I didn't know that, how does it do that? It sounds like a reason to use harmonic distortion
Please be assured that I was not having a go at you. Your Class D experience seems to be diametrically opposed to mine, that is all.
My personal experience is that I have found it easier to build valve gear which remains stable. I always seem to get oscillation (either parasitic or loop) from SS builds. Maybe its my age too: I can see valve stuff better because it is bigger!hellokitty123 said:
I don't kid myself that valves sound better than SS; when properly done (e.g. PP, GNFB), they should sound the same.
A distorting amp will add harmonics (and intermodulation); it will not add overtones.fb2017 said:
Sorry, I wasn't clear, it was this part I meant. Are you referring to "Simultaneous Masking"? It begs the question how audible the harmonics are in a complex signal.
Each one have his own reasons/theories about this, you can see some here
Intriguing measures on tube amp
With all due respect, you have not a clue about to make ECC82/12AU7 sing, as I said before on another thread, in SRPP the sound is transparent, and REAL measurements made by member DualTriode are amazing.
Obviously you did not see them.
Just in case
http://www.diyaudio.com/forums/tube...ply-design-input-6sn7-srpp-2.html#post5408462
I've seen those results, and have also seen better results with different tube choices or different topologies.
I have plenty a' clue on how to make the 12AU7 sing, but using something more linear is just how I like to start. Drop a 6CG7/6SN7 into the same circuit and you will usually get lower distortion 90% of the time, with little to no changes in parts values. Lower distortion is surely more transparent than higher distortion, is it not? At least to my ears and those I've built and listened with, the 6SN7 types win every time in listening tests and distortion measurements, for no real difference in cost or complexity.
I definitely think the 12AU7/ECC82 does best at much higher voltages than it is commonly used, and it's not a terrible tube, just not my first choice for a linear voltage amplifier stage. It's a great way to go for cathode followers or split load splitters if you run it nice and hot, but where it measures best may not be best for longevity without special considerations.
Different strokes for different folks. If you want to use a specific tube, no harm in that.
I try not to get baited into audiophile terms and religious arguments, but that's just the way I see things.
I wonder...the best amp I've ever heard was both extremely realistic sounding and extremely euphonic. Those are two things that normally should not go together.
I'm still reeling from the supposed fact that tubes all sound the same in their linear range. I've always worked on the assumption that they sound better than SS somehow, all things being equal.
The implications are blowing my mind. Maybe I should just not use tubes at all for linear purposes.
Well, there's too many things going on in a given design to give a blanket statement that all tubes sound the same in their linear range, as you can use some nonlinearities to iron out other nonlinearities, ask a few dudes for their pet topology and you will get different circuits that will work just as well. Some tubes sound and measure poorly as a whole, even when set up in a good operating range that would otherwise give low distortion, simply by how they interact with other sections of the circuit doing their thing. Here's where the waters get murky and choppy, and often quite interesting.
Sort of on topic, but if you were to ask me to explain the best sounding amplifier that I ever heard (that was the second lowest overall distortion I have ever measured from tubes) it would be a bit unexpected-
400 volt regulated supply for whole amplifier (maida)
Lundahl input transformer with 4-1 voltage gain (forget the model offhand)
6SN7 mu follower for voltage gain
6SN7 concertina
EL84 parallel triode connected push pull, 50W, 5k James iron, biased right at max dissipation(!) With fixed bias
About 1/10 of output fed back to input of mu follower for global feedback (between transformer and grid)
I don't have measurements handy (the amp and a bunch of my gear and folders are in storage still) but the best way to describe the "sound" was spooky. the absolutely most engaging, almost unreasonably wonderful sounding setup I've ever heard, and it gave my AMB B22 a thick layer of dust after nearly a year of solid use as my main rig. Only reason it isn't still around, is I had an electrolytic cap explode, taking out the maida, and one of the output transformers, along with a few other misc parts. I'll have to ask my former business partner if he minds if I post the pics and schematic, since we were working on trying to do a commercial model that never went anywhere, but I'm not sure if we ever sold the design.
As for the lowest measuring amp, well, it sounded like crap
It was a choke loaded 6SN7 with an input transformer, an interstage transformer/phase splitter, feeding triode connected EL34, with local feedback to the EL34 cathodes via extra windings on the transformer, and global feedback to the 6SN7 cathode. It measured and sounded like a cheap/generic chip amp
dull and lifeless is how I would describe it.Can you tell we had a thing for iron?
Now that I'm a bit older, and not subject to the whims of a guy with more money than sense, I do things a bit more sensibly.
Last edited:
Why use an input transformer?
This is why I don't trust measurements. Too often have they betrayed the actual sound. There are other variables going on.As for the lowest measuring amp, well, it sounded like crap
Nothing I've designed has sounded lifeless or boring but that might have something to do with the fact I only use open loop concepts and I strive to remove all possible imperfections in the design. However I have heard things that sounded terrible and measured really good. The THD graph almost seems useless.
Measurements are fun and all, but not the final word. Sometimes you just need to trust your own ears.
I used to really, really rely on them, but after selling off most my test equipment and gear a couple years ago, I find I have a lot more fun building and testing. I try to stay out of the audiophile "woo" and go off of electronic theory. I like to go off of data sheets, curves, and good design practices. I use a little bit of global feedback, if the circuit will benefit. I start with linear tubes, so they dont need as much work to iron out the details. If my circuit is set up correctly, the DC voltages pan out good, and it sounds great, I'm happy. Most of my testing is with a basic multimeter, and a bunch of good lossless tracks that I'm used to enough that I'm likely to notice if anything sounds "off" or gross.
An example... My current main amp is a 1.3W push pull using the 6SN7 as output tubes, and it sounds fantastic. I would say honestly up there right with the aforementioned "best" build, in all ways except at lower power. I haven't run it for distortion measurements, and likely won't. I know that it's probably pretty low, and it sounds great. If anyone builds one and measures it, I would love to see the results, merely out of curiosity. In my application, it just plain works. You don't hear the amp, you hear the music, or dialog, or background... It simply spits out an accurate sounding and feeling representation of whatever you feed it, without changing anything for the worse. I'm happy with it, and heartily recommend it to others.
Besides, we're building with tubes. This hobby is a near-lawless wasteland, where anything that sounds "good" and doesn't catch fire is considered a success to the average guy. The design is as much technology as it is art, as much logical as it is spiritual for some. Not everything has to be perfect. Sometimes we choose to use tubes simply because we like them.
Its as much about the journey as it is the destination. Crack open a cold one, get cozy, and build something the way you like it.
I come off as a bit dismissive of some design choices and topologies at times, and it comes off a bit more negative than I intend it to, so I apologize to any I may rub the wrong way with that (especially to the SRPP guys) as its not my intent. I just like certain topological choices more than others, and show a heavy bias for/against tubes that I like, for various reasons.
I used to really, really rely on them, but after selling off most my test equipment and gear a couple years ago, I find I have a lot more fun building and testing. I try to stay out of the audiophile "woo" and go off of electronic theory. I like to go off of data sheets, curves, and good design practices. I use a little bit of global feedback, if the circuit will benefit. I start with linear tubes, so they dont need as much work to iron out the details. If my circuit is set up correctly, the DC voltages pan out good, and it sounds great, I'm happy. Most of my testing is with a basic multimeter, and a bunch of good lossless tracks that I'm used to enough that I'm likely to notice if anything sounds "off" or gross.
An example... My current main amp is a 1.3W push pull using the 6SN7 as output tubes, and it sounds fantastic. I would say honestly up there right with the aforementioned "best" build, in all ways except at lower power. I haven't run it for distortion measurements, and likely won't. I know that it's probably pretty low, and it sounds great. If anyone builds one and measures it, I would love to see the results, merely out of curiosity. In my application, it just plain works. You don't hear the amp, you hear the music, or dialog, or background... It simply spits out an accurate sounding and feeling representation of whatever you feed it, without changing anything for the worse. I'm happy with it, and heartily recommend it to others.
Besides, we're building with tubes. This hobby is a near-lawless wasteland, where anything that sounds "good" and doesn't catch fire is considered a success to the average guy. The design is as much technology as it is art, as much logical as it is spiritual for some. Not everything has to be perfect. Sometimes we choose to use tubes simply because we like them.
Its as much about the journey as it is the destination. Crack open a cold one, get cozy, and build something the way you like it.
I come off as a bit dismissive of some design choices and topologies at times, and it comes off a bit more negative than I intend it to, so I apologize to any I may rub the wrong way with that (especially to the SRPP guys) as its not my intent. I just like certain topological choices more than others, and show a heavy bias for/against tubes that I like, for various reasons.
Last edited:
Using NFB trades lower order harmonics for higher order ones until a certain point.
If I recall, you need to use at least 20db of NFB to eliminate the increase in high order harmonics that it creates. So either no NFB or a lot of it. This is a measurable proven fact of NFB. One of the many reasons I don't use it.
Maybe, I never experienced such a tolerance build up though. Usually there's the initial "wow" factor, if the "wow" factor persists after a week, you know it's golden. There are varying degrees of euphony but generally speaking if you have a difficult time putting the music down and keep telling yourself "just one more" because it feels so good to listen then you've got it.
Last edited:
Member
Joined 2009
Paid Member
You can get used to anything. Except words do have a definition, you can try and change the meaning through usage but it takes more than one person.
I have a question about cathode followers. Or in this specific case WCFs.
I already have a power buffer design that has unmeasurable distortion. I don't need a tube for this.
I'm soon going to build a WCF buffer with 6c33c tubes. Should I expect anything special out of them? Or is it just an overly expensive method of creating a power buffer?
I was contracted by somebody to build this for him with the hopes of it sounding better than his WCF stage made of 6080 tubes. Am I wasting his and my time? Lol I feel like I don't know anything about tubes anymore, my perceptions are shattered.
I already have a power buffer design that has unmeasurable distortion. I don't need a tube for this.
I'm soon going to build a WCF buffer with 6c33c tubes. Should I expect anything special out of them? Or is it just an overly expensive method of creating a power buffer?
I was contracted by somebody to build this for him with the hopes of it sounding better than his WCF stage made of 6080 tubes. Am I wasting his and my time? Lol I feel like I don't know anything about tubes anymore, my perceptions are shattered.
Hah you caught me. What I said was mainly aimed toward H2 and using harmonic structure as a trend toward how something will sound. I still suspect H3 might be heard as harshness and I stay far away from anything higher than that as higher order distortion strays from natural distortion and into electrical distortion which our brains did not evolve to filter out.
Last edited:
You remind me of my Guru Hellokitty. He spent all his money into his audio builsiness designing & building SET amps for years & till this day he’s still riding a motorbike . Audio is a tuff business to be inn but having said that, he’s one of the most knowlegable audio guy that I’ve met. Might be tuff for you guys to swallow but aside from designing there’s really philosophy involved in making an amp sing. This has to come from exprience which also equals an art of sorts. There’s no such thing as not adding anything to the sound. Whatever we use, resistor, topology, tube type etc all affects sound. It’s all a matter of getting everything to gel together to become as one.
Cheers
Cheers
Yeah, but unfortunately it's impossible to test everything. I'm trying to form guidelines for myself to be efficient.
I've only officially built a few of the two dozen or so designs I've created. I've spent endless hours in spice and theory but relatively small amount of time actually building, even then it's been very expensive.
The biggest problem I've had is building prototypes in a mechanically stable fashion. Unfortunately my circuits almost always have sensitive biasing circuits and are relatively complex compared to pure tube circuits. The act of combining complex low voltage circuits with high voltage circuits has been a source of endless troubles for me both electronically and mechanically. It's been a money pit figuring out a methodology of building stable prototypes.
I've only recently been able to finally build a few of my designs due to a limited sponsorship from someone who was impressed with them.
Most of my practical experience building with tubes has been during modding existing amps or during proof of concept prototyping.
Only in the last month or so have I built fully stable versions of some of my designs. The performance is off the charts compared to pretty much anything that I ever recall seeing, but I've been designing with the thought that tubes are plain and simply superior to SS comparatively.
Well, my SS designs are the cleanest, silkiest, smoothest, most hi-fi things I've heard, and my current tube line stage using the 4P1L sounds......like nothing.
If I power buffer the dac and level shift the line stage, they both sound the same.
Maybe this is impressive for some people, maybe not, I don't really have a benchmark to go off of other than my own listening experiences and spec sheets/graphs I've seen. Certainly it's better than any commercial product I've heard. But all of my tube designs go to waste if they sound like nothing. endless hours wasted. I need to figure out how to make them produce the magic I've heard in the past.
At the moment i can only make assumptions based off of this 4P1L stage and what you people tell me.
Spending money to find truth and proper practice is a very expensive endeavor. I wonder how many thousands I could have saved if I had someone to teach me. Any bones you can throw would be appreciated.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.