??? Not understanding your question. V = I * Z, Z is fixed. Plug in which ever is constant of I and V and you get the other.
Yesterday I have made a mistake at P.26/Post#251.
Allow me to bring self corrections:
Actually it is the ratio Fs/Qts which has meaning of bandwidth (of resonance). This makes my second statement False.
It turns that Qts/Fs has the physical meaning of inverse bandwidth, with following common-sense conclusions:
- since bandwidth is proportional to loss, and the later has terms that are less dependent on manufacture but on material properties (tight indirect QC via supliers) and ME concept design (tight internal RnD control previous to product release) then manufacture wise for a serious product it makes sense to expect to see Bandwidth unchanged, aka. constant Qts/Fs or its inverse, as common feature.
- Small bandwidth (large inverse Fs/Qts) indeed calls for Horn designs, 'cause:
a) they have cut-off HP behaviour, more exactly exponential decay, which is the most very effective way to isolate a sharp resonance (assuming a standard X-over, i.e. no shunt notch conjugate is present) and
b) a small bandwidth exists if under or critical matching, which 50% of times means critical matching aka. best coupling efficiency, which in turn is a needed condition to drive a cone.
- Large bandwidth (small inverse Fs/Qts) indeed matches no-resonant (closed) or low-resonant (BP and BR) enclosures.
With Z-leveling actions you might help efficient motors be used in closed enclosures(for example). This is where you heading out now but TSP and Z*(f) consideration is a must! Have fun wasting your time without them, least to mention confusing those who just started to learn.
With titanic work if considering TSP and Z*(f) you will surely help the amplifier-speaker interface, but the final acoustic result will still be unsure. A horn acting outside resonance might still be much cleaner and purer, but surely it will be easier, simply because it does not have to handle the phase inversion and group delay bunching from so many (aligned) resonances.
If I will see here a solution+TSP+Z*(f)+method and a rigorous analysis language - inclusive definition of terms - I will be much surprised...
Why? Mixing in the same sentence the "Q" of LP or HP filter (not BP, no mention that it means "loss" and proceeding further without quota signs)... with true Q from resonances (like Qes,Qts, notch etc) simply invites to sum their inverses. Test question: is this permitted and if yes, in what conditions and limits?
Allow me to bring self corrections:
Actually it is the ratio Fs/Qts which has meaning of bandwidth (of resonance). This makes my second statement False.
It turns that Qts/Fs has the physical meaning of inverse bandwidth, with following common-sense conclusions:
- since bandwidth is proportional to loss, and the later has terms that are less dependent on manufacture but on material properties (tight indirect QC via supliers) and ME concept design (tight internal RnD control previous to product release) then manufacture wise for a serious product it makes sense to expect to see Bandwidth unchanged, aka. constant Qts/Fs or its inverse, as common feature.
- Small bandwidth (large inverse Fs/Qts) indeed calls for Horn designs, 'cause:
a) they have cut-off HP behaviour, more exactly exponential decay, which is the most very effective way to isolate a sharp resonance (assuming a standard X-over, i.e. no shunt notch conjugate is present) and
b) a small bandwidth exists if under or critical matching, which 50% of times means critical matching aka. best coupling efficiency, which in turn is a needed condition to drive a cone.
- Large bandwidth (small inverse Fs/Qts) indeed matches no-resonant (closed) or low-resonant (BP and BR) enclosures.
With Z-leveling actions you might help efficient motors be used in closed enclosures(for example). This is where you heading out now but TSP and Z*(f) consideration is a must! Have fun wasting your time without them, least to mention confusing those who just started to learn.
With titanic work if considering TSP and Z*(f) you will surely help the amplifier-speaker interface, but the final acoustic result will still be unsure. A horn acting outside resonance might still be much cleaner and purer, but surely it will be easier, simply because it does not have to handle the phase inversion and group delay bunching from so many (aligned) resonances.
If I will see here a solution+TSP+Z*(f)+method and a rigorous analysis language - inclusive definition of terms - I will be much surprised...
Why? Mixing in the same sentence the "Q" of LP or HP filter (not BP, no mention that it means "loss" and proceeding further without quota signs)... with true Q from resonances (like Qes,Qts, notch etc) simply invites to sum their inverses. Test question: is this permitted and if yes, in what conditions and limits?
Other than nitpicking, ditto here.
In fact, it dispels several myths.
Indeed there are also elements I want to return to and how it relates to the alternative crossover of the Usher S520.
That means I am likely to come back and quote some of what John has said and see them in the context of the way that I do things.
Indeed it is the V/I function I want to get to, but I implore, I need to be allowed to, step by step, describe the thinking or 'philosophy' behind the crossover. Here I define philosophy as the fundamental thinking behind the approach used.
Now hold that thought, because here we have something I want to return to. But I need to build a case and eventually come to elements that John just mentioned, and basically comes down to that V = I*Z function. While I may have my own deas, when we get to that point, I will not be averse to ask John questions and wanting to know his thoughts.
But just quickly, two points.
One, I want to correct the notion that I am in the 'current-drive' faction. That is not entirely true. This is where Esa Merilainen and I have a different, but not necessarily incompatible ways of looking at things. I am far more interested in voltage drive and getting voltage drive right.
Secondly. I want a more hand-on discussion, an engineer's way of looking for solutions (read stuff about Jørn Utzon, although architect, came up with engineering solutions that amaze) and while there is a definite role for the maths (there has to be), we need all of us to explain what we see the maths saying, not just the numbers.
Maybe just one more.
I have been described in terms that I am not comfortable, so I am only going to say it once. I am actually quite conservative, although many seems to have gathered a different view. If you ever spoke to me in person, you would get an entirely different impression of me.
I think if we view the best we have in each other, and bring our joint ability to the discussion, this could be a really interesting topic.
So John, I do hope you stick around.
When Vout, the V gets converted to I, the current, by Z. The current is still the end product, as I see it.
On this I have a number of questions, but I want to keep my powder dry for now.
Like, how accurate is...? No, will leave for later.
Joe is there any chance that you will stop telling us what you are going to talk about, and actually say something that isn't about you?
True, each tile of the Sidney Opera House was different and cast in situ/pre-fab in the construction yard.
I can see where some of the problems arise. Maths are a primary way of looking at a problem and it is not just about numbers.
Quite. An engineer will use the maths to describe and solve the problem. Not sure how one would describe a purely empirical approach to the problem without risking being derogatory.
There's a difference between having a PhD in applied math and not having a basic knowledge of complex algebra. Interesting that the Philbrick house technical journal was called "The Lightning Empiricist" while most of those involved were quite facile in math. My impression is that we are not trying to account for poorly understood or second and third order issues here but a problem that is well served by the standard first order electro-mecanical modeling and mathematical analysis..
Last edited:
And I am guilty of completely forgetting derivations but remembering the end equation that I can use to get something done!
I know this is DIY, but... considering the educational value... is here any possibility for us to vote a minimal set of proper technical conditions for this thread to stay open? Proposals?
My questions, or first proposals:
1. Title tells that main purpose for this X-over is to work well for both CS and VS, in a "compatible" manner. Right or misunderstanding? If right, then is there any other merit to justify technical discussions beyond this point? If Yes, then proper technical statements and Proofs are expected without delay.Maybe restart in a better (="proper") and complete (="full") manner?
2. Just to enable all of us (owners or not of Usher S520) to verify the quality and potential of various statements before engaging any DIY effort, then the TS+Z*(f) characteristic of drivers are expected without delay. These should have been stated in Page1, in my opinion.
Joe, I am convinced all this middle-way is very reasonable and doable to you, of course.We are all waiting...but not for Godot. If you cannot deliver now, for various reasons, it is OK. You can close this thread for now, and reopen the topic when you are ready.
All the best,
Ionmw
My questions, or first proposals:
1. Title tells that main purpose for this X-over is to work well for both CS and VS, in a "compatible" manner. Right or misunderstanding? If right, then is there any other merit to justify technical discussions beyond this point? If Yes, then proper technical statements and Proofs are expected without delay.Maybe restart in a better (="proper") and complete (="full") manner?
2. Just to enable all of us (owners or not of Usher S520) to verify the quality and potential of various statements before engaging any DIY effort, then the TS+Z*(f) characteristic of drivers are expected without delay. These should have been stated in Page1, in my opinion.
Joe, I am convinced all this middle-way is very reasonable and doable to you, of course.We are all waiting...but not for Godot. If you cannot deliver now, for various reasons, it is OK. You can close this thread for now, and reopen the topic when you are ready.
All the best,
Ionmw
These speakers, in their characteristics, claim to have constant impedance:
EVA | Davis Acoustics
•Un filtre avec compensation d’impédance, atténuateur à impédance constante,…
I have consulted to Davis Acoustics but I have not received a reply.
Does anyone have information about these speakers, the impedance graphic? Thanks
EVA | Davis Acoustics
•Un filtre avec compensation d’impédance, atténuateur à impédance constante,…
I have consulted to Davis Acoustics but I have not received a reply.
Does anyone have information about these speakers, the impedance graphic? Thanks
Nothing of their website, but the specs of one of their speakers says 4-8 Ohm, so that is hardly constant. Let us know what they say, maybe some of the speakers do? I am kinda hoping they are...?
Everybody who has followed this thread with interest understands that Joe committed something close to audio heresy by suggesting that impedance matters. Really lighten up folks. It's up to you. 😀
I think flattish impedance is a pretty damn good idea. It works better with imperfect amplifiers, and every speaker I build seems to sound better for it. 😎
Whether you get it PERFECT is your call, but better is better IMO. But there are some quite interesting techniques for flat impedance. How nice does this one look? 5" plus 1" tweeter. Not you regular flat baffle crossover. But worth a go IMO. These sort of ideas add interest to my hobby.
I think flattish impedance is a pretty damn good idea. It works better with imperfect amplifiers, and every speaker I build seems to sound better for it. 😎
Whether you get it PERFECT is your call, but better is better IMO. But there are some quite interesting techniques for flat impedance. How nice does this one look? 5" plus 1" tweeter. Not you regular flat baffle crossover. But worth a go IMO. These sort of ideas add interest to my hobby.
Attachments
Well, like anything in this world, generalities don't hold. If an amp is sensitive to impedance phase or impedance amplitude variation, flattening Z may be beneficial, but there are other considerations. For example look at the power delivered into a speaker at the woofer resonance where the impedance may be around 40 ohms and resistive. For 2 V input that's 0.1 Watt, 0.05 amps. Now, flatten the Z so that it is around 6 ohms across the entire frequency range. The same power is delivered to the speaker but the amp must deliver 0.666 Watts, 0.333 amps, in power and a across the load it sees. That is an 8.25 dB increase in power at 6.666 times the current. So, for a given volume level you would need an amp with 8.25 dB greater output into a lower impedance load. Not to mention than now any component of the signal at higher frequencies will be riding that higher current level meaning there is greater possibility to exceed the amps current capability. It's just not a black and white issue.
John, once again we get a lesson in basics about stuff we already know. We wouldn't be doing anything to stress the amplifier, we are experienced enough to know where the dangers are. Every speaker I have designed in fourty years have been amplifier friendly and anybody who knows me knows that is paramount.
We can always find technical reasons why something shouldn't work. There was a time, like thousands of years, when it was crazy to think that iron boats could float, so why would somebody build them? Waste of time. Until they realised there was something called displacement.
This subject only really matters if it makes the speaker sound better !!!
For whatever reasons we may or may not know? But brainstorming now seems frowned upon - even if it leads to nowhere it can still be fun and might lead to something surprising because somebody says something so left field that makes a synaptic connection in somebody's else's head and gets results that was not predictable any other way. It seems social media no longer lends itself to such 'fun' that could be very beneficial. Instead we just seem to get these battles of egos.
Is there something that has been overlooked when we say that flattening the impedance leads to a noticeable increase in clarity and not then having to defend our listening acuity and be told we are imagining it?
That was the topic, that or just getting your hands on a pair of Usher S520, the crossover is reversible if you don't like the result, they are cheap, lot's available - at least I have made the opportunity available.
I would love to have an open discussion about, it would at least be 'fun' to see where it leads.
We can always find technical reasons why something shouldn't work. There was a time, like thousands of years, when it was crazy to think that iron boats could float, so why would somebody build them? Waste of time. Until they realised there was something called displacement.
This subject only really matters if it makes the speaker sound better !!!
For whatever reasons we may or may not know? But brainstorming now seems frowned upon - even if it leads to nowhere it can still be fun and might lead to something surprising because somebody says something so left field that makes a synaptic connection in somebody's else's head and gets results that was not predictable any other way. It seems social media no longer lends itself to such 'fun' that could be very beneficial. Instead we just seem to get these battles of egos.
Is there something that has been overlooked when we say that flattening the impedance leads to a noticeable increase in clarity and not then having to defend our listening acuity and be told we are imagining it?
That was the topic, that or just getting your hands on a pair of Usher S520, the crossover is reversible if you don't like the result, they are cheap, lot's available - at least I have made the opportunity available.
I would love to have an open discussion about, it would at least be 'fun' to see where it leads.
Last edited:
- Status
- Not open for further replies.