I have to wonder if you're now in an area where some minor system tweak would be the deciding factor in which part you preferred, be it cables, some cap in the circuit, etc. Tweeter level?
It would also be interesting if you could get somebody else to plug in the parts based on a coin flip or something, then see if you can really tell the difference. I know the pros and cons of this type of test, but it still says quite a lot if you can or can't consistently find the same characteristics in the parts. It would also help if someone did that, then you listened for a couple days to see if one or the other really caused more listening fatigue.
It would also be interesting if you could get somebody else to plug in the parts based on a coin flip or something, then see if you can really tell the difference. I know the pros and cons of this type of test, but it still says quite a lot if you can or can't consistently find the same characteristics in the parts. It would also help if someone did that, then you listened for a couple days to see if one or the other really caused more listening fatigue.
HiFiNutNut said:
Hi Bill
You don't mention (or i've missed it) the 12B4 topology. Is it a cathode follower? In any case i am not familiar with 12B4 but something must be really broken if it sounds like a lousy opamp. Unless, it is a cathode follower
The last tube pre i built (mercury rectified, transformer coupled) would be quite easy to tell apart from a 627. In fact even my modest Pass Aleph 1.0 clone is simply incomparable to any opamp.
Oshifis,
I believe the opa134 should sound the same as the opa2134.
I have not heard of THS4031/4032. If anybody sends me one I am happy to test it. I am not buying any more opamps for this round of opamp tests unless I am sure that they are better than opa627 or LM4562.
Conrad,
I do have consistent results.
Peranders,
I don't have a LME49710. It would be in the same family as LM4562.
Analog_sa,
My 12B4A is only a simple common-cathode with no cathode bypass. It is basically a mono construction with oversized / over spec / high quality parts. It initially had only LRCRCRCRC supply but, having no time to play around I paid Joe R to install twin Allen Wrights superegs on it. I have briefly compared it to Allen's FPV preamp at Joe R's place and they sounded different - I personally preferred the 12B4A sound. 12B4A is not everyone's cup of tea. Many tube lovers love the beautiful harmonics generated from tube gears, a kind of artificially enlarged sound stage, artificial depth, etc, which can be very nice, I confess. Marantz 7, Audionotes, etc, come into my mind. However, what I am pursuing is not "good sound" but accurate reproduction of sound. I found the 12B4A to be quite accurate. My friends commented that it is solid state sound without the harshness of the solid state. I agree.
On another note, I found accurate sounding components connected to a system with a lot of distortions can sound worse. For example, I believe the Meridian G08 and the Marantz SA11 are very accurate. They sounded very fine on my previous set up. But on my friends' "audiophile" systems, they sounded cold, sterile and even "terrible". Sometimes the inaccuracy of a component (like the Marantz 7) can mask a lot of nastiness of sound in a problematic system and make it sound much nicer. An accurate component, like the 12B4 or opa627, having no or little sound of its own, exposes the problems of other components in the signal path.
Use the opa627 in an accurate system and see if it sounds bad!
I believe the opa134 should sound the same as the opa2134.
I have not heard of THS4031/4032. If anybody sends me one I am happy to test it. I am not buying any more opamps for this round of opamp tests unless I am sure that they are better than opa627 or LM4562.
Conrad,
I do have consistent results.
Peranders,
I don't have a LME49710. It would be in the same family as LM4562.
Analog_sa,
My 12B4A is only a simple common-cathode with no cathode bypass. It is basically a mono construction with oversized / over spec / high quality parts. It initially had only LRCRCRCRC supply but, having no time to play around I paid Joe R to install twin Allen Wrights superegs on it. I have briefly compared it to Allen's FPV preamp at Joe R's place and they sounded different - I personally preferred the 12B4A sound. 12B4A is not everyone's cup of tea. Many tube lovers love the beautiful harmonics generated from tube gears, a kind of artificially enlarged sound stage, artificial depth, etc, which can be very nice, I confess. Marantz 7, Audionotes, etc, come into my mind. However, what I am pursuing is not "good sound" but accurate reproduction of sound. I found the 12B4A to be quite accurate. My friends commented that it is solid state sound without the harshness of the solid state. I agree.
On another note, I found accurate sounding components connected to a system with a lot of distortions can sound worse. For example, I believe the Meridian G08 and the Marantz SA11 are very accurate. They sounded very fine on my previous set up. But on my friends' "audiophile" systems, they sounded cold, sterile and even "terrible". Sometimes the inaccuracy of a component (like the Marantz 7) can mask a lot of nastiness of sound in a problematic system and make it sound much nicer. An accurate component, like the 12B4 or opa627, having no or little sound of its own, exposes the problems of other components in the signal path.
Use the opa627 in an accurate system and see if it sounds bad!
OK, last night I spent a couple of hours to have another go at the LM4562 and OPA627, after which, I was still undecided. A few more days are needed.
I first plugged in the LM4562 for 20 minutes. This time I played a familiar symphonic music. It did not sound good at all. So bad that I had doubts on all of my previous tests. I could not help swapping the LM4562 with the OPA627, so the good sound came back. There must be something wrong there. I suspected that the contacts to the DIP8 socket were bad with the LM4562, as I found the contacts were quite loose when I inserted the LM4562 previously. My guess was right. I replaced the OPA627 with the LM4562 once again and made sure the contacts were as tight as possible. Now the LM4562 sounded like a LM4562.
This year I learnt a lesson. Loose contacts of loudspeaker cable joints can destroy the sound. When I designed and tested my speakers, the wires were screwed on some terminals only. In time the copper and other metal materials could change shape from the pressure of the screw and got oxidised. The added resistance was only one issue, loose contacts, while measuring OK with a multi-metre, could not deliver large current on music transients, making the sound dull or bad. It was not so obvious while I was testing my speakers. I just found no matter what I did to the speakers they did not sound good! Until one day, the loose contacts were so bad that the problem was exposed. I soldered the connection joints and all of a sudden, my speakers were upgraded 3 levels up with a bit of solder!
The same can apply to plugging in an opamp into a DIP socket, and others.
Back to the OPA627 and LM4562:
I am still undecided on which one I will be using. I have not found out if on long listening I will be tired of the LM4562 or not. I have found out that I won't be tired of the OPA627 on long listening.
The LM4562 still sounds super. Very dynamic, very rich, very fast, life like. In these areas, it beats the OPA627. However, when it comes to human voices, in comparison, the OPA627 delivers smooth, accurate, sweet, beautiful and very convincing sound, while the LM4562, to a very small extend, sounds a bit bright, making human voices a bit "harder" than "softer".
So if you are looking for a sound that is sweet, mellow, laid-back, calm and especially you are fuzzy with vocals, get the OPA627, for sure.
If you are looking for a sound that is exciting, inspiring, forward, dynamic and especially you are looking for the feeling of "live-performance" on instrumental music, get the LM4562, for sure.
I don't know which is better, or which I will prefer, yet. It is like comparing the Kawai piano to the Yamaha piano. The Kawai is mellow sounding, the Yamaha is bright sounding. Both are similarly good and are at the same level. Ironically, when I bought a piano for my kids a few months ago, I spent most of my time looking at the Kawai but ended up buying the Petrof, a decision made within 20 minutes.
I first plugged in the LM4562 for 20 minutes. This time I played a familiar symphonic music. It did not sound good at all. So bad that I had doubts on all of my previous tests. I could not help swapping the LM4562 with the OPA627, so the good sound came back. There must be something wrong there. I suspected that the contacts to the DIP8 socket were bad with the LM4562, as I found the contacts were quite loose when I inserted the LM4562 previously. My guess was right. I replaced the OPA627 with the LM4562 once again and made sure the contacts were as tight as possible. Now the LM4562 sounded like a LM4562.
This year I learnt a lesson. Loose contacts of loudspeaker cable joints can destroy the sound. When I designed and tested my speakers, the wires were screwed on some terminals only. In time the copper and other metal materials could change shape from the pressure of the screw and got oxidised. The added resistance was only one issue, loose contacts, while measuring OK with a multi-metre, could not deliver large current on music transients, making the sound dull or bad. It was not so obvious while I was testing my speakers. I just found no matter what I did to the speakers they did not sound good! Until one day, the loose contacts were so bad that the problem was exposed. I soldered the connection joints and all of a sudden, my speakers were upgraded 3 levels up with a bit of solder!
The same can apply to plugging in an opamp into a DIP socket, and others.
Back to the OPA627 and LM4562:
I am still undecided on which one I will be using. I have not found out if on long listening I will be tired of the LM4562 or not. I have found out that I won't be tired of the OPA627 on long listening.
The LM4562 still sounds super. Very dynamic, very rich, very fast, life like. In these areas, it beats the OPA627. However, when it comes to human voices, in comparison, the OPA627 delivers smooth, accurate, sweet, beautiful and very convincing sound, while the LM4562, to a very small extend, sounds a bit bright, making human voices a bit "harder" than "softer".
So if you are looking for a sound that is sweet, mellow, laid-back, calm and especially you are fuzzy with vocals, get the OPA627, for sure.
If you are looking for a sound that is exciting, inspiring, forward, dynamic and especially you are looking for the feeling of "live-performance" on instrumental music, get the LM4562, for sure.
I don't know which is better, or which I will prefer, yet. It is like comparing the Kawai piano to the Yamaha piano. The Kawai is mellow sounding, the Yamaha is bright sounding. Both are similarly good and are at the same level. Ironically, when I bought a piano for my kids a few months ago, I spent most of my time looking at the Kawai but ended up buying the Petrof, a decision made within 20 minutes.
Hi Bill,
I am about to order a bunch of 2N5484 JFETs in a little over a week due to other commitments. I will test them and send you 2 matched pairs so you can make a cascode with a matched bias of 1.2 to 1.3 mA to bias the OPA627s to Class A. Employ them and listen. Then decide!
I am not sure how long it will take to mail them to you from New York. So, can you wait?
Also, don't forget to pay attention to the entire circuit that surrounds the op amps! Conrad already suggested that this may be beneficial. I agree.
Regards//Keith
I am about to order a bunch of 2N5484 JFETs in a little over a week due to other commitments. I will test them and send you 2 matched pairs so you can make a cascode with a matched bias of 1.2 to 1.3 mA to bias the OPA627s to Class A. Employ them and listen. Then decide!
I am not sure how long it will take to mail them to you from New York. So, can you wait?
Also, don't forget to pay attention to the entire circuit that surrounds the op amps! Conrad already suggested that this may be beneficial. I agree.
Regards//Keith
Interesting about connections. I always think my system sounds better when I clean or just unplug and replug all the interconnects. Sockets are *not* preferred for maximum performance, especially with wide bandwidth chips, but how else to change them? I use sockets on almost everything. The popular view is that machined contact gold flashed sockets are best, but actual tests some years ago showed that tinned (matches the ic lead) high pressure double leaf contacts were actually lower resistance and more reliable over time. I still use the machine contact type- I never claimed to be completely logical. Does anybody use a bit of DeOxit on chip leads?
HiFiNutNut said:
thanks, your comparison has made very interesting reading.
I recently built ESP's linkwitz active crossover project, and was actually recommended here to just use the OPA2134.... I was not terribly happy with the end result, which was a little mushy and dark. Then I dropped the LM4562 into the same circuit instead, and I thought there was a positive and clear difference. I really find it to be very clear and neutral, if such a thing exists, and all the things you describe above. I have not tried the OPA627 and at the price I'm not sure if I would. One big thing that the LM4562 has going for it is that the price is quite reasonable.
I can confirm this.
I managed to improve LM4562 performance by pulling the output in class A (4K7 Arcol resistor or 3mA CCS from output to -15V), and applying noise-gain manipulation (placing a resistor between both inverting and non-inverting input). I experimented with 5 ... 100 times the NFB resistor value.
http://www.analog.com/Analog_Root/s...ols/interactiveTools/stability/stability.html
I used small film caps for decoupling (between +15 and -15V). Very important! is to place a 82 Ohm resistor in series with the OP-amp output to reduce the influence of any capacitive load, this is a must when using LM4562.
Resistors are very important, I prefer Arcol 0.1%, and Vishay bulk metal foil resistors. It's also possible to use high-wattage resistors like Beyschlag 1W matal film. (less noise). Resistor values should be kept as low as possible to keep noise levels down.
OP-amps seem to perform worst when used as unity-gain buffer (filters), and there is very little to do about it, except using noise-gain manipulation.
After experimenting with OP-amps for a long time period, I no longer use them for High-End audio, even the very best aren't acceptable anymore after listening to discrete JFET Op-amps.
Discrete JFET amplifiers (even simple ones) outperform OP-amp (buffers) easily. They provide much higher resolution, and more fluent, balanced sound.
Well designed tube OP-amps / buffers (< 0.001% THD) can provide even better performance.
The power amplifier also has big impact, especially on digital sound quality.
By connecting the NFB loop directly to the output stage (standard configuration), the speakers will disrupt the NFB loop, smearing / distorting the sound, as each voice coil produces an ac voltage when moving in the strong magnetic field.
I now connected the NFB loop to the driver stage, isolating the speaker from the NFB loop. The MOSFET power amplifier design I used allows for this. With the highly increased resolution and transparence of the power amplifier, it becomes much easier to evaluate different (OP) amps.
Here is a link that provides some interesting information about this subject:
http://www.altmann.haan.de/splif_page/
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.