i agree, single ended tube amps are no global feedback types in general and so open loop gains are quite low...so here we see the difference between tube and ss amps as being about large difference in open loop/closed loop gains...
Leach expounded on this in the 80's...when the TIM debate was a hot topic then....Leach and Cordell had a rather animated debate about this at the Audio Magazine in the 80's..
Cordell is still around and i wonder how much his reasoning has changed since..
Thanx Chris for reminding everyone, you are not the 1st to state that in this thread, it needs emphasizing.
If one wants to design an amplifier with no regard to the speaker system used, the amplifier should have lowish output impedance then it doesn’t care (there are exceptions) what speakers it is hooked up to.
That eliminates talk of most SE designs.
dave
Sorry who says?
The absolute RESOLUTION of a reproduction system is the vital element, nothing to do with linearity, but I am sure on that score you would also be shocked to see the THD of many "hi end" headphones.
Sennheisers are a blunt old hammer compared with an ESL, and there are now ESL headphones.
I have nothing against hearsay, but attempting to use it as validation in a critical analysis situation is doomed to failure,- at which point I give up here straight away.
Some people like to be in denial.
I am not one of them.
In my general experience, class A designs or designs that are made to avoid producing higher order harmonics that have distortions hovering around -120db or less all seem to sound the same. Second harmonic doesn't seem to be audible in any amount you can expect to see in a properly functioning amplifier.
Variable impedance is an option.
As I stated earlier a variable impedance super triode was something I was going to make for a friend and probably make a few PCBs if people wanted some but I struggle to see the point beyond 50W amp with a glowy light on it.
Even SET designs without loop feedback can have adequately low output Z, if properly proportioned. Loudspeaker damping R is the sum of the driver's voice coil resistance plus source effective resistance, so the amplifier plus wiring R only needs to be small compared to VC resistance to not matter. A factor of 5 is plenty in the real world of loudspeakers. More is more, but not important.
The conjugate to this is the improvement in all distortions that comes along with the lighter loading that gives better damping. Why, oh why, do folks building SET amplifiers compulse about how many WATTS they make? And on a linear scale, rather than log, which is extra ridiculous.
Could we all be adults and agree to never use the term WATTS as a figure of merit for amplifiers? dBW is acceptable at the adults table. WATTS is reserved for heat generation, or at the children's table.
Fat chance.
All good fortune,
Chris
The conjugate to this is the improvement in all distortions that comes along with the lighter loading that gives better damping. Why, oh why, do folks building SET amplifiers compulse about how many WATTS they make? And on a linear scale, rather than log, which is extra ridiculous.
Could we all be adults and agree to never use the term WATTS as a figure of merit for amplifiers? dBW is acceptable at the adults table. WATTS is reserved for heat generation, or at the children's table.
Fat chance.
All good fortune,
Chris
Another way of saying what we have been talking about for some time.
A Lot of sets have a damping factor less than 5.
I don’t know whether that is true or not, but it does point out that all these things change with frequency & dynamics. The single numbers we have been throwing around are a place holder for that.
That does not only apply to SETs. I find in general people think they need a lot more than they do.
Perhaps a symptom of the whole male “mine is bigger than yours” syndrome.
dave
You do realize that any particular loudspeaker output requires exactly the same voltage and current at the terminals, irrespective of the nature of the amplifier and irrespective of any other networks in parallel to the loudspeaker. Right? Just so we're on the same page.
All good fortune,
Chris
All good fortune,
Chris
Thank you.
I am not sure how to answer your question, can you be a bit more precise? I suspect we agree, but I don't want to say yes unless I fully understand you.
I suppose my position, backed up by measurement, is that the dB-SPL of the driver is proportional to the current and current only.
What I have also been able to confirm, let us say there is a 30 degree current phase angle, then we hear that sound delayed by 30 degrees. Again, this is really definitive proof that we are listening to the current of the amplifier.
Think of it this way, the voltage of the amplifier is simply the potential for current, but the actual current produced, this is what we end up listening to.
It was not easy to come to that conclusion, but there are now significant people who agree with me.
Cheers, Joe
PS: That 30 degrees shows up in crossover development, you have to allow for the phase shift.
Last edited:
Are they able to post some evidence so we can have a proper discussion?
I posted evidence, quite a bit, three times on blowtorch and got ignored, except the third time Scott Wurcer realised that there were numbers that corroborated the change of current with dB-SPL output, but it was rounded off to a dB. He then asked if that correlation would apply not rounded off to 0.1dB and I said that I will take up that challenge, in time.
Would you like me to post that same information here for the FOURTH time?
Maybe you won't ignore it this time? Maybe you will look at the numbers, look at the measurement, and at least try to understand them? Because clearly the last time you did not.
You know, I am sorry to say it, but I detect a degree of intellectual dishonesty when you ask for something that I have already furnished you.
Please think about that, OK?
.
Thank you for calling it "evidence" and if it does not show what I think it does, then you tell me. What does it show? You have the floor.
BTW, consider that in the long game, are you prepared to be on the wrong side? Look at my byline below, I stick my neck out, I am at least brave, I am willing to be wrong in order to eventually maybe right. Please don't be just another armchair critic, please.
OK, tell me what the "evidence" shows.
And yes, maybe I can counter that and have a good discussion. Maybe even convince you? Can we both have an open mind? That is the only way to learn. I know that I have an open mind, proved it many times!
Let's take a moment to entertain an idea, in the spirit of the original thread subject.
I'm sure many of you are aware of the mu follower circuit (often full of associated discussion, but little in the way of finished designs) and many of you have read Morgan Jones "Designing Valve Amplifiers" books. In the 3rd edition he brings up for analysis the so-called "beta follower" variation- which gives gain as high as the tube is capable of, and according to him it can present distortion as low as the tube is capable of.
I don't know about you guys, but this sounds like a good candidate to build up an excellent gain stage to test the idea that the tube has it's own "sound" to contribute. It should be fairly easy to pick a nice good all-around type of triode, say the 6SN7, and build up a basic preamplifier using it. Use a mosfet source follower or opamp as a buffer, and attenuate the output to match with other devices that are being evaluated. One could try several different amplification devices, opamps, jfets, mosfets (enhancement and depletion mode), NPN or PNP, transistors, even transformers.
For those allergic to mixing the bit of solid state in there, a basic mu follower or grounded cathode stage could be used instead, with the mu follower needing a bit more voltage, but if the triode is run at the same DC operating points it should be a fair comparison. Run a triode cathode follower instead of a mosfet source follower.
You would need a very good double-blind A/B switching arrangement, a control (or several) and would ideally do the tests with many different people to get a good set of data to dig through to formulate a decent answer.
What do you guys think? Sounds like the ultimate preamp topology shootout could be born from this.
I'm sure many of you are aware of the mu follower circuit (often full of associated discussion, but little in the way of finished designs) and many of you have read Morgan Jones "Designing Valve Amplifiers" books. In the 3rd edition he brings up for analysis the so-called "beta follower" variation- which gives gain as high as the tube is capable of, and according to him it can present distortion as low as the tube is capable of.
I don't know about you guys, but this sounds like a good candidate to build up an excellent gain stage to test the idea that the tube has it's own "sound" to contribute. It should be fairly easy to pick a nice good all-around type of triode, say the 6SN7, and build up a basic preamplifier using it. Use a mosfet source follower or opamp as a buffer, and attenuate the output to match with other devices that are being evaluated. One could try several different amplification devices, opamps, jfets, mosfets (enhancement and depletion mode), NPN or PNP, transistors, even transformers.
For those allergic to mixing the bit of solid state in there, a basic mu follower or grounded cathode stage could be used instead, with the mu follower needing a bit more voltage, but if the triode is run at the same DC operating points it should be a fair comparison. Run a triode cathode follower instead of a mosfet source follower.
You would need a very good double-blind A/B switching arrangement, a control (or several) and would ideally do the tests with many different people to get a good set of data to dig through to formulate a decent answer.
What do you guys think? Sounds like the ultimate preamp topology shootout could be born from this.
I once built a beta follower to use within a phono preamp, and over the few weeks I lived with it I failed to get it to sound good. Whereas mu followers consistently come out neutral in my experience.
Hi 6vheater,
An amplifier through a cable?
Another tube?
A reverb tank?
A volume control?
A feedback network?
An inter-stage or coupling transformer?
There are a ton of different things a tube can drive apart from a speaker. Everyone focused on a speaker, but the question is perfectly valid driving all kinds of other audio devices. All these things can and will impact the sound you hear at the end of the chain. A headphone is just another type of speaker, and mine are 600 ohm in case you were wondering.
-Chris
No, I think you are too caught up in a side topic maybe.
Well, let's see.
An amplifier through a cable?
Another tube?
A reverb tank?
A volume control?
A feedback network?
An inter-stage or coupling transformer?
There are a ton of different things a tube can drive apart from a speaker. Everyone focused on a speaker, but the question is perfectly valid driving all kinds of other audio devices. All these things can and will impact the sound you hear at the end of the chain. A headphone is just another type of speaker, and mine are 600 ohm in case you were wondering.
-Chris
Another type of speaker??
Since when has 600ohm studio line level been used to drive a "speaker"?
Mine are 2K ohm, but I will opt for ESL at the drop of a hat.
Can you tell me how much your little weeny amp will drive a near short circuit at 15khz?
..thought so... sounds to me more like trying to get the client to fit the hat, rather than the hat made to fit the client
How much more can your logic resources stretch?
Agreed. The off topic discussions have been of interest to me so I've been tolerating them but it would benefit everyone if they were placed into their own threads as comments directly related to the thread at hand get swept under the rug pretty quickly.
IMO a super triode is far better suited as the load can be run directly from the tubes plate or cathode (or grid if you're into inverted triodes) without really loading it more than you wish it to. Variable loading, variable distortion, variable output impedance. The sound by design will intrinsically be that of the triode and yet it can directly drive loud speakers of arbitrary power levels.
I already have some pretty sweet super triode designs prepped for PCB but haven't bothered yet because I'm not sure people would be interested. If there are interested people let me know so I can have an excuse to do so and I'll make my own thread for it.
I've tried what you are proposing, using an ultra low distortion buffer (in this case it was a sandman bridge) and putting tubes and such in the path to see if they sound different. This is why I made the thread because I couldn't hear a difference, most of the tests I did were done a while ago though.
I've tried an array of mosfets and bipolars arranged to emulate triodes, the curves look just like DHTs and they use no GNFB, and I've also tried 6sn7s and 4P1ls.
They were arranged in a configuration where there is two of them operating in anti current phase in so one triode (or emulated triode) cancels the distortion of the other before amplifying the signal, also gyrator/mu follower loaded. Actually I think I merged a gyrator and a current mirror for that one (sorry I'm not at home so I can't post a schematic).
I failed to hear anything noticeably different, not that there was no difference, but it just wasn't noticeable to me when I compared them. I will say they sounded darn good though.
I've also tried just a simple 6sn7 with a resistor load and a battery bias on grid, cathode grounded. Varied the bias around, only got varying levels of good or bad, no euphonics.
I've also tried direct drive 6080 tubes with a special variation of the WCF that ensure perfect current sharing between triodes and they were just simply inferior to a good solid state output stage.
I never stated any load. The thread is not about loads unless it is some universal trait related to the topic.
I simply said that I prefer using headphones for testing purposes because they are more detailed than any speaker normal people can attain and they don't have room characteristics to mess with the results. If you disagree that's fine, I would opt to say you've never heard a pair of high end headphones if that is your opinion, however you should slow down and stop making presumptive and aggressive accusations to everyone at every turn.
Last edited:
- Status
- Not open for further replies.