the Silverface I remember ran PPP 6L6 with a 4H/90mA smoothing choke in its screen supply and no UL taps - I think my Bassman 100 was that way.
http://www.thevintagesound.com/ffg/schem/twin_reverb_aa769_schem.gif
I didn't know they had made a series of UL amps - here's a schematic - seems like UL would be ok for bass but maybe not guitar (?)
http://www.thevintagesound.com/ffg/schem/twin_reverb_sf_135_schem.jpg
what is the approximate Z-out of a single push-pull pentode pair of EL34 with ~10-13dB nfb from the output transformer's tap back to the inverter? I think Dynaco was said to have 20dB nfb in their UL connected Stereo 70.
http://www.thevintagesound.com/ffg/schem/twin_reverb_aa769_schem.gif
I didn't know they had made a series of UL amps - here's a schematic - seems like UL would be ok for bass but maybe not guitar (?)
http://www.thevintagesound.com/ffg/schem/twin_reverb_sf_135_schem.jpg
what is the approximate Z-out of a single push-pull pentode pair of EL34 with ~10-13dB nfb from the output transformer's tap back to the inverter? I think Dynaco was said to have 20dB nfb in their UL connected Stereo 70.
Peter, if you can let go for a minute of your conviction that I am wrong, I can help you understand what you're missing here.
For the moment, please forget everything you know about the Thiele-Small model. I think it's keeping you from seeing something much simpler.
Okay: a loudspeaker cone and voice coil have mass. The cone is suspended by a surround and spider which have springiness (compliance). The materials used in the surround and spider, and any induced currents allowed to flow in the voice coil, cause loss of energy from the moving mass - mechanical damping.
Now, any time you have a mass suspended by a spring, you get a mechanical oscillator. When you also have mechanical damping, you get what is called a damped harmonic oscillator ( Damped Harmonic Oscillator ).
A damped harmonic oscillator has a number of properties. One is that it has a mechanical resonant frequency. Another is that it has a Q - a mechanical Q.
The moving parts of a loudspeaker constitute such a damped harmonic oscillator. As such, it has a resonant frequency, and a Q. This is always true - whether it is in a box or not, whether there is a port or not, whether you connect it to an amplifier output or not.
The Q of a mechanical oscillator determines its frequency response near the resonant frequency, and also its transient response, via a Fourier transform. This also applies to the moving parts of a loudspeaker. The Q of the damped harmonic oscillator determines its frequency response and time response.
Notice that we have had no need of the Thiele-Small analysis at all so far. All we needed was a little understanding of a damped harmonic oscillator.
What Thiele and Small did was work out a way to calculate this single mechanical Q of the driver, given other relevant parameters, such as suspension compliance, moving mass, compliance of air sealed in a box, if any, and the output resistance of the amplifier driving the box.
All the Thiele-Small parameters and the equations connecting them all exist for only one reason: so you can calculate the one single Q that matters, i.e., the Q of the damped harmonic oscillator created by the suspended speaker cone.
I am sorry you find me difficult to talk to; I have exactly the same problem with you!
Perhaps it is a second language issue? I have found your posts cryptic at best and unhelpful at worst. For example:
Here you say my calculation (which calculation?) is off, but do not point to an actual error. Then you quote the subscripts Thiele/Small used in their paper, which is nice, but contributes nothing to the discussion - the main point was that a high amplifier output impedance raises the mechanical Q of the attached driver so much that you can no longer obtain a flat frequency response, even in an infinite baffle or infinitely large sealed box. Finally, you threw in a little jibe that suggests that I have no idea how the Thiele-Small model works.
So how am I going to respond to that post, which flatly says I'm wrong, but contains nothing to explain why? I did the only logical thing, which was to ask you what exactly you believed to be wrong.
I still haven't got an answer, but it seems you got hung up on the subscript "c" on some of the Q parameters, and therefore didn't understand that the whole moving system acts as a single damped mechanical oscillator, has only one mechanical Q. Let's hope the first half of this post has helped clarify that for you.
(Caveat: in a tuned-port system, there are actually two coupled damped harmonic oscillators. I hate to think how much additional argument that might cause here, so let's leave it at that.)
Thank you, that is extremely kind of you.
What makes you think I haven't? I worked in loudspeaker R&D years ago; I also created a complete mathematical model for a driver at the time, which did not make the transformations to electrical components that Thiele and Small used (for example, converting mechanical mass to electrical inductance).
Chris' posts have all been helpful, and I don't think there is any disagreement between Chris and myself in this thread.
-Gnobuddy
English is my first language, lol, I'm just to lazy since I knew that this was not going
to go well.
So, here is the problem, you think you know better than T&S and probably anyone else
on the planet as far as speaker design goes.
I went to college, fully understand a damped mechanical oscillator. Studied T&S theory
for an undergraduate project where I used that style of analysis to develop a model for
transmission line speakers - so yes I understand this stuff.
I understand NOW that because as you state: " I also created a complete mathematical
model for a driver at the time, which did not make the transformations to electrical
components that Thiele and Small used (for example, converting mechanical mass to
electrical inductance)."
You are going to use language that is not inline with those of us who've accepted the T&S
nomenclature. I feel like I'm talking to someone who doesn't understand speaker design
or is from another planet. Yes there is a language barrier you think you have something
better than T&S theory and so you don't hear what people are saying.
Really, I don't think that you understand T&S because you think yours is better. The
majority of T&S analysis assumes a voltage source. They briefly explain how to include
the source impedance/resistance for cases where it does play a roll but, generally, a zero
source impedance is assumed.
The fact that you suggest that all tube amps have high output impedance also says to me
that you do not have experience with them. I've used plenty of them and read the specs
and most competent tube designs do not have >5 ohm output impedance unless it was
deliberate. Some tube designs had a pot for variable damping.
Your calculations do not pass a sanity check, and you offered them in complete disregard
of the mountain of test reports and specs out there for tube amps. You live in your own
world where you are right - that is fine enjoy it.
You remind me of an internet audio troll that went by "soundminded" who didn't need
any of that T&S stuff because he knew Newton's laws of motion.
Last edited:
Yes, I think it was the later Silverface that were UL and the highest power, several here
are UL, some have it in the title:
www.thetubestore.com - Later Fender Guitar Amp Schematics
I don't think that people like them for lead but they did try UL. Perhaps it was what gave
the Silverface a bad reputation.
I suppose it's all semantics here, but loudspeakers are electro-mechanical devices. While I can see what you mean, I'd phrase it a little differently.
Here are the definitions I usually work to:
Fs - resonant frequency of the speaker cone, from the spring-mass system of the cone and suspension.
Qms - mechanical damping from the suspension and surround
Qes - electrical damping from the voicecoil when it has a very low impedance across it (most solid-state amps)
Qts - mathematically derived from the first two, 1/Qts = 1/Qms + 1/Qes. That's the total Q of the speaker, lumping the effects of mechanical and electrical damping.
Qtc - the total Q of the speaker in a cabinet. This is a function of Qts (original Q of the speaker), and some of the other T/S parameters (Vas particularly). When a speaker is put in a cabinet, Qtc > Qts, unless the cabinet volume is very very much larger than Vas, in which case Qtc will tend towards Qts.
Fc - the resonant frequency of the speaker cone in the cabinet.
Lets also remember that higher Q means less damped. When there's more than one lot of damping at play, the lowest Q is the biggest contributor. If we're waving a paddle through custard and then add a fan on the other end, the custard is still doing most of the work.
Now, when we drive a speaker from something that is not a very low impedance source, some damping is lost because the braking currents won't be as large (slightly simplistic, but works for now). So, the effective Qes will increase. Given that Qes is usually much smaller than Qms, Qts is dominated by Qes (see custard above). If Qes effectively increases, Qts also effectively increases. We can then see that Qtc will rise due to an increased source impedance.
We can also see something useful drop out - by adding a couple of ohms in series with a bass driver, we can make it more suitable for OB use by raising its effective Qts to something that'll give a flatter response in a dipole system. Note that dipoles have no cabinet, so Qtc doesn't really apply.
What Gnobuddy has said is very similar to this, but the way of describing things above is the most clear to me.
Chris
Peter - you are right about only one thing, you and I live in different universes. In your world Thiele/Small seems to be religion, in mine it is one of several different ways to understand the same electromechanical problem, each one offering different advantages and disadvantages.
Since we cannot break the communication barrier, best we ignore each others posts from now on. Have a nice life.
-Gnobuddy
Chris, thank you for your input, and I agree with your post. As you say, we are saying pretty much the same thing.
T/S converted a mechanical problem into an electrical network, because they were experienced in using analytical tools used in AC network theory. Their solution was groundbreaking, and has remained the foundation used in thousands of loudspeaker designs.
It is not, however, the most intuitive of models to understand, particularly when you go beyond the usual driver-in-a-box speaker system. The Thiele/Small conversion of mechanical parameters into electrical ones makes it harder to see some relationships that are much more obvious when you look at the whole thing as the mechanical system that it is.
-Gnobuddy
This T/S discussion stopped being useful after the first exchange when both sides dug in their heels. Please take your silly urinating contest somewhere offline.
Qes is electrically generated mechanical damping, Qms is mechanical in origin. Qts is the combination - and one side of the argument insisting on calling it 'total mechanical damping' confuses it with Qms for most people who know something about T/S, hence the silly fight that has ensued in this thread.
For those reading this thread thinking they are going to learn something - the effect of Qes dominating Qts is simple to observe with a low output impedance amplifier. With the speaker connected and amp off, tap the cone of a reasonably low resonance low Qes, and high Qms woofer. THen turn the amplifier on and tap the cone. Big difference - the amp has "control" over the woofer. Now insert a potentiometer between amp and woofer and vary the resistance to simulate various levels of output impedance. At some point , with rising resistance, Qms begins to dominate Qts rather than Qes - here you will no longer hear much difference in the amp off vs. amp on scenarios when tapping the cone.
This seems much more intuitive than a bunch of arguing about terminology.
Qes is electrically generated mechanical damping, Qms is mechanical in origin. Qts is the combination - and one side of the argument insisting on calling it 'total mechanical damping' confuses it with Qms for most people who know something about T/S, hence the silly fight that has ensued in this thread.
For those reading this thread thinking they are going to learn something - the effect of Qes dominating Qts is simple to observe with a low output impedance amplifier. With the speaker connected and amp off, tap the cone of a reasonably low resonance low Qes, and high Qms woofer. THen turn the amplifier on and tap the cone. Big difference - the amp has "control" over the woofer. Now insert a potentiometer between amp and woofer and vary the resistance to simulate various levels of output impedance. At some point , with rising resistance, Qms begins to dominate Qts rather than Qes - here you will no longer hear much difference in the amp off vs. amp on scenarios when tapping the cone.
This seems much more intuitive than a bunch of arguing about terminology.
Ron, did you read my last post?
I'm pretty sure we've got it all cleared up.
Now that we've managed to scare OP away, shall we get back on with designing a vented box for a bass guitar speaker?
Chris
To the OP, have you tried putting the driver into a free simulator such as Unibox?
I would consider tuning to the low 40s since that is the lowest note on a 4 string bass, or
do you play a 5 or 6 string? You can adjust the amplitude response at Fb by adjusting
the size of the box to avoid boom.
You could use an adjustable port to try different tunings.
I'd probably build it as two 2X10 boxes that you could stack for portability.
I would consider tuning to the low 40s since that is the lowest note on a 4 string bass, or
do you play a 5 or 6 string? You can adjust the amplitude response at Fb by adjusting
the size of the box to avoid boom.
You could use an adjustable port to try different tunings.
I'd probably build it as two 2X10 boxes that you could stack for portability.
sim has bump - in real life may not look the same - 110 liters for two 10 seems "big"
https://i.imgur.com/1Xx3lx8.jpg
https://i.imgur.com/1Xx3lx8.jpg
I understand the sentiment, but please read back through the thread to see which side of the table all the urine came from. It may take two to tango, but it only one to urinate!
I left the thread to avoid urine splatter, before I had a chance to continue to talk about why it actually helps the understanding to see the problem for what it is: a single damped mechanical system. Seeing it this way may be new to some, but it is almost certainly the way Thiele/Small themselves saw it, and I find it actually helps one to understand the various results that Thiele/Small calculated from their mathematical model. Give me a minute, and I'll give some examples.
Quick recap: a speaker is a mass on a spring, with damping from the mechanical floppy bits. A mechanical damped harmonic oscillator.
Connect the voice coil to an amp with zero (or a reasonably low) output impedance, and you get additional mechanical damping from the current induced in the voice coil. In turn, the total mechanical Q drops, from the value Qms, towards the value Qts.
Now forget the T/S equations and think of the mass-on-a-spring with damping. What happens if you increase damping? Yup, lower Q. Exactly the same result you get from the Thiele/Small model. Perfect agreement between the damped mechanical oscillator approach, and the Thiele/Small electrical filter math.
(Incidentally, adding more damping also causes the resonant frequency to drop a hair, usually too little to matter. This also happens in both T/S and the purely mechanical models.)
There's more: mount the speaker in a sealed box. Thiele/Small says that the resonance frequency rises, and so does the total mechanical Q.
And that's exactly the same conclusion you get when you look at the damped harmonic oscillator approach, too: the sealed air in the box makes the spring stiffer, which produces two effects: a raised resonant frequency, and a raised mechanical Q. Look at the equations for Q and Fo of a damped harmonic oscillator, you'll see what I mean. Once again, perfect agreement between the purely mechanical mass/spring/damping model, and the T/S electrical filter model.
There's still more: mount the speaker in a ported box. The Thiele/Small analysis gets complex enough to require lots of calculator time, or a computer and some suitable software to solve, but it tells you that (a) you now have two resonant frequencies, not just one, (b) one of the resonant frequencies comes from the mass of air in the port "bouncing" on the compliance of air in the box, (c) the outputs from the port and speaker cone add, with phase shifts and amplitudes that vary with frequency, in such a way that the combination acts like a 4th-order high pass filter at low frequencies.
If you notice, that "mass of air in the port, bouncing on springy air in the box" explanation, which usually goes along with most introductory discussions of the Thiele/Small paper, is in fact entirely a mechanical one! And it aligns exactly with what you get when you solve the mathematics for two coupled damped harmonic oscillators; one, the speaker, and two, the mass of air in the port bouncing on the "springiness" of the air in the box.
To me, it's entirely clear that Thiele/Small actually saw the problem for what it actually was in the real world (a mechanical one), and then converted it into a fictitious electrical problem, purely because it allowed them to use the mathematics of AC electrical filter theory, which in turn let them solve the original mechanical problem. It was a masterstroke, a very clever way to solve a difficult, and until then, poorly understood problem, using the mathematical tools that were available and practical to use back then, decades ago.
Today, we have easy access to computers, and software that models mechanical models just as easily as electrical filter networks. So we have the luxury of using either Thiele/Small's approach, or simply putting the mechanical system equations into Matlab, or Gnu Octave, or your mathematical software package of choice.
...that sounds like almost exactly what I said in my post #21. I don't think you and I have opposing viewpoints here at all, you know.
And, for the record, I have no reverence in terminology for terminology's sake, however, we all have to use words to get our message across. One person's "words" might be seen as another persons "terminology".
In other words, I don't care what it's called, I just want to understand how it works, but that's usually easier to do if I call a spade a spade, rather than a Furflefligger-maxxtrup.
-Gnobuddy
I don't want to mess with new software right now.. (is enough with the software that I learned at work (Im civilian engineer))
So,... yes, I have a four string bass,... 41 Hz the lowest note.. maybe even lower than this in drop D tuning..
And yes.. I was thinking to do two boxes 2x10"... For this is that I pretend to calculate a 2x10" box..
So, guys. At this time I'm totally stuning with yours info...
Maybe I should change the perspective of this thread. Maybe I should asking to you to directly design a 2x10" box with the drivers that I post earlier. So,.. if you were in my shoes,.. what would you do?
~64.6 L tuned to ~42 Hz with a ~10.16 cm [4" dia.] x 13.10 cm [5.16"] long vent.
GM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.