications, appears to suggest a rather discriminating choice of
components, such that initial tinkering might focus on, if not be
restricted to the input coupling caps: the 10uF Elna Cerafine
cylinders clad in sombre red. A comment by the designer himself in the
apparently superceded yahoo discussion group for diyparadise suggests
that 4.7uF BG N caps should be a reasonable replacement /
upgrade. Alternatively, one might follow conventional wisdom and get
rid of the electrolytics here altogether. I hadn't yet come across a
report of MKS caps replacing the Cerafines, so, since MKS is widely
available, cheap and small, for my test I opted to try Wima MKS02
3.3uF. The result: sound was somewhat smoother, perhaps more neutral;
there seemed to be less detail; overall there was a loss of presence
and the amp - in my view - certainly became less engaging. So the
Cerafines were rather good after all. The MKS caps were thus quickly
removed, the Cerafines returned.
Non-polar Black Gates might have been next, but then I opted for a
pair of Mundorf MCap Supremes, 2.2uF. They could be acquired more
quickly: here a case of local expediency over possible rationality,
understandable perhaps considering that other "boutique" caps have
been in transit for several weeks, remain in transit. The Mundorfs
arrived a few days later, were wired in. Result: the edginess is
gone; the detail remains. From a music reproduction point of view,
strings are now on par with piano - of significance imaginably
depending on one's listening preferences. No longer is the amp's sonic
signature idiosyncratic. As one might surmise, the Mundorfs have been
left in place. Such caps are big however, ca. 25x55mm, so if space is
an issue, BG's might be a fine choice too, given especially that the
stock Cerafines are very good.
Tthis is too long a post for too trivial remarks, but there
you have it, Charlize, my ramble...
Well, since this thread was one of the web's resources that got me
interested in the Charlize, I thought it might be time to share my
thoughts too. First post on this forum also, so here goes.
Ordered one Charlize from yeo about a month ago. Received it in a week
or so, quite quickly. I inquired at that time whether a 12V 2A smps
would be sufficient. Yeo suggested at least 2.5A.
Charlize sat in her primal wireless state for a while until I finally
swung my soldering iron in her direction. Brute-forced an aluminum
open enclosure, installed binding points, 2.1mm DC jack, input jack,
attached hookup wire in about a couple hours. No on-off switch and run
through a volume pot cum passive pre. Adequate test configuration I
suppose. The only suitable smps I had lying about was that for an
external Seagate usb 3.5" hdd, rated @ 12V 2A. It was summoned,
plugged in, for immediate ad hoc service. Measured output voltage to
preempt any nasty surprises.
Initial listening impressions: Firstly, it worked. No flame, no
smoke. Good start. I could concur - more or less - with a great deal
of the reports on the Tripath sound in general, and Charlize in
particular. For instance, on the following: surprisingly good control
of complex works, include orchestral/symphonic. Great retrieval of
detail. Very low background noise. Unusually good performance for
everything with vibrato - piano, electronic music and the like. Better
performance on such acoustic profiles however, than say on voice,
choral works, strings, and so on. On the Charlize in particular,
especially given its size, I did not detect the bass shyness that one
might have expected a priori from such a small form factor device. Nor
was the amp bright. Nevertheless, there was a slight -- comment peut
dire? -- edginess or sharpness to the overall musical presentation. I
supposed this was part of the Charlize (or even Tripath) signature
sound, one which some have described as holographic, hyperreal and so
on. In sum I was very pleased. A so-called keeper as some are wont to
say. The word "civilized" came to mind. Given the low noise floor,
level of detail and resolution, reasonable imaging and spatiality,
Charlize is rather well-suited to listening, critical listening even,
at low volumes. The amp certainly recommends itself for classical
music.
Charlize DIY.
A perusal of the numerous, but not innumerable URL's on Charlize
modifications, appears to suggest a rather discriminating choice of
components, such that initial tinkering might focus on, if not be
restricted to the input coupling caps: the 10uF Elna Cerafine
cylinders clad in sombre red. A comment by the designer himself in the
apparently superceded yahoo discussion group for diyparadise suggests
that 4.7uF BG N caps should be a reasonable replacement /
upgrade. Alternatively, one might follow conventional wisdom and get
rid of the electrolytics here altogether. I hadn't yet come across a
report of MKS caps replacing the Cerafines, so, since MKS is widely
available, cheap and small, for my test I opted to try Wima MKS02
3.3uF. The result: sound was somewhat smoother, perhaps more neutral;
there seemed to be less detail; overall there was a loss of presence
and the amp - in my view - certainly became less engaging. So the
Cerafines were rather good after all. The MKS caps were thus quickly
removed, the Cerafines returned.
Non-polar Black Gates might have been next, but then I opted for a
pair of Mundorf MCap Supremes, 2.2uF. They could be acquired more
quickly: here a case of local expediency over possible rationality,
understandable perhaps considering that other "boutique" caps have
been in transit for several weeks, remain in transit. The Mundorfs
arrived a few days later, were wired in. Result: the edginess is
gone; the detail remains. From a music reproduction point of view,
strings are now on par with piano - of significance imaginably
depending on one's listening preferences. No longer is the amp's sonic
signature idiosyncratic. As one might surmise, the Mundorfs have been
left in place. Such caps are big however, ca. 25x55mm, so if space is
an issue, BG's might be a fine choice too, given especially that the
stock Cerafines are very good.
Tthis is too long a post for too trivial remarks, but there
you have it, Charlize, my ramble...
An excellent report Danon!
I have not removed the input caps from Charlize but I did on the Autocostruire 202o amp and that did make an improvement. When I have replaced electrolytics in the signal path, I have not found much (if any) advantage in using polyesters so I now go for polypropylene (or better).
Using a linear PSU seems to remove that "edginess or sharpness to the overall musical presentation" but you lose some of the plus points at the same time. I was also told by one of the guys who bought a Skynet 8080 from me that it clearly out-performed the other three SMPS he had tried. (sorry for the blatant plug)
I have not removed the input caps from Charlize but I did on the Autocostruire 202o amp and that did make an improvement. When I have replaced electrolytics in the signal path, I have not found much (if any) advantage in using polyesters so I now go for polypropylene (or better).
Using a linear PSU seems to remove that "edginess or sharpness to the overall musical presentation" but you lose some of the plus points at the same time. I was also told by one of the guys who bought a Skynet 8080 from me that it clearly out-performed the other three SMPS he had tried. (sorry for the blatant plug)
Actually, Nuuk, in no small measure did your review of the Tripath
kits direct my attention to Charlize; ultimately, quite possibly to
that infamous state known on Audiogon as "Done for now" (here anyways,
as regards spkr amplification). So, my gratitude for the review!
Couple of further notes. I did wonder whether a 2-3uF cap would be
sufficient. The Tripath Wiki seems to suggest such a capacitance is
appropriate for the T-amp but is a bit ambiguous on whether this
extends say to all 2020 implementations, including for example
Charlize. If anyone is of the view that greater capacitance is better,
or appropriate for Charlize input coupling capacitance -- equations
etc welcome -- I would be interested to know. My listening suggests
that 2.2uF is fine, no detectable bass roll-off, but certainty of
course is better than belief. Also, since Charlize does, as yeo has
observed, have a nice bottom end, it would be a pity if perception of
her delights were merely partial.
Other possible issue. Phase inversion. If Charlize inverts phase, and
assuming what is upstream (e.g., a DAC or whatever) does not, this
might suggest that external speaker leads should be connected to
Charlize with reverse polarity; or that the amp should be wired up to
binding posts with reversed polarity accounting for inverted output.
Perhaps someone has investigated.
kits direct my attention to Charlize; ultimately, quite possibly to
that infamous state known on Audiogon as "Done for now" (here anyways,
as regards spkr amplification). So, my gratitude for the review!
Couple of further notes. I did wonder whether a 2-3uF cap would be
sufficient. The Tripath Wiki seems to suggest such a capacitance is
appropriate for the T-amp but is a bit ambiguous on whether this
extends say to all 2020 implementations, including for example
Charlize. If anyone is of the view that greater capacitance is better,
or appropriate for Charlize input coupling capacitance -- equations
etc welcome -- I would be interested to know. My listening suggests
that 2.2uF is fine, no detectable bass roll-off, but certainty of
course is better than belief. Also, since Charlize does, as yeo has
observed, have a nice bottom end, it would be a pity if perception of
her delights were merely partial.
Other possible issue. Phase inversion. If Charlize inverts phase, and
assuming what is upstream (e.g., a DAC or whatever) does not, this
might suggest that external speaker leads should be connected to
Charlize with reverse polarity; or that the amp should be wired up to
binding posts with reversed polarity accounting for inverted output.
Perhaps someone has investigated.
I noted that the input and feedback resistor values on Charlize and the Autocostruire 2020 amps were the same so I assume that the 4.7 uf polyprops that I am using with the latter would also work fine with Charlize.
With regards to absolute phase, I do connect the speaker terminals the 'wrong' way to 'correct' the inverted signal. I don't think that this is too important though unless you are using the amp in a multi-amped system!
With regards to absolute phase, I do connect the speaker terminals the 'wrong' way to 'correct' the inverted signal. I don't think that this is too important though unless you are using the amp in a multi-amped system!
Re phase inversion - thanks for the confirmation. I think I do recall someone observing inverted output by using a test CD (decware forum possibly). I do agree that most probably would not hear the difference. Nevertheless there doesn't seem to be any reason to inhabit an inverting universe if unnecessary. Time for snap soldering again..
danon said:Re phase inversion - thanks for the confirmation. I think I do recall someone observing inverted output by using a test CD (decware forum possibly). I do agree that most probably would not hear the difference. Nevertheless there doesn't seem to be any reason to inhabit an inverting universe if unnecessary. Time for snap soldering again..
danon, could you please inform what test CD was used to determine the right polarity? I got totally lost with the polarities, since either way I plug the speaker terminals, the sound remains the same (I went deaf, probably).
...and "inverting universe IS [indeed] unnecessary"
These last few posts about phase inverting have set me thinking. I have an Autocostruire 2020. I use a subwoofer. I find it works much better out of phase, whereas when I use it with the Sonic T it sounds really odd and I have to put it in phase. Is this because the 2020 should have the speaker wires reversed ?
irss, after my recollection the particular test cd was not mentioned. I just tried to ferret out the post (using google of course) but can't seem to find it. I should note that phase inversion was already discussed in the first couple dozen posts of this thread. An educated guess: possibly it was an old cd from stereophile.
So I dug a bit further and found the following article:
The Stereophile Test CD
Robert Harley & John Atkinson, February, 1990
URL: http://stereophile.com/reference/176/
Quote: "[...] include a track to allow listeners to check how sensitive they are to absolute-phase inversion; [...]"
I don't have any - test cd's (or Stereophiles for that matter) - so can't check myself.
So I dug a bit further and found the following article:
The Stereophile Test CD
Robert Harley & John Atkinson, February, 1990
URL: http://stereophile.com/reference/176/
Quote: "[...] include a track to allow listeners to check how sensitive they are to absolute-phase inversion; [...]"
I don't have any - test cd's (or Stereophiles for that matter) - so can't check myself.
Perception of Audio Phase
Tripathistas, any others interested in phase, might find following sensible
article interesting:
Perception of Audio Phase
http://home.att.net/~lemon.j.russell/audphase.htm
As a further note: I can imagine an enterprising engineer or
other offering a public domain phase test mp3 file. MP3 format
because less wasted bandwidth. Public domain because it could then be
hosted or mirrored by anyone. Perhaps some such already exists.
Tripathistas, any others interested in phase, might find following sensible
article interesting:
Perception of Audio Phase
http://home.att.net/~lemon.j.russell/audphase.htm
As a further note: I can imagine an enterprising engineer or
other offering a public domain phase test mp3 file. MP3 format
because less wasted bandwidth. Public domain because it could then be
hosted or mirrored by anyone. Perhaps some such already exists.
Wow, that Stereophile article is a blast from the past!
Imagine editing a CD with only 600 megs of HD space. The stone age!
Actually, I remember those days. Pro Tools in the late 80s. It was cool, but took a heck of a setup.
Anyhow, the Tripath chips are inverting, positive in = negative out. The opamp front end is inverting. For the bridged chips, no worries, just reverse the speaker leads. When I build my amps, I just do it inside the case.
It is easy to see the reverse polarity with an oscilloscope and a sawtooth wave. If the waveform leading edge is the ramp, the trailing edge will be the sharp drop-off. Once the polarity is switched, you will see the opposite. Hearing it is another matter.
As for the Russell Lemon phase article, you can generate his tone pairs on the computer using something like the NCH Tone Generator
You can even save the tones as wave files and burn them to CD. FWIW, a quick test of some of his tone pairs did not generate 1Hz beats on my computer. (I do not know the distortion specs or freq precision of the NCH generator.)
Imagine editing a CD with only 600 megs of HD space. The stone age!
Actually, I remember those days. Pro Tools in the late 80s. It was cool, but took a heck of a setup.
Anyhow, the Tripath chips are inverting, positive in = negative out. The opamp front end is inverting. For the bridged chips, no worries, just reverse the speaker leads. When I build my amps, I just do it inside the case.
It is easy to see the reverse polarity with an oscilloscope and a sawtooth wave. If the waveform leading edge is the ramp, the trailing edge will be the sharp drop-off. Once the polarity is switched, you will see the opposite. Hearing it is another matter.
As for the Russell Lemon phase article, you can generate his tone pairs on the computer using something like the NCH Tone Generator
You can even save the tones as wave files and burn them to CD. FWIW, a quick test of some of his tone pairs did not generate 1Hz beats on my computer. (I do not know the distortion specs or freq precision of the NCH generator.)