carlosfm
Thanks for the link. It answered two of the questions I asked above. I agree that a good preamp is more convenient than tuning up a buffer for those of us that have uncooperative living spaces but I still believe the best bang for the buck comes from an integrated gainclone with a buffer and phono stage in the same box so the selection list for the amp section is listed:
the inverted "T" Clone
the "X" Clone
the NIGCWIPC or
one of the discreet buffered gainclones (Nuuk, help us decide here)
The X Clone seems the darkhorse here simply because of the questionable benefits of this design without balanced sources. The inverted T has been discredited by some but I think that with a little more tweaking and a buffer this might surprise us all. The NIGCWIPC has been tested by at least one member and he has found it robust and sounding quite well.
Thanks for the link. It answered two of the questions I asked above. I agree that a good preamp is more convenient than tuning up a buffer for those of us that have uncooperative living spaces but I still believe the best bang for the buck comes from an integrated gainclone with a buffer and phono stage in the same box so the selection list for the amp section is listed:
the inverted "T" Clone
the "X" Clone
the NIGCWIPC or
one of the discreet buffered gainclones (Nuuk, help us decide here)
The X Clone seems the darkhorse here simply because of the questionable benefits of this design without balanced sources. The inverted T has been discredited by some but I think that with a little more tweaking and a buffer this might surprise us all. The NIGCWIPC has been tested by at least one member and he has found it robust and sounding quite well.
Konnichiwa,
Hmmm, my OWN take would probably be an UNBUFFERED inverting gainclone, which uses the "lead/lag compensation" trick to fix the stability issue (in other words 220pF in series with 10K between inverting & noniverting input).
The 100K...220K linear track potentiometer (PEC Carbon?, alternatively generic tandem carbon track) together with the 10K input impedance of the IGC manage fine to give a proper "log" volume control.
A 330k/10K feedback circuit makes sure with an 18V transformer that around 0.5V will clip the amplifier if the volume is all the way up, way enough for any modern source, output probablky in the 25W ballpark.
Then add a "Analogue Addicts" Phonostage (also often called "El Cheapo") internally which has been covered elsewhere here many times and in detail and use up a pair of OPA637 for that if you can, if MC stepup is needed the single 2SK170 (et al) per channel is fine.
The resultant "integrated" amp is minimal, will sound pretty good, has an excellent MM (MC option) phono which equally sounds pretty good and will probably do for a lot of people.
I personally would probably limit any Solid State work to that kind of thing (it is worth doing more when doing commercial gear) but the result will take some beating, on a shoestring budget too.
Sayonara
yldouright said:
Hmmm, my OWN take would probably be an UNBUFFERED inverting gainclone, which uses the "lead/lag compensation" trick to fix the stability issue (in other words 220pF in series with 10K between inverting & noniverting input).
The 100K...220K linear track potentiometer (PEC Carbon?, alternatively generic tandem carbon track) together with the 10K input impedance of the IGC manage fine to give a proper "log" volume control.
A 330k/10K feedback circuit makes sure with an 18V transformer that around 0.5V will clip the amplifier if the volume is all the way up, way enough for any modern source, output probablky in the 25W ballpark.
Then add a "Analogue Addicts" Phonostage (also often called "El Cheapo") internally which has been covered elsewhere here many times and in detail and use up a pair of OPA637 for that if you can, if MC stepup is needed the single 2SK170 (et al) per channel is fine.
The resultant "integrated" amp is minimal, will sound pretty good, has an excellent MM (MC option) phono which equally sounds pretty good and will probably do for a lot of people.
I personally would probably limit any Solid State work to that kind of thing (it is worth doing more when doing commercial gear) but the result will take some beating, on a shoestring budget too.
Sayonara
We can always count on herr Loesch to shed some light on a darkened thread and make alight the darkened fiber of our thoughts. Welcome Kuei Yang Wang to my humble thread. The phono stage would indeed be and El Cheapo type circuit with some form of AD video op-amp. CarlosFM has been very happy with with AD815 pre so there is corroboration of your now three year old suggestions to use these AD video chips. Like you, I generally like less parts but so many members have agreed that a buffer is mandatory to get the most out the LM chips that I have been convinced the additional complexity is warranted. Two points I would like to make:
The rail voltage decided upon should really be a function of the builders planned speaker system. In my own case, I have some 4 ohm speaker drivers so I would use 12V rails if I were to build the amp section with LM3875s and 16V rails with the LM3886s.
Using a lower rail voltage necessitates a smaller multiplier so I thought a buffer would help add some gain. Obviously, I'm not the authority here but that is my opinion based on what I have read so far.
Have you had a chance to look at Mauro's two circuits?
The rail voltage decided upon should really be a function of the builders planned speaker system. In my own case, I have some 4 ohm speaker drivers so I would use 12V rails if I were to build the amp section with LM3875s and 16V rails with the LM3886s.
Using a lower rail voltage necessitates a smaller multiplier so I thought a buffer would help add some gain. Obviously, I'm not the authority here but that is my opinion based on what I have read so far.
Have you had a chance to look at Mauro's two circuits?
Konnichiwa,
Video Op_Amp's are not very suited for the job, you will find them rather noisy.
The OPA637 is the best compromise between noise, speed and sound for MM Pickups.
For MC Pickups, you may consider a AD811 in inverting configuration on the input for that, Bipolar CFB Video op-amp's are good in very low impedance circuit and some have low noise there, but for MM Pickups and MC with stepup, low noise Fet input Op-Amp's tend to be quite good for that kind of job and sound great there.
Buffers are in fashion now, BUT practically non of them have ever implemented the ACTUAL configuration I recommended. By now we have buffered, T-Network IGC's with regulated supplies and all that jazz. No doubt all these items have merit and are interesting to try and all, but there is also merrit in doing things simple.
Well, yes, this depends somewhat, but if the "soft" PSU derived from the original Gaincard is kept, then much higher rail voltages are actually needed.
I am not sure I follow, what are you refering to?
Yes, don't much care for that way of doing things myself, otherwise quite smart circuitry, but the multiple feedback loops can get out of hand and can result in poorer transient behaviour. If tuned right this may very well be better than the chip on its own, but it can equally be worse very easily... One would have to build and test the thing.
Sayonara
yldouright said:
Video Op_Amp's are not very suited for the job, you will find them rather noisy.
The OPA637 is the best compromise between noise, speed and sound for MM Pickups.
For MC Pickups, you may consider a AD811 in inverting configuration on the input for that, Bipolar CFB Video op-amp's are good in very low impedance circuit and some have low noise there, but for MM Pickups and MC with stepup, low noise Fet input Op-Amp's tend to be quite good for that kind of job and sound great there.
yldouright said:
Buffers are in fashion now, BUT practically non of them have ever implemented the ACTUAL configuration I recommended. By now we have buffered, T-Network IGC's with regulated supplies and all that jazz. No doubt all these items have merit and are interesting to try and all, but there is also merrit in doing things simple.
yldouright said:
Well, yes, this depends somewhat, but if the "soft" PSU derived from the original Gaincard is kept, then much higher rail voltages are actually needed.
yldouright said:
I am not sure I follow, what are you refering to?
yldouright said:
Yes, don't much care for that way of doing things myself, otherwise quite smart circuitry, but the multiple feedback loops can get out of hand and can result in poorer transient behaviour. If tuned right this may very well be better than the chip on its own, but it can equally be worse very easily... One would have to build and test the thing.
Sayonara
Sorry to be late for the party, but somehow this thread escaped my notice. Damn job keeps interfering with my hobbies.
To answer one question, after we got the last of the stability problems worked out, the GCSS will work nicely with a single ended input (just ground the negative input to maintain the gain). Compared to the more standard GC variants, driven single ended it compares quite well, but keep in mind I never did build a totally tricked out fully tweaked basic or GCSS version. Single ended it didn't really present a significant advantage over the basic version, at least to my ears. Balanced input was another story, where it kept the clarity but added more body to the presentation. The one advantage of the GCSS, independant of input configuration, is that the sound of the amp seems less dependant on the power supply configuration, probably primarily due to the cancellation effect of the balanced topology. To the GCSS the difference between capacitance only, regulated, capacitance snubbered, and regulated snubbered was a lot less noticeable, whereas with the basic GC I could hear distinctive differences. I keep hoping that Peter Daniel or Carlosfm might take a wander into GCSS land so I could find out their opinions, but no luck so far (hint, hint).
My opinion, for what it's worth. If your using single-ended input, your efforts are better spent on getting the power supply right than adding the complexity of the SuSy. I'd also vote for the 3886 over the 3875 now that I've had a chance to work with both.
I'll try to pay attention here from now on.
Cheer, Terry
To answer one question, after we got the last of the stability problems worked out, the GCSS will work nicely with a single ended input (just ground the negative input to maintain the gain). Compared to the more standard GC variants, driven single ended it compares quite well, but keep in mind I never did build a totally tricked out fully tweaked basic or GCSS version. Single ended it didn't really present a significant advantage over the basic version, at least to my ears. Balanced input was another story, where it kept the clarity but added more body to the presentation. The one advantage of the GCSS, independant of input configuration, is that the sound of the amp seems less dependant on the power supply configuration, probably primarily due to the cancellation effect of the balanced topology. To the GCSS the difference between capacitance only, regulated, capacitance snubbered, and regulated snubbered was a lot less noticeable, whereas with the basic GC I could hear distinctive differences. I keep hoping that Peter Daniel or Carlosfm might take a wander into GCSS land so I could find out their opinions, but no luck so far (hint, hint).
My opinion, for what it's worth. If your using single-ended input, your efforts are better spent on getting the power supply right than adding the complexity of the SuSy. I'd also vote for the 3886 over the 3875 now that I've had a chance to work with both.
I'll try to pay attention here from now on.
Cheer, Terry
carlosfm
What is reason you wouldn't use a current feedback op-amp on a phono stage?
Kuei Yang Wang
I thought I remembered reading that the video op-amps had the best noise to gain ratio and was very well suited as a phono stage gain device. Is there a way to use one of these video op-amps for both the MM and MC? I thought we might switch in a MM circuit that would throw away the gain without leaving a greater noise floor. There have been reports of some people having success with the T-Clones by throwing away the gain and thereby also reducing noise. I'm still researching exactly how this was done but if you know what I'm talking about and can elaborate I would be thankful. I mentiion the T-Clones because they were also reported to be noisy by some members but some members have successfully implemented the "T" layout presumably by throwing away gain. This technique is also mentioned in the X-Clone thread and endorsed by Nelson Pass as a way to get a better sounding gainclone. Speak of the devil
Terry, you must have sensed that I was thinking of you when I wrote that last comment.
What is reason you wouldn't use a current feedback op-amp on a phono stage?
Kuei Yang Wang
I thought I remembered reading that the video op-amps had the best noise to gain ratio and was very well suited as a phono stage gain device. Is there a way to use one of these video op-amps for both the MM and MC? I thought we might switch in a MM circuit that would throw away the gain without leaving a greater noise floor. There have been reports of some people having success with the T-Clones by throwing away the gain and thereby also reducing noise. I'm still researching exactly how this was done but if you know what I'm talking about and can elaborate I would be thankful. I mentiion the T-Clones because they were also reported to be noisy by some members but some members have successfully implemented the "T" layout presumably by throwing away gain. This technique is also mentioned in the X-Clone thread and endorsed by Nelson Pass as a way to get a better sounding gainclone. Speak of the devil
Terry, you must have sensed that I was thinking of you when I wrote that last comment.
Konnichiwa,
Nope, just not true. Check the datasheets.
OPA627/637 has around 5nV|/Hz and practically no current noise. The MM Cartrige is usually an inductor around 0.5-1H so across a lot of the audio range the noise of the Op-Amp is comparable to that generated by Cartridge & load resistor. Add any bipolar input op-amp and you add a lot of current noise, and your low noise goes to hell.
The AD815 due to current noise has a net Ein of around 3uV+ for 100Hz - 20KHz while the OPA637 manages < 0.8uV under the same conditions, or respectively -64db and -76db below 5mV respectively.
Not easily, the AD815 will be marginal and it appears to be indeed one of the best....
Sayonara
yldouright said:
Nope, just not true. Check the datasheets.
OPA627/637 has around 5nV|/Hz and practically no current noise. The MM Cartrige is usually an inductor around 0.5-1H so across a lot of the audio range the noise of the Op-Amp is comparable to that generated by Cartridge & load resistor. Add any bipolar input op-amp and you add a lot of current noise, and your low noise goes to hell.
The AD815 due to current noise has a net Ein of around 3uV+ for 100Hz - 20KHz while the OPA637 manages < 0.8uV under the same conditions, or respectively -64db and -76db below 5mV respectively.
yldouright said:
Not easily, the AD815 will be marginal and it appears to be indeed one of the best....
Sayonara
Kuei Yang Wang
The OPA627/637 is good for a gain of 10x according to what I've read before its noise starts increasing geometrically, the video op-amps we are discussing hold their noise value up to a gain of 60x and higher so if you take that into consideration, how am I incorrect when I say that they have superior gain to noise ratios?
The OPA627/637 is good for a gain of 10x according to what I've read before its noise starts increasing geometrically, the video op-amps we are discussing hold their noise value up to a gain of 60x and higher so if you take that into consideration, how am I incorrect when I say that they have superior gain to noise ratios?
yldouright said:
With current-feedback op-amps you should use feedback resistors values recommended by the manufacturer, or close to that.
Also, you can't use caps on the feedback loop.
To help even more, they have have high(er) input dc-offset, which you would really need to protect with input coupling caps if you don't want to destroy your cartridge.
As Kuei says, the OPA637 is very good for the job, has low noise, and sounds amazing biased to class-A at around 10ma.
The 'El Cheapo' is a very good phono stage.
I made one a couple of years ago (MM), with OPA637 biased to class-A.
I compared that to a Linn Linto (connected to a full-spec LP12 / Linn Akiva) and the 'El Cheapo' was better.
The 'El Cheapo' doesn't sound cheapo.
metalman said:
I'm presently using a bridged version of LM4780 with (only) a balanced input. After trying that, I don't think I will go back to single ended LM3875 (unless I try it in bridged version).
As you noted, a bridged version seems to give more body to the presentation without loosing any clarity. Also spaciousness and dynamics increase.
Just today, I replaced ML380S preamp with S&B102 line stage in a completely balanced configuration and I really like what I hear now. There is a certain coherence and ease of presentation I didn't experience in a long while. I will be soon trying a new silver version.
But this was only possible with a new DAC I got yesterday (that has low impedance balance outputs).
I tried similar configuration with TDA1543 passive stage and S&B102 used as single ended to balanced converter didn't work well at all. Using it in fully balanced setup is another story.
As to the SuSy, this would require another active stage in front of the amp, and I'm not sure if the trade off is worth that. But considering the simplicity of such circuit, I might try it one day
IMO, the future of GCloning is in fully balanced topologies.
Peter Daniel said:
Passive stage... sure.
The source is the most important component on any system.
It makes everything after it sound good or bad, it rules.
Single-ended or balanced, the difference is not night and day, if well done.
You can't compare amps when you also change the source component.
metalman said:
Well, somebody suggested that I should try different things than my "beloved" TDA1543 DAC, so I purchased recently some other DACs for comparison purposes. I have now Benchmark DAC-1. Having seen all those good reviews, I couldn't resist not to try it. Well, I tried it and tomorrow it goes back on Audiogon.
I also found a very good deal on ML360S and bought it. I didn't really like it that much, when it replaced TDA1543. Although sort of refined and well behaved, it was lacking immediacy and energy the smaller, (beloved) DAC shows. It is almost like those quotes, I found on AA, perfectly apply here:
"Many reviewers harp on such descriptions as 'smooth, analog-like', etc. I would rather make my comparisons to the effect of live music....which is neither necessarily smooth, or analog-like! But it IS emotional, driving, very present, and full-range (airy highs AND taught, very deep bass). One is simply moved by the music, the intent of the artist, and dazzled by the notes, the beat, the emotion of the singer, the life of the recording. CDs played through (such) components will grab your atention. Take John Hammonds' latest cd, Ready For Love. I happen to know John personally, and when I say that it is erie to hear his music on my system, I mean it. His voice is so real and alive, all that is missing is the cigarette smoke. This cd sounds like John and the band were having fun. That is the kind of energy that comes across; it's more than just the notes.
... and it has the best bass I have heard (by that I mean tightness and impact, not just 'boom') and it delivers the energy of music, the emotion, the reason why an artist has bothered to put the music to disc in the first place. This is not just a matter of good sound, which I realize now is what I used to be into: nice, warm, lovely sound. No, this is direct music like you will get live. Not colored, syrupy 'sound'. It is not for the faint of heart. But it is for those of us who want to get into the essence of what is on the disc."
But, I thought that I can somehow improve things and since ML has exactly the same output stage as the preamp (6 ohm output impedance), I decided to remove preamp completely and use passive, S&B102 based, line stage. Well, it was a right move and the sound is pretty good now, different in character from what I was getting with NOS DAC and ML380S preamp, but at least comparable
PS: To be completely clear, the TDA1543 DAC I'm using presently is not exactly the same version that I offer as a kit. It has some parts upgraded: 4 x BG N 1000/50 caps are used in main filtering stage (replacing BG NX 25/50), Caddock TF020 replace Rikens and V-Cap TFTF replace BG N in coupling. Those parts alone add about $600 to the price tag though
Attachments
Peter Daniel said:
You can't change two components at once and compare.
It's not conclusive.
Make a single-ended TDA1541A NOS dac with Sin(x)/x compensation and active stage and compare again with that balanced output dac.
Also, read below.
Peter Daniel said:
Right, agreed.
You will never hear this out of a NOS dac with passive output.
carlosfm said:
I was using balanced amp in all cases, as the preamp has balanced output. And I was using single ended outputs from both DACs, What is your point then as I still don'tt get it.
And this is exactly what you are getting from NOS DAC with passive output:
"...to the effect of live music....which is neither necessarily smooth, or analog-like! But it IS emotional, driving, very present, and full-range (airy highs AND taught, very deep bass). One is simply moved by the music, the intent of the artist, and dazzled by the notes, the beat, the emotion of the singer, the life of the recording. CDs played through (such) components will grab your atention."
If you claim it's not the case, you simply didn't make it right. But I'm not really surprised, observing your conclusions in other areas.
Peter Daniel said:
Well, some deaf people out there?
A NOS dac without Sin(x)/x compensation has a marked high frequency roll-off.
Passive stage has high output impedance, of course it doesn't work with a passive pre.
I don't know how some can't clearly detect this on a listening test, because it's very clear.
Maybe my standards are higher than normal?
Oh well...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.