Hi borisov57
Yes I did, on the IV / differential amplifier modules in the DI8M. Sound quality improved (smoother / higher resolution). It's a simple and cheap mod that's worth trying, it usually makes sound smoother, and enhances resolution.
I added a switch to toggle NGM on / off, so I could hear the effect it had on the original circuit.
Yes I did, on the IV / differential amplifier modules in the DI8M. Sound quality improved (smoother / higher resolution). It's a simple and cheap mod that's worth trying, it usually makes sound smoother, and enhances resolution.
I added a switch to toggle NGM on / off, so I could hear the effect it had on the original circuit.
Hi lumanauw,
Yes in fact I did, some years ago. I used OPA2604 in an 8th order Butterworth filter for a NOS DAC. But like noted before, using an OP-amp as unity gain buffer is not a great idea for High-End audio applications.
It would be better to use simple discrete JFET buffer amplifiers instead.
Sound quality degrades as more Op-amps are placed in the signal path....
Yes in fact I did, some years ago. I used OPA2604 in an 8th order Butterworth filter for a NOS DAC. But like noted before, using an OP-amp as unity gain buffer is not a great idea for High-End audio applications.
It would be better to use simple discrete JFET buffer amplifiers instead.
Sound quality degrades as more Op-amps are placed in the signal path....
Spartacus,
Have you posted your results in this forum? Any links to some good discussions / threads would be beneficial.
Keith,
Thank you very much for your very kind offer which I could not decline but have to. I would love to test your class A bias but I am still worried about heat dissipation. These chips are designed to be used in class B (or AB) operation, not class A. Of course I can bias it to class A provided that the heat can be dissipated. If I had the OPA627 in the DIP package, I would test your method. But I have only two tiny OPA627 in SOIC mounted back to back on a browndog, so heat will be an issue if used in class A. Have you got a class A opa627 buffer already? If so, you could send it to me. The postage won't be more than sending me the 2N5484. I could then compare it to my class B opa627 buffer in my set up and publish the results here. I will send the opa627 buffer back to you in N.Y after the comparison.
Conrad,
A few months ago, I brought my multimeter to my friends' place. He made a living by repairing electronics. My MM, when the probes were shorted, showed 0.5R. He sprayed something on them. There was no rubbing, scratching, wiping or any other sorts of cleaning. The MM magically showed 0R then. Both of us laughed and were pretty amazed. It was the first time he used that particular sort of spray. I still don't know what it is. I will ask him next time as I want to buy one.
ecdesign,
Thanks for the input. Can you describe more about your discrete JFET opamp?
Have you posted your results in this forum? Any links to some good discussions / threads would be beneficial.
Keith,
Thank you very much for your very kind offer which I could not decline but have to. I would love to test your class A bias but I am still worried about heat dissipation. These chips are designed to be used in class B (or AB) operation, not class A. Of course I can bias it to class A provided that the heat can be dissipated. If I had the OPA627 in the DIP package, I would test your method. But I have only two tiny OPA627 in SOIC mounted back to back on a browndog, so heat will be an issue if used in class A. Have you got a class A opa627 buffer already? If so, you could send it to me. The postage won't be more than sending me the 2N5484. I could then compare it to my class B opa627 buffer in my set up and publish the results here. I will send the opa627 buffer back to you in N.Y after the comparison.
Conrad,
A few months ago, I brought my multimeter to my friends' place. He made a living by repairing electronics. My MM, when the probes were shorted, showed 0.5R. He sprayed something on them. There was no rubbing, scratching, wiping or any other sorts of cleaning. The MM magically showed 0R then. Both of us laughed and were pretty amazed. It was the first time he used that particular sort of spray. I still don't know what it is. I will ask him next time as I want to buy one.
ecdesign,
Thanks for the input. Can you describe more about your discrete JFET opamp?
Hi homemodde,
I would try an AD823 instead of AD826 & AD829 at the I/V stage.
Suppose you use a 16-bit DAC with FS output 2mA. The current of the LSB would be 2mA / 65536 = 30.5nA.
The input bias current of AD826 and AD829 is 3uA. You can figure out how many bits they have eaten for biasing. Furthermore, this current flows in one direction only, into the OpAmp in this case. You lost some bits in the positive half cycle and not on the negative side. Harmonic distortion would increase as a result.
While the AD823 has input bias current of 3pA(typ.) and 25pA(max).
The advantage is obvious.
Besides, the distortion level of AD823 is only 4 to 6dB higher than that of opa627 at 20KHz. The AD823 is a dual and much cheaper. You may give it a try.
xtaltech
I would try an AD823 instead of AD826 & AD829 at the I/V stage.
Suppose you use a 16-bit DAC with FS output 2mA. The current of the LSB would be 2mA / 65536 = 30.5nA.
The input bias current of AD826 and AD829 is 3uA. You can figure out how many bits they have eaten for biasing. Furthermore, this current flows in one direction only, into the OpAmp in this case. You lost some bits in the positive half cycle and not on the negative side. Harmonic distortion would increase as a result.
While the AD823 has input bias current of 3pA(typ.) and 25pA(max).
The advantage is obvious.
Besides, the distortion level of AD823 is only 4 to 6dB higher than that of opa627 at 20KHz. The AD823 is a dual and much cheaper. You may give it a try.
xtaltech
Last night I wanted to draw a conclusion which opamp, i.e. opa627 or lm4562 to be used in my active XO project. They are needed for the buffer, (shelving) LP/HP, notch, gain and shellen key stages. For the test, I still used the same unity gain buffer.
Instead of playing familiar music, this time I played music that I had not listened to for a long time and the music I was less familiar with. I used the opamps in turn twice, mainly listening to the music from the next room / lounge, but sometimes walked to the music room. There are two large openings / doors from the music room to the lounge so the sound was clearly heard. I thought this allowed me to have a better perception of the overall musical presentation, even though the room was included into the picture.
This time I was quite sure that I liked the opa627 better than the lm4562. I felt that the sound from the opa627 was more refined and controlled. I started to think that the "live sound" of the lm4562 may come from some type of distortions, despite the doses being very small. This could happen. For example, odd order harmonics in certain frequency range seem to add some "presence" and "realism" to the music. In Photoshop, adding certain type of "noise" can sharpen the images. The more naturalness and sweetness of human voices from the opa627 may suggest that it is a more accurate performer. While the lm4562 was exciting to listen to, when listening to the opa627 alone, and I did that for days, told me that the chip was audibly faultless. The chip really appears to have no sound of its own, only producing whatever music fed into it.
Of course, again, this result was highly subjective. I expect that if the opamps are used differently I may have different results.
But for now, for me, opa627 is the winner.
Given that LM4562 is such a good chip at a much lower price than the opa627, I wonder if this will eventually force BB to drop its price, or perhaps to produce another "low cost" chip that replaces the opa2134 and compete with the lm4562. Technologies are advancing. The chips are now getting better and better.
Instead of playing familiar music, this time I played music that I had not listened to for a long time and the music I was less familiar with. I used the opamps in turn twice, mainly listening to the music from the next room / lounge, but sometimes walked to the music room. There are two large openings / doors from the music room to the lounge so the sound was clearly heard. I thought this allowed me to have a better perception of the overall musical presentation, even though the room was included into the picture.
This time I was quite sure that I liked the opa627 better than the lm4562. I felt that the sound from the opa627 was more refined and controlled. I started to think that the "live sound" of the lm4562 may come from some type of distortions, despite the doses being very small. This could happen. For example, odd order harmonics in certain frequency range seem to add some "presence" and "realism" to the music. In Photoshop, adding certain type of "noise" can sharpen the images. The more naturalness and sweetness of human voices from the opa627 may suggest that it is a more accurate performer. While the lm4562 was exciting to listen to, when listening to the opa627 alone, and I did that for days, told me that the chip was audibly faultless. The chip really appears to have no sound of its own, only producing whatever music fed into it.
Of course, again, this result was highly subjective. I expect that if the opamps are used differently I may have different results.
But for now, for me, opa627 is the winner.
Given that LM4562 is such a good chip at a much lower price than the opa627, I wonder if this will eventually force BB to drop its price, or perhaps to produce another "low cost" chip that replaces the opa2134 and compete with the lm4562. Technologies are advancing. The chips are now getting better and better.
HiFiNutNut,
The input stage is a dual N-JFET (PMBFJ620, SOT363, low noise / high-gain). It's fed by a cascode current source (2 x N-JFET)
The signal is then fed to a cascode amplifier (2 x P-JFET), the cascode amplifier is fed by a N-JFET current source.
The signal is then fed to an output bufer (J-FET / N-FET pair)
The amplifier has 7 JFETs, 1 Dual JFET, and 8 resistors.
The circuit requires some tuning, due to JFET tolerances. Low DC offset voltage can be achieved by using a DC servo circuit.
It should be possible to place this JFET OP-amp on a very small PCB, as all parts are available in SMD.
THD @ 1 KHz is below 0.002%.
The input stage is a dual N-JFET (PMBFJ620, SOT363, low noise / high-gain). It's fed by a cascode current source (2 x N-JFET)
The signal is then fed to a cascode amplifier (2 x P-JFET), the cascode amplifier is fed by a N-JFET current source.
The signal is then fed to an output bufer (J-FET / N-FET pair)
The amplifier has 7 JFETs, 1 Dual JFET, and 8 resistors.
The circuit requires some tuning, due to JFET tolerances. Low DC offset voltage can be achieved by using a DC servo circuit.
It should be possible to place this JFET OP-amp on a very small PCB, as all parts are available in SMD.
THD @ 1 KHz is below 0.002%.
contact enhancer?
I agree about the superior sound of the OPA627, especially when class-A biased. I was surprised at how much better it was than my previous favorite, the OPA2132, especially in the all-important midrange..
HiFiNutNut wrote:
I do remember a hearing about a demo where someone wiped the back of a flashlight bulb with some "Cramolin" (like Progold), and the previously dim flashlight shone brighter... Interesting!
-chas
I agree about the superior sound of the OPA627, especially when class-A biased. I was surprised at how much better it was than my previous favorite, the OPA2132, especially in the all-important midrange..
HiFiNutNut wrote:
Q: Anyone else on this forum ever experienced anything like this using contact cleaners or enhancers, such as Caig products?
I do remember a hearing about a demo where someone wiped the back of a flashlight bulb with some "Cramolin" (like Progold), and the previously dim flashlight shone brighter... Interesting!
-chas
Interesting reading.
One thing that strikes me.. since you put all this time and effort into deciding which IC sounds best, why not using them in a bypass test to see what they actually do to the signal?
A/B:ing two components against eachother is interesting but IMO it's even more interesting to know HOW the DUT affects the signal as compared to no DUT at all.
Maybe it turn out that some of these alternatives (627, 4562, discrete j-fet..) colors the signal and some don't..?
Just my thoughts.
For the record I'm a believer of transparent electronics and in the end if needed color the signal only when poorly produced music call for it.
/Peter
One thing that strikes me.. since you put all this time and effort into deciding which IC sounds best, why not using them in a bypass test to see what they actually do to the signal?
A/B:ing two components against eachother is interesting but IMO it's even more interesting to know HOW the DUT affects the signal as compared to no DUT at all.
Maybe it turn out that some of these alternatives (627, 4562, discrete j-fet..) colors the signal and some don't..?
Just my thoughts.
For the record I'm a believer of transparent electronics and in the end if needed color the signal only when poorly produced music call for it.
/Peter
Alas, though the 627 is a good audio part, it's claim to fame is for parameters not necessarily key to audio performance. Those parameters are difficult to achieve and screen for, so the cost will probably remain high. Which keeps demand low. Which keeps cost high. Somebody in marketing should figure out that they could sell four times as many at half the price.
Pan said:
Hi Peter
Many years ago i thought the bypass test was indeed best and it really helped me design some pretty good sounding (at the time) gear. It is not so easy to implement though.
It's biggest problem is the quality of the input signal. The Marantz HDAMs have a stronger signature than some of the opamps mentioned above and would certainly make a similar sonic signature less obvious.
If you have a high end analogue source and discrete or tube amplification all these opamps will sound very invasive in a bypass test. If your source is discrete solid state analogue a tube stage under test will have an easily discernible signature as well.
It's only easy using a "normal" cd player with a lousy multi-opamp output stage. Then most additional opamps sound amazingly transparent
I am happy to disclose the non-secrete Marantz SA11 with HDAMs.
A year ago I was looking for a truly hi-end CD player and I borrowed some of the best CD players from the only two hi-end shops in Sydney, as well as from friends for evaluation. I made A-B comparisons at home with my system. Brands included Burmester, Mark Levinson, Audionote, Meridian G08, Musical Fidelity, Audio Research, ...
Our ex-forum member Hi-Fryer told me that his Marantz SA11 sounded better than his Audio Research CD player for CDs. He brought his Audio Research CD player to my home and we listened to it for a whole evening. While he asked me to go to his home to pick up his Marantz SA11, I turned up at his home and he was not there any more. I was saddened because I believed that he had just passed away from his long suffering from blood cancer. Just before he passed away, he still showed enormous enthusiasm in Audio, as he did not want to sell me his GB300 kits (Greg Ball's MOSFET amp) because he wanted to build them for his subs. Note that he spent one hour drive to my home (as he lost his way in the dark) while his health condition was that he couldn't get up by himself from my sofa without help. His house was full of HiFi equipment and his most beloved system was C.J. He was a tube fanatic and never liked SS sound except the Marantz SA11. He lived his life and enjoyed life and the HiFi hobby to the maximum and was never mentally let down by his suffering from blood cancer. We only knew each other for a short period of time but he was a true inspiration to me.
HiFryer told me, forget about all previous Marantz, regardless of prices and statuses. This new Marantz was completely different sounding. Because the Marantz SA11 was by far the CHEAPEST (at RRP AUD$4,250, or USD$3,900 at today's rate) comparing to all other brands listed in above, I took HiFryer's words and bought a brand new one from a cheaper source without listening to it first. In my CD player contests, I much preferred the Meridian G08 but I would not be able to get it for under USD$6,000 at the time.
I initially found the Marantz SA11 sounding hard and non-musical. After some run-in and improvement on the overall system (mainly speakers), I now agree the Marantz SA11 sounds as good as I can expect from a CD player. It possibly beats everything I have heard, and I have taken it to my friends' places and compared it to others. I paid $45 for a service manual and studied the schematic. I also upgraded its components (with 40+ parts), which, when I look back, seem to be unnecessary (except the output caps). There is no opamp used in the signal path. The HDAMs look fantastic. The HDAMs are simply Marantz' discrete JFET circuits for the analogue output stage on this machine and indeed, the components of the HDAMs occupy the majority of a big board. It looks massive but comparing to opamps there are very few parts. I have to say on this machine the HDAMs sound a lot better than any opamps. The Marantz SA11 renders sound like no other players. A violin sounds a violin, not viola or cello. When I was at high school I played violin and I performed in a group. I have also a top of the range European Petrof upright piano (A135) at home that gives me the "live" reference sound when I design and build my system. I have not compared the Marantz to the G08, but obviously, all CD players I mentioned above, perhaps with the exception of the G08, in my view, do not render sounds as accurate as the Marantz. It is a very transparent player. Once you replace those output Elna Silmic caps with good MKPs, the Marantz house sound would be gone.
I would say the Marantz SA11 is the best buy. People are starting to realise that. In the Hi-Rez forum, in the past months, people preferred the Marantz (Marantz SA7 and SA11S2) over the Cary306, Esoteric, ... basically everything else. The Marantz has been the hottest topic in the Hi-Rez forum for a few months now.
Digital / CD technologies are advancing rapidly, and we can not compare yesterday's players with today's players.
A year ago I was looking for a truly hi-end CD player and I borrowed some of the best CD players from the only two hi-end shops in Sydney, as well as from friends for evaluation. I made A-B comparisons at home with my system. Brands included Burmester, Mark Levinson, Audionote, Meridian G08, Musical Fidelity, Audio Research, ...
Our ex-forum member Hi-Fryer told me that his Marantz SA11 sounded better than his Audio Research CD player for CDs. He brought his Audio Research CD player to my home and we listened to it for a whole evening. While he asked me to go to his home to pick up his Marantz SA11, I turned up at his home and he was not there any more. I was saddened because I believed that he had just passed away from his long suffering from blood cancer. Just before he passed away, he still showed enormous enthusiasm in Audio, as he did not want to sell me his GB300 kits (Greg Ball's MOSFET amp) because he wanted to build them for his subs. Note that he spent one hour drive to my home (as he lost his way in the dark) while his health condition was that he couldn't get up by himself from my sofa without help. His house was full of HiFi equipment and his most beloved system was C.J. He was a tube fanatic and never liked SS sound except the Marantz SA11. He lived his life and enjoyed life and the HiFi hobby to the maximum and was never mentally let down by his suffering from blood cancer. We only knew each other for a short period of time but he was a true inspiration to me.
HiFryer told me, forget about all previous Marantz, regardless of prices and statuses. This new Marantz was completely different sounding. Because the Marantz SA11 was by far the CHEAPEST (at RRP AUD$4,250, or USD$3,900 at today's rate) comparing to all other brands listed in above, I took HiFryer's words and bought a brand new one from a cheaper source without listening to it first. In my CD player contests, I much preferred the Meridian G08 but I would not be able to get it for under USD$6,000 at the time.
I initially found the Marantz SA11 sounding hard and non-musical. After some run-in and improvement on the overall system (mainly speakers), I now agree the Marantz SA11 sounds as good as I can expect from a CD player. It possibly beats everything I have heard, and I have taken it to my friends' places and compared it to others. I paid $45 for a service manual and studied the schematic. I also upgraded its components (with 40+ parts), which, when I look back, seem to be unnecessary (except the output caps). There is no opamp used in the signal path. The HDAMs look fantastic. The HDAMs are simply Marantz' discrete JFET circuits for the analogue output stage on this machine and indeed, the components of the HDAMs occupy the majority of a big board. It looks massive but comparing to opamps there are very few parts. I have to say on this machine the HDAMs sound a lot better than any opamps. The Marantz SA11 renders sound like no other players. A violin sounds a violin, not viola or cello. When I was at high school I played violin and I performed in a group. I have also a top of the range European Petrof upright piano (A135) at home that gives me the "live" reference sound when I design and build my system. I have not compared the Marantz to the G08, but obviously, all CD players I mentioned above, perhaps with the exception of the G08, in my view, do not render sounds as accurate as the Marantz. It is a very transparent player. Once you replace those output Elna Silmic caps with good MKPs, the Marantz house sound would be gone.
I would say the Marantz SA11 is the best buy. People are starting to realise that. In the Hi-Rez forum, in the past months, people preferred the Marantz (Marantz SA7 and SA11S2) over the Cary306, Esoteric, ... basically everything else. The Marantz has been the hottest topic in the Hi-Rez forum for a few months now.
Digital / CD technologies are advancing rapidly, and we can not compare yesterday's players with today's players.
I think I will do the bypass test when I have the time.
My Marantz SA11 has a 2.2uF output cap. If I bypass the opamp and only use the volume pot, currently 47k, which then connects to the power amp with 10k input impedance without the input cap, it is not ideal but should still work.
I am pretty sure a wire is far more transparent than even the opa627.
But since I am building an active XO, I have to use an opamp or two, at least for the frequencies from 150Hz up. After my opamp tests, I am now thinking about using a single opamp instead of two opamps, even if I have to reduce the slope of the XO.
I am a strong believer that active components like opamps and some passive components like capacitors colour the sound to various extents. Less components in the signal path to accompolish the same electrical objectives is better.
My Marantz SA11 has a 2.2uF output cap. If I bypass the opamp and only use the volume pot, currently 47k, which then connects to the power amp with 10k input impedance without the input cap, it is not ideal but should still work.
I am pretty sure a wire is far more transparent than even the opa627.
But since I am building an active XO, I have to use an opamp or two, at least for the frequencies from 150Hz up. After my opamp tests, I am now thinking about using a single opamp instead of two opamps, even if I have to reduce the slope of the XO.
I am a strong believer that active components like opamps and some passive components like capacitors colour the sound to various extents. Less components in the signal path to accompolish the same electrical objectives is better.
analog_sa
Interesting. I must ask though, your bypass tests where blind and scientifically performed?
HiFiNutNut
Looking forward to your bypasstests.
Don't get me wrong here. I have heard differences in gear that I and other people thought was not possible so I'm not after you in any way here.. it's just that for me it's always more interesting when people actually have prooven the audible difference.
/Peter
Interesting. I must ask though, your bypass tests where blind and scientifically performed?
HiFiNutNut
Looking forward to your bypasstests.
Don't get me wrong here. I have heard differences in gear that I and other people thought was not possible so I'm not after you in any way here.. it's just that for me it's always more interesting when people actually have prooven the audible difference.
/Peter
A friend and I had two identical headphone amps, one using AD711's and one with OPA627's. We checked output levels using a multimeter across the load (a pair of Grado RS series 'phones), to assure consistent loudness.
In A/B/ X comparisons (not blind, but if switching back and forth enough times while enjoying the music, even the person performing the switching can lose track of which is which), there was clearly a preference for one unit, which turned out to be the OPA627's, which consistently sounded much more natural, more dynamic and gave a better sense of the recorded "space". Now the AD711's sound very nice- especially when biased into class-A, but the OPA627's made the sound more like a fine class-A J-fet or tubed headphone amp.
Granted, not a scientifically-controlled "blind", or "double-blind" test, but convincing enough for me. I was surprised at the differences heard between them, and was skeptical at first, not expecting to hear more than a subtle shift of timbre.
-Joe
In A/B/ X comparisons (not blind, but if switching back and forth enough times while enjoying the music, even the person performing the switching can lose track of which is which), there was clearly a preference for one unit, which turned out to be the OPA627's, which consistently sounded much more natural, more dynamic and gave a better sense of the recorded "space". Now the AD711's sound very nice- especially when biased into class-A, but the OPA627's made the sound more like a fine class-A J-fet or tubed headphone amp.
Granted, not a scientifically-controlled "blind", or "double-blind" test, but convincing enough for me. I was surprised at the differences heard between them, and was skeptical at first, not expecting to hear more than a subtle shift of timbre.
-Joe
peranders said:
Yeah, I really like the AD8610 and unlike the 627 it also comes in a dual, the AD8620. It's main drawback is it's +/- 12V supply limit but that can also be an advantage. I've put AD8620's in a few satellite receivers that had single polarity +12V to ground supplies and it really woke them up. It runs well at low supply voltage.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.