|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
180.200.183.199
In Reply to: RE: Tube sound - how much is due to the caps? posted by Bold Eagle on December 13, 2015 at 06:58:28
Any DF above 10:1 is enough IMO&E. Most PP valve amps can achieve that.
I certainly don't think it has ever been the material used in the coupling caps / their internal structure / or even those used in the PSU whose job is to store and deliver DC current.
I think it might have been, and still might be to do with: i)fairly low NFB / basic linearity of the active devices themselves and, ii) very high energy storage in the PSU.
? Thanks to the square of the voltage factor in the storage equation. This factor dominates the storage numbers as squaring even 250 volts is a big number compared to squaring numbers well below 100V.
These are things that actually do distinguish valve amplifier circuits from SS circuits. And might be where we should look.
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Follow Ups:
Tim,
I still believe in the source impedance of the amp + the cable impedance acting with the speaker impedance as a frequency dependent voltage divider that changes the frequency response of the speaker. The theory and math works, I have run several experiments on that, and other well know authorities hold the same opinion. A DF of 10 corresponds to a 0.8 ohm series impedance in the amp, that's quite a bit. My "belief" is not in DF per se; but in the series impedance that the DF represents.
DF is not a measured property, it is a calculated "rule of thumb" factor based on the measured source impedance. It has much less to do with damping than it does with the "voltage divider" effect of the series impedance of the amp and cables interacting with the speaker's modulus of impedance as it varies with frequency. Speakers with perfectly flat impedance have no effect.
My post was not whether the caps made ALL the difference; but whether they contribute, and if they do, how much of the difference they account for. Since bypassing my electrolytic coupling caps makes a noticeable difference in a solid state amp; do the film or paper, or PIO caps make a noticeable difference in a tube amp, as compared to electrolytics?
Jerry
And in most valve power amps that means everywhere.
Having had a PIO coupling cap blow up on me - and take out an OPT in a friend's Radford STA25 - I won't use them. I find them a bit bloated in the bass.
I still think that the two important distinguishing features of PP valve amps cf SS PP amps are the ones that make for the different sound. ? High voltages giving storage that is orders of magnitude higher than SS can hope for, and far less loop NFB.
Can't speak for SET amps.
I once borrowed two well made SE UL 13E1 18w RMS amps for a few weeks. Yet we preferred the system bi-amped into a passive HP @3.5khz with the 'once were LEAKs' giving about 26wpc in total. Pentode 20wc and triode 6wpc.
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Tim,
Calculate the Joules of storage in a Dynaco Mk 3 with a 500 V supply and a 40 uF filter cap, and my NAD C350 with 50 Volts PS and 40,000 uF. It's a long, long way from orders of magnitude.
Jerry
Your comparison is a long, long way from apples to apples.
Deliberate or subconscious, :-)?
If we compared an SS amp that was contemporary with the Dynaco Mk3 and of the same rated power, the storage in the SS amp would be much lower than in the Dynaco, no question
If we recapped a Mk3 today using HV electrolytic caps that would still fit, and used an SS rectifier which would also be fair, where would the Dynaco be?
Warmest
Tim Bailey
Skeptical Measurer & Audio Scrounger
Tim,
I just am questioning the "orders of magnitude" comment. An order of magniture is 10:1, orders (plural) would be 100's or 1000's. That seemed a bit much. Do you have examples to back that up?
Perhaps more important is the impedance of the power supply, since that determines the rate at which the stored energy can be released.
And since when does contemporary factor in. In the first place, there were no SS amps contemporary with the Mk III at introduction. The first I know of was the Acoustech model 1 which had 3000 uF of capacitance at 80 volts. That's still about double the Mk III.
But it's still off the main point. First, can we agree that an OPT tube amp will sound different than a direct coupled SS amp of similar power. Second, I still maintain that source impedance is a major player in the difference in sound between them. And Third, is the difference in coupling caps any part of the difference in sound? I limited my question to coupling caps in my original post and I was not saying it was a major part, only that it might contribute, and asking if it did, how much.
If you read back over the many replies, hardly anybody addresses the original question. Most of what I got back are digressions into power supplies, lectures on EE theory, a lecture on high efficiency speakers, etc. Any question why I'm getting a little testy on this?
Kindest Regards,
Jerry
If you read back over the many replies, hardly anybody addresses the original question.
Perhaps it is because we don't agree with your analysis of what constitutes "tube sound". You find it to be a quantitative FR balance thing where others find it to be of a qualitative nature. For me, it is a level of linearity where the music exhibits superior depth retrieval and focus. Adding resistors to my old Stasis amp to "simulate" the higher source resistance did absolutely nothing to improve its ability to float voices in space like the VTLs. It may have changed the frequency balance a touch here and there, but that's it.
Returning to your question, the type of caps affect the sound when in the signal path of any amplifier. Better ones provide more transparency and top end resolution. Bypassing the 'lytics with more linear films is a partial step in that direction. The sound of caps is not a tubes vs SS question.
Indeed, the sound of the very best SS and tube gear today has converged. Largely because solid state gear has returned to simpler topologies with less corrective feedback (if any) and tube gear uses better passive components (including output transformers) along with balanced operation and heroic power supplies.
BTW, the Audio Research REF75SE with 75 watts/channel has a 500 joule power supply. :)
"Better caps provide more transparency and top end resolution." Your comment. Since tube amps use better quality caps than the typical SS amp, doesn't that mean a qualitative edge for a tube amp over SS?
I also thing you're attributing things to me that I have not said. I have said that the higher source impedance of tube amps does interact with the impedance of the speaker to form a frequency dependent voltage divider. However, I have also stated in this forum that it only works with a speaker who's impedance varies by 4 or 5 to 1 from minimum to maximum. Your electrostatics are not a suitable test. If you doubt this, look at some of the amplifier tests in Stereophile, where the interaction shows up clearly in the amp's frequency response when measured with a simulated speaker load. It's a premise that is well proven and follows from basic EE theory. John Atkinson wrote an article on it, and it was in their archives. I can e mail you a copy if you can't find it; but I'll need your current e mail.
However, that's not enough. Even if you EQ out that frequency response shift, (and I have) there is still some remaining difference in the sound. I'm simply suggesting that the inherently better caps might be part of the answer.
Jerry
Since tube amps use better quality caps than the typical SS amp, doesn't that mean a qualitative edge for a tube amp over SS?
The SS amps in my experience that do use a small one for DC blocking purposes also use high quality models -while the majority are direct coupled and don't require them at all like my '81 Threshold.
I also thing you're attributing things to me that I have not said.
Perhaps I was confused by these observations:
"Second, I still maintain that source impedance is a major player in the difference in sound between them.
...at least 70% of the difference between tube amp with output transformers and direct coupled SS amps lies with the higher source impedance of the tube amps. "
70% huh?
where the interaction shows up clearly in the amp's frequency response when measured with a simulated speaker load.
Not every speaker shares the "simulated" roller coaster response curve. Even then, the results vary by less than a db . Room effects throughout the range can easily overwhelm such a minor EQ variance anyway. Such is a quantitative change, not a qualitative one.
I'm simply suggesting that the inherently better caps might be part of the answer.
Well, sure if you find a SS amp that still uses 'lytics in the signal path! That wasn't the case when I resistively loaded the direct coupled Stasis and found little "improvement".
The source impedance issue is not that difficult to sort out. Cap coupled and transformer coupled (McIntosh) transistor amps have high source impedance, so they should sound like tube amps if that was the cause of tube sound. Then there are OTL tube amps (which I have never heard, unfortunately!). I have heard both cap coupled and transformer coupled sand amps, and, while they sounded good to me, they did not sound like tube amps. I actually own a Marantz and a Sansui cap coupled amp.
I believe the resistor trick was to make the amp work better with speakers designed in the tube era, which sound over damped when driven by modern amps with a high damping factor.
Then there are OTL tube amps (which I have never heard, unfortunately!).
The two I've heard were exceptional. The first were Joule Electra Rite-of-Passage monos using MiG Foxbat 6H33 outputs driving Magneplanar 20.1s at Sea Cliff. Naturally, the rest of the system was darn good!
The other was Atma Sphere MA-2 monos driving Brian Walsh's U-1s in his incredible new room. The sound there was simply fabulous . The front end was commensurately good and his speakers have the current PX panels with higher output capability than mine. Jud Barber is no longer making the former amps, but I'd sure consider Ralph Karsten's work!
Impressive, but I don't think I could afford to retube them let alone buy them! The Mini Beast is more my style.
Dave
That's a pretty cool design! Class A OTL output like Atma Sphere with 6H30 input tubes.
Also fairly simple as OLT's go. One day I'd like to build one.
Dave
Joule Electra's product line used that big a$$ triode in multiple variations.
I was cross confusing another modern Russian military tube, the 6H30 dual triode used in my ARC preamp.
Arguably the first engineer to use both of those tubes was inmate Victor Khomenko with his BAT products. He grew up in a family who lived near and worked for the original Svetlana tube plant in St. Pete.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: