Up until just recently if anyone had told me that rectifiers sound differently I would have responded with something to the effect that they were crazy. But, I swear they do have different sonic qualities and no I haven't been drinking. I first thought I heard a difference when I experimented with different rectifiers in my 6SN7/6SN7 Aikido P-to-P line stage. I started off with a home made SS plug in rectifier and then progressed to a 5Y3. I thought the
SS sounded a little punchy. I then tried a 5AR4 a JJ that I had handy and then a Japaneese Mullard 5AR4. The Mullard sounded better. Then I tried a Chinese 5Z4P and it sounded really smooth. Then I tried a #80 in a octal to 4 pin adapter and it sounded the best of all.
I ordered (2) #83's to try.
So, the question is am I right? Do rectifiers really sound different. Even the wife admitted she heard some difference after telling me just play the damn thing.
I also tried different rectifiers in my Stereo 70. The Japaneese Mullard first followed by the JJ followed by the SS plug in followed by a 5U4. To my ears the Mullard sounded really good followed by the 5U4. The 5U4 has a really nice smooth midrange and this is after setting the bias to the correct setting because it was lower with the 5U4 than the 5AR4.
Am I totally nuts?
SS sounded a little punchy. I then tried a 5AR4 a JJ that I had handy and then a Japaneese Mullard 5AR4. The Mullard sounded better. Then I tried a Chinese 5Z4P and it sounded really smooth. Then I tried a #80 in a octal to 4 pin adapter and it sounded the best of all.
I ordered (2) #83's to try.
So, the question is am I right? Do rectifiers really sound different. Even the wife admitted she heard some difference after telling me just play the damn thing.
I also tried different rectifiers in my Stereo 70. The Japaneese Mullard first followed by the JJ followed by the SS plug in followed by a 5U4. To my ears the Mullard sounded really good followed by the 5U4. The 5U4 has a really nice smooth midrange and this is after setting the bias to the correct setting because it was lower with the 5U4 than the 5AR4.
Am I totally nuts?
Maybe. But maybe not.
First thing I would check is the actual delivered voltages. Second thing would be what the ripple waveforms look like. Third, I'd check the amplifier output for noise differences. If all of those look about the same, I'd consider doing a more controlled listening test- susceptibility to unconscious preconception is an attribute of normal human brains. So if there isn't anything there, you're not nuts (or you might be, but this has nothing to do with it), but if there is, you've got some interesting experiments to do.
First thing I would check is the actual delivered voltages. Second thing would be what the ripple waveforms look like. Third, I'd check the amplifier output for noise differences. If all of those look about the same, I'd consider doing a more controlled listening test- susceptibility to unconscious preconception is an attribute of normal human brains. So if there isn't anything there, you're not nuts (or you might be, but this has nothing to do with it), but if there is, you've got some interesting experiments to do.
I had already looked at B+ differences and written them down.
Location for listening is exactly the same. One easy chair positioned 10 ft away from the speaker arrays.
Even had others switch the rectifiers with me not knowing which was which and they recorded the results.
SS rectifiers highest B+ highest ripple.
5AR4's within 5 volts of each other ripple roughly the same.
5U4 lowest B+ ripple slightly lower.
#80 B+ same as 5Y3 ripple same.
Don't know if pink noise as source and calibrated microphone with output to plotter would show any differences.
Location for listening is exactly the same. One easy chair positioned 10 ft away from the speaker arrays.
Even had others switch the rectifiers with me not knowing which was which and they recorded the results.
SS rectifiers highest B+ highest ripple.
5AR4's within 5 volts of each other ripple roughly the same.
5U4 lowest B+ ripple slightly lower.
#80 B+ same as 5Y3 ripple same.
Don't know if pink noise as source and calibrated microphone with output to plotter would show any differences.
So each rectifier produced different B+ and ripple, so of course they would make the amp sound different. Control for these things so that all rectifiers produce the same and then test. Somewhat difficult as it would require each rectifier be in a different circuit, which may then sound different.
Probable virtually possible to make a controlled experiment, so we will always be uncertain o the differences.
Or said more simply: they sound different cuz they are different.
Probable virtually possible to make a controlled experiment, so we will always be uncertain o the differences.
Or said more simply: they sound different cuz they are different.
In the case of a PSU containing chokes, the different rectifier resistances will offer different amounts of damping of subsonic resonances.
My guess is that the main effect will be changes in harmonic structure of the ripple. How much effect this has will depend on PSRR of the amplifier, and its supply IMD performance.
My guess is that the main effect will be changes in harmonic structure of the ripple. How much effect this has will depend on PSRR of the amplifier, and its supply IMD performance.
Well then you're not comparing those rectifiers anymore..... you're comparing different circuits altogether
YES, as you say, I guess the outcome is they ARE different, but simply because of voltage drop and internal resistance, which makes the supply different (and is no surprise by the way!).... but it's not because of magical properties of such or such tube or rectifier... with good engineering most types or rectifiers should be able to produce an acceptable output!
I have mentioned two individual tests. Power supply in the line stage contains no chokes.
The stereo 70 has a choke in the power supply.
The power supply on the Aikido is extremely clean. The Dynaco is cleaner than stock form will measure if desired.
Haven't put my head in a vice to prevent movement yet but have used the same distance from the equipment and same relaxed seating.
As mentioned possibility of different internal resistances internal to the rectifier tubes. Power supply difference due to resistance changes between tubes resulting in a 5 volt or so difference between the same model number of tube. Don't know if 5 volt B+ difference is enough to hear.
Could there be any capacitence difference between the tubes that might result in a apparent sonic difference?
The stereo 70 has a choke in the power supply.
The power supply on the Aikido is extremely clean. The Dynaco is cleaner than stock form will measure if desired.
Haven't put my head in a vice to prevent movement yet but have used the same distance from the equipment and same relaxed seating.
As mentioned possibility of different internal resistances internal to the rectifier tubes. Power supply difference due to resistance changes between tubes resulting in a 5 volt or so difference between the same model number of tube. Don't know if 5 volt B+ difference is enough to hear.
Could there be any capacitence difference between the tubes that might result in a apparent sonic difference?
This is such an old hat that i find it incredible it's even worth discussing. Of course rectifiers sound different. Even if followed with the fanciest shunt reg, they still sound distinctly different. Anyone with ears knows this. And no, it cannot be explained with Ohm's law or gross simplifications.
1. Our audio memory is deceptive.
2. Very slight changes in our head position will affect the character of what we think we are hearing.
Here is a example of a recent experience I have comparing recordings played on the same equipment.
I compared two DVDA discs I created from one song file recorded from an LP at 24/192 resolution. The discs were burned on different computer DVD recorders and different with 2 different programs and two different brands of discs.
I layed on the couch to listen through the first song on each disc. I paid attention as critically as I could to the levels of the highs and lows at various points in the song.
I switched discs to the second one, and thought immediately that the levels were softer.
I switched back to the first disc and was struck by the new indicision I was havibg on the difference I thought i had heard on the other disc.
I swithced again to the second disc and realized I could hear everything identically to the the first disc.
I looked at the file structures with a hex editor and the files were identical.
Take your own first impressions with a "table spoon" of salt. I would have sworn I had heard a major difference on my first comparison that turned out to be nothing.
20
Rectifier is non-linear device, filter is resonant device, so of course it would be reflected on dynamics! That's why I stabilize critical voltages in my amps. It is cheap today, unlike when voltage and current stabilization was considered as economical only for measurement devices and similar critical equipment.
I found experimentally that string of zeners and source followers is enough for tube based power amps. However, even in such case type of filter is critical: resonant tank before source follower is odd: filtering ripples way below the level that is needed it adds non-linear, frequency dependent complex resistance that leads to problems on peaks of signal envelope. Avoid chokes when use pass regulators, even so simple like source followers.
I found experimentally that string of zeners and source followers is enough for tube based power amps. However, even in such case type of filter is critical: resonant tank before source follower is odd: filtering ripples way below the level that is needed it adds non-linear, frequency dependent complex resistance that leads to problems on peaks of signal envelope. Avoid chokes when use pass regulators, even so simple like source followers.
A very interesting post. It may be just "poor listening skills" but it may be more than that. Why would you compare files once you knew they sounded identical? Did you have a nagging doubt, that after all, they didn't? Wanted some additional proof?
And, of course, the fact that files are identical on the optical media bears no relevance on how they sound. Remember the clock?
From a guitar amp point of view, the rectifier makes a big difference in performance.
SS - would maintain a tighter voltage / less diode drop and sag. This would allow for tighter punchier bottom end, faster/stiffer response.
Tube Rectification - Slower sounding and more dynamic in response to playing, due to having sag during attack and peaks.
Having said that, if you had a tube rectifier that was over sized and able to maintain the desired voltage without sagging I doubt you would hear as much of a difference.
In my opinion SS is more practical for music reproduction.
Tube is good for some music production, sometimes not.
SS - would maintain a tighter voltage / less diode drop and sag. This would allow for tighter punchier bottom end, faster/stiffer response.
Tube Rectification - Slower sounding and more dynamic in response to playing, due to having sag during attack and peaks.
Having said that, if you had a tube rectifier that was over sized and able to maintain the desired voltage without sagging I doubt you would hear as much of a difference.
In my opinion SS is more practical for music reproduction.
Tube is good for some music production, sometimes not.
It wasn't a planned listening experiment. I had the first disc for a long time and was getting a second DVDA burning system set up so I used an old 24/192 sound file as a test burn. I listened to it and then thought about doing the listening comparison to the old disc. I was really just doing a test on the burning system which had a different burner that burned at higher speeds and a Linux based software compared to the Windows program the old disc was burned on.
Remember the clock?
The discs were played on the same DVDA player in their native 24/192 resolution. The files were never converted or down-converted from original 24/192 wav files except as is necessary for the DVDA formating program to create the DVDA file system.
Several years ago I decided to do some rectifier tube rolling in my little 6V6 PP amp. The amp uses a 5U4,but I tried just about every rectifier with the same pinout directly,or indirectly heated.
I finally came back around to the 5U4GT...I preferred it over the 5U4GB's I had on hand. -Yes,I could hear a difference between the GT and GB tubes.
Lately it's had the Russian "Winged C" 5U3C/5C3S rectifier in there,and I might like it even more than the 5U4GT.
I put it all down to the various tubes impedance/voltage drops. The B+ was different for every tube -that had to account for at least part,if not all of the difference I was hearing.
I finally came back around to the 5U4GT...I preferred it over the 5U4GB's I had on hand. -Yes,I could hear a difference between the GT and GB tubes.
Lately it's had the Russian "Winged C" 5U3C/5C3S rectifier in there,and I might like it even more than the 5U4GT.
I put it all down to the various tubes impedance/voltage drops. The B+ was different for every tube -that had to account for at least part,if not all of the difference I was hearing.
This has nothing to do with the way the clock is derived and the effect pit jitter has on it. The fact that different blanks and different burners produce different sounding copies has been discussed and explained ad infinum on many forums.
Careful when those arrive. They probably won't like working in your existing amp without PS mods.