Hi Russ, avr300
I've tried removing the 15n caps too. Leo first tried this, and was astonished how much better it sounded. As was I. (I'm using just CD source material).
Several other people have heard the same. It really isn't subtle. But I nearly didn't try it on the basis that, even if the filtering did no good, it could, in theory, do no harm, so why not leave it in.
Well it seems theory is deficient. I've no idea what's going on here, but I like it!
I've tried removing the 15n caps too. Leo first tried this, and was astonished how much better it sounded. As was I. (I'm using just CD source material).
Several other people have heard the same. It really isn't subtle. But I nearly didn't try it on the basis that, even if the filtering did no good, it could, in theory, do no harm, so why not leave it in.
Well it seems theory is deficient. I've no idea what's going on here, but I like it!
Hi francolargo
It's to maintain optimum screening (I assume- ask Auricap). In a cylindrically wound cap, it minimises stray signal pickup if the outer foil is connected to the path with lowest impedance to ground. This marked by the black wire (Hovland caps were the same, only with a green wire). Usually, for a coupling cap, this means that the red wire should be connected to the output terminals, and the black to the circuit driving it
Paul
It's to maintain optimum screening (I assume- ask Auricap). In a cylindrically wound cap, it minimises stray signal pickup if the outer foil is connected to the path with lowest impedance to ground. This marked by the black wire (Hovland caps were the same, only with a green wire). Usually, for a coupling cap, this means that the red wire should be connected to the output terminals, and the black to the circuit driving it
Paul
For wound film caps there in some instances a marking, in your case the lead color, that designates the outer foil, so there are instances (in theory) where it may be better to have the outer foil connected one way around.
I suspect that in practice in makes little or no difference.
Regards
I suspect that in practice in makes little or no difference.
Regards
I am using Auricaps, which don't even remotely fit the board but have nice multistranded flexible leads. For a number of years I've been using them wherever transparency is needed. But just tonight I discovered something I didn't know about them (plus a review confirming small polarity effects):
* In signal path applications the black lead is "in" and the red lead is "out."
* In power supplies the black lead should be connected to ground and the red lead to the + or - voltage.
*In loudspeaker crossovers the red lead should be connected toward the + terminal on the drive unit.
Huh? Any comments on this by the "cognoscenti"? My caps are apparently in backwards because before reading this I would never have soldered a black lead into a location labeled "plus" and vice versa for red leads!I'll try changing them the next time I have everything apart...
What sonic difference?
Could you possibly explain the sonic difference heard without the 15 nF caps? I guess that these are part of an analog filter stage, to remove out of band artifacts-what are the ramifications of allowing these artifacts through?
Russ: what might be the R that applies to these caps for filtering, perhaps raising the corner frequency some might provide sonic benefits without eliminating this filter alltogether?
Could you possibly explain the sonic difference heard without the 15 nF caps? I guess that these are part of an analog filter stage, to remove out of band artifacts-what are the ramifications of allowing these artifacts through?
Russ: what might be the R that applies to these caps for filtering, perhaps raising the corner frequency some might provide sonic benefits without eliminating this filter alltogether?
Basically the filter is formed between output impedance of the I/V stage (which is dominated by the I/V resistor) and the cap. In the stock configuration (150R I/V) the corner is about 65khz. The filter is 1st order.
If you increase the I/V resistor you will get a lower corner. For example 357R would get about 30khz corner.
The presence and value of the cap are totally up to the user. Though I would never suggest omitting it.
Cheers!
Russ
If you increase the I/V resistor you will get a lower corner. For example 357R would get about 30khz corner.
The presence and value of the cap are totally up to the user. Though I would never suggest omitting it.
Cheers!
Russ
For what it's worth, nobody I have asked has ever made such comments.
I would take anything like that with a grain of salt. I am not saying people aren't being honest, just that my experience is definitely not the same.
Hi Paul
Krell KAV400ix via NAC A5 into PMC OB1. I know the combination is intrinsically bright but I heard the same via KT88 triode valve amp in my man cave system. Yes, it tames the glassiness but sounds as dynamic/open as without the caps.
I have one of the Legato 2.0.0 boards with Takman metal film IV resistors, Charcroft for 4.7k and 1k resistors, Blackgate N 100uF bypassed with Vishay/Roederstein MKP1837 100nF as DC blocking caps. Panasonic FM for powersupply decoupling caps. Balanced to unbalanced conversion with metal can LME49720HA.
The powersupply uses Hexfreds to Nichicon Superthrough 4700uF and 0.47R CRC to Placids. When using the Krell the connection is balanced and unbalanced to valve amp.
As you can see I've pushed this board a bit, but it responds brilliantly to component upgrades. I use PC based audio analysis software via a EMU 0404 soundcard to check that I don't mess up:-0.
Cornelius
Krell KAV400ix via NAC A5 into PMC OB1
. I know the combination is intrinsically bright but I heard the same via KT88 triode valve amp in my man cave system. Yes, it tames the glassiness but sounds as dynamic/open as without the caps.I have one of the Legato 2.0.0 boards with Takman metal film IV resistors, Charcroft for 4.7k and 1k resistors, Blackgate N 100uF bypassed with Vishay/Roederstein MKP1837 100nF as DC blocking caps. Panasonic FM for powersupply decoupling caps. Balanced to unbalanced conversion with metal can LME49720HA.
The powersupply uses Hexfreds to Nichicon Superthrough 4700uF and 0.47R CRC to Placids. When using the Krell the connection is balanced and unbalanced to valve amp.
As you can see I've pushed this board a bit, but it responds brilliantly to component upgrades. I use PC based audio analysis software via a EMU 0404 soundcard to check that I don't mess up:-0.
Cornelius
Hi Cornelius
My I/V resistors are also 1W Takman metal films, and the op amp a metal can LME49720HA- snap! (I've also done some other mods and use Paul Hynes series regs- but not TP approved)
When using the Bal/SE stage, try bypassing the output caps- you should find the dc offset between the Legato's output pairs is small enough (a few mV) for this
I use valve amps with only local series feedback- very different to your solid state system, and even your KT88 amp will have some global feedback. Which may explain our somwhat different experiences. My treble is the sweetest it's ever been!
Paul
My I/V resistors are also 1W Takman metal films, and the op amp a metal can LME49720HA- snap! (I've also done some other mods and use Paul Hynes series regs- but not TP approved)
When using the Bal/SE stage, try bypassing the output caps- you should find the dc offset between the Legato's output pairs is small enough (a few mV) for this
I use valve amps with only local series feedback- very different to your solid state system, and even your KT88 amp will have some global feedback. Which may explain our somwhat different experiences. My treble is the sweetest it's ever been!
Paul
I can't explain it, but have a look a some FFT's
Picasa Web Albums - Cornelius van Wyk - Legato Filtering
Picasa Web Albums - Cornelius van Wyk - Legato Filtering
That...
Does not seem right at all? Can anyone explain what we are looking at here? Are we looking at the output of the Legato alone, or the output of another component being fed the Legato's signal?
Why would a filter effective about 500 kHz have any effect here?
I am considering going 10nF REL polystyrene here, with R1-4 at 180R.
I can't explain it, but have a look a some FFT's
Picasa Web Albums - Cornelius van Wyk - Legato Filtering
Does not seem right at all? Can anyone explain what we are looking at here? Are we looking at the output of the Legato alone, or the output of another component being fed the Legato's signal?
Why would a filter effective about 500 kHz have any effect here?
I am considering going 10nF REL polystyrene here, with R1-4 at 180R.
I'm pretty sure my method of measuring is ok but electronics is just a hobby, so I might be completely off base. PC - 27" iMac Quad core 2.66GHz, 8Gb ram running Win7 in bootcamp with Virtins Multi-Instrument 3.2. Emu 0404 usb external sound card. Signal to DUT via Emu 0404 and M2Tech Hiface using ASIO and MME drivers with similar results. Sampling at 24bit and 96kHz using Flattop and Blackman Harris windowing also with repeatable results. The test signal(SPDIF) was at 24/96 but I get similar results with 16/44.1.
Using Audiotester software in Windows and Mac the Scope in OSX I also get repeatable results with numerous devices including Cambridge Audio DacMagic(latest), DIY 1541aS2 DAC, Musical Fidelity TriVista DAC and Devilsound V2 DAC, so I'm fairly confident about the results even though they baffle me. There is quite a few reports online about the accuracy of the Emu 0404 as measurement device. It's quite tricky to get Windows to not interfere with sample rate conversion and you can see obvious jitter sidebands when you've got it wrong
What's been encouraging is that when ever I've heard something that bothers me after a mod/"upgrade", it was obvious when I measure that something is wrong. Also I noticed that I like some low order distortion and seem to be sensitive to phase and pre-ringing with impulse response testing. I can only postulate that there is some phase/timing changes with the filtering that's minimised with 1.8nf whilst higher order distortion is decreased enough not to bother me any more as with no filtering.
This is a interesting hobby, as it is pretty obvious that we all have a unique taste in sound and DIY allows us to tune equipment to our liking
Cornelius
Using Audiotester software in Windows and Mac the Scope in OSX I also get repeatable results with numerous devices including Cambridge Audio DacMagic(latest), DIY 1541aS2 DAC, Musical Fidelity TriVista DAC and Devilsound V2 DAC, so I'm fairly confident about the results even though they baffle me. There is quite a few reports online about the accuracy of the Emu 0404 as measurement device. It's quite tricky to get Windows to not interfere with sample rate conversion and you can see obvious jitter sidebands when you've got it wrong
What's been encouraging is that when ever I've heard something that bothers me after a mod/"upgrade", it was obvious when I measure that something is wrong. Also I noticed that I like some low order distortion and seem to be sensitive to phase and pre-ringing with impulse response testing. I can only postulate that there is some phase/timing changes with the filtering that's minimised with 1.8nf whilst higher order distortion is decreased enough not to bother me any more as with no filtering.
This is a interesting hobby, as it is pretty obvious that we all have a unique taste in sound and DIY allows us to tune equipment to our liking
Cornelius
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.