Just to complete with the sort of info present in the sim, the woofer has got Qms=3.8, Qes=0.2. Foam suspension and a rather big motor.
Not to mention (implied earlier?) the conjugate circuit keeps reactive currents out of the feedback loop within which the low(er) Zo is crafted?
Well, let me explain why your conclusion about the decay is incorrect. You don't have to buy it but here it is. I assume your woofer also has a rising impedance due to voice coil inductance. That said, when you dive the woofer with a current source, or a voltage source with a high series resistance between woofer and amp, you alter (extend) the high frequency cut off of the woofer response because the VC inductance no longer acts as a low pass filter. The extended high frequency response does give a faster rise and a faster initial decay, and of greater magnitude because of the additional high frequency content in the impulse. But the long time decay is governed by the low frequency cut off and Q. The problem is, this decay is low in amplitude and often difficult to see over the noise unless you have a very quiet measurement room.
Below is another simulation of the Q = 0.43 woofer with and without a low pass filter at 2k Hz to mimic the effect of VC inductance. Notice the case without the LP filter shows a sharper rise and initial decay, with greater amplitude and under shoot. But the long time decay tracks identically for both. So, if you believe this make the current drive superior I can accept that. But considering that the woofer will be low pass filtered it is of little consequence, IMO. However, the higher Q of the woofer when current driven will be an issue that would necessarily have to be accounted for, passively or actively, if the bass response is not to exhibit an peak at resonance and significant over hang.
An externally hosted image should be here but it was not working when we last tested it.
John, now you are being rather offensive, this has just revealed your real agenda and it is not nice. And you have not been careful in examing the data, as I shall clearly make the point:
Do you have access to an amplifier with 270 Ohm output impedance?
You have clearly not noted, or just chosen to ignore, the extreme nature of those graphs.
This is a comparison that even you are not able to make, as I doubt that you have a amplifier there with 270 Ohm output impedance.
That FR deviation is one that FEW will ever hear.
I just don't understand why it has to come to this? Where is the cool scientific attitude of carefully examining the evidence, the nature of the evidence and making judicious conclusions from them? This is not what you do.
The deviation with a perfect solid state amplifier (zero Z) and a typical Single Ended Triode (circa 5 Ohm) Is:
Above 400 Hertz, deviation at two points are about 0.25dB.
Below 400 Hertz, deviation 0.3dB max around 220 Hertz.
With some of my designs I have easily achieved LESS than that.
Even if you could hear that, then fine. But my simple point was that FR does not explain the improvement spoken about.
I stand by that statement.
But thank you, at least you trust the accuracy of my graphs, since you feel confident to use them, albeit in a less than fair way.
I will take that as a compliment. 😀
Last edited:
What we are seeing here is just a rivalry between two speaker designers, but the odd thing, the rivalry is ALL ONE WAY!
As somebody who knows a LOT of speaker designers, I have never come across anyone like John, I have not come across another speaker designer with whom I cannot make some connection, even though we may have wildly different ideas on some things, there will always be something we can agree upon. That makes it a great fraternity and one I have enjoyed for decades. Why John has to be so contrarian towards me, I just don't understand, but take solace from his uniqueness.
BTW, John, what's this about smoke? I don't smoke. 😀
Hope that gave you a good laff - and yes, my face is straight, at least the mirror says so. 😉
As somebody who knows a LOT of speaker designers, I have never come across anyone like John, I have not come across another speaker designer with whom I cannot make some connection, even though we may have wildly different ideas on some things, there will always be something we can agree upon. That makes it a great fraternity and one I have enjoyed for decades. Why John has to be so contrarian towards me, I just don't understand, but take solace from his uniqueness.
BTW, John, what's this about smoke? I don't smoke. 😀
Hope that gave you a good laff - and yes, my face is straight, at least the mirror says so. 😉
I believe that the driver 45 is talking about is a full range so there will be no LP.
dave
Hi Joe,
Does it matter that the test amplifier has a 270 ohm output impedance? How close or far off can it be? 330 ohms, 250 ohms?
I do understand to replicate your results exactly, the identical amplifier would have to be used for the remote test. But the testing by outside sources could be different, but still show the validity of your example ... or not.
I don't think we need the personal comments back and forth, but what is needed is some clear questions and answers back and forth. Stick to the science, not wishful thinking.
If people ask clear questions, can a clear answer be provided?
Now, Steve, Joe doesn't need a knight to defend him. If you want to play the role of interpreter, that's fine. But there is no need to add remarks or talk down to other members, some of who have a great deal of knowledge and experience. What I'm hearing from you is mostly an institutional mindset. Try not to play by those rules as they usually fall flat outside of an institution. You aren't being cute either.
-Chris
Does it matter that the test amplifier has a 270 ohm output impedance? How close or far off can it be? 330 ohms, 250 ohms?
I do understand to replicate your results exactly, the identical amplifier would have to be used for the remote test. But the testing by outside sources could be different, but still show the validity of your example ... or not.
I don't think we need the personal comments back and forth, but what is needed is some clear questions and answers back and forth. Stick to the science, not wishful thinking.
If people ask clear questions, can a clear answer be provided?
Now, Steve, Joe doesn't need a knight to defend him. If you want to play the role of interpreter, that's fine. But there is no need to add remarks or talk down to other members, some of who have a great deal of knowledge and experience. What I'm hearing from you is mostly an institutional mindset. Try not to play by those rules as they usually fall flat outside of an institution. You aren't being cute either.
-Chris
Yes, that sounds about right. Everyone here who questions your claims, you deem to have an "agenda" (you word not mine, because I know you will deny having said that). Yet you never examine the technical analysis that these folks provide, you just fall back to the same claims or deny you ever said what you said (it's obviously a problem with social media if you write 2+2=5 and people think that's what you meant) or, worse, ascribe evil intent to those who ask questions. Do you never wonder why so many of your threads devolve to this sort of acrimony? Have you never considered that you might have something to do with it?
So now, I know, you will send me a pointless and unwelcome private message (which I will post here if you do) and the moderators will, correctly, remove this post. I just get so tired of reading things like your last two posts.
PS: Here's a fun exercise for Joe: go back through your own posts and count the number of first person singular pronouns in them. Now go through the posts of those you call the "usual suspects" and do the same. See any difference? Now ask yourself why.
Last edited:
I posted the same experiment weeks ago with the woofer that Earl uses same result.
I have no agenda, Joe, so please don't try to in that on me. I've been out of the speaker design game for several years now and have no intention of reentering it. At 69 I'm more interested in enjoying my other interests, exotic sports cars, tennis and sailing. Anyway, I apologize for any offensive comment. But if I go by what you have posted previously, about amps don't matter, do I now have to accept that you speaker sounds the same whether driven by a single ended triode or a solid state amp?
If you feel your speaker sounds better with comp networks fine. But you simply haven't provided any evidence, measurements or analysis of why it should. Nothing but antidotal evidence and you statements that it does.
Now, you claim you have better clarity with the networks in place, regardless of amplifier. However, if the amp has significant output Z, and I'm talking on the order of 5 ohm, like your triode amp, how can you make that conclusion? It's not an issue of how much the response changes without out Z when the comp networks are in place that is relevant to that argument. You could flatten the speaker Z perfectly and I would agree that the FR would be constant in that case. But to compare how the speaker sounds with and w/o comp networks in place when the amp has 5 ohm output Z is another issue. In such a case I would suspect you would see greater changes in FR and this would render the comparison invalid. I don't have the exact impedance data for the drivers in your speaker but I have developed reasonable approximations to them. And I have further made parametric variations to those impedance models. They certainly point to changes in the FR on the order of 3 dB and more when I drive the system with an source with 5 ohm output Z and remove the comp networks. So, if you will, it's not about how the response varies with Z out when the comps are in place, it's about how the FR changes when Z out deviates from zero with and w/o the comps.
If you feel your speaker sounds better with comp networks fine. But you simply haven't provided any evidence, measurements or analysis of why it should. Nothing but antidotal evidence and you statements that it does.
Now, you claim you have better clarity with the networks in place, regardless of amplifier. However, if the amp has significant output Z, and I'm talking on the order of 5 ohm, like your triode amp, how can you make that conclusion? It's not an issue of how much the response changes without out Z when the comp networks are in place that is relevant to that argument. You could flatten the speaker Z perfectly and I would agree that the FR would be constant in that case. But to compare how the speaker sounds with and w/o comp networks in place when the amp has 5 ohm output Z is another issue. In such a case I would suspect you would see greater changes in FR and this would render the comparison invalid. I don't have the exact impedance data for the drivers in your speaker but I have developed reasonable approximations to them. And I have further made parametric variations to those impedance models. They certainly point to changes in the FR on the order of 3 dB and more when I drive the system with an source with 5 ohm output Z and remove the comp networks. So, if you will, it's not about how the response varies with Z out when the comps are in place, it's about how the FR changes when Z out deviates from zero with and w/o the comps.
Last edited:
Ok, fine. That doesn't change the character of the long time decay of the impulse, i.e. the point I was making. If current drive gives more extended and useful highs from a full range drive I'm ok with that. It doesn't change what is happening around resonance. That was pretty much the point of Pass's paper and his use of RLC shunts to eq the bass response, in other words, adjust the system Q.
Last edited:
In the context that John used, yes, there is. I was graphing something that showed two extremes and he misinterpreted the data into a 'smoke' comment that was rather crude.
The deviation centered around 200-400 Hertz had nothing to do with the issue, if the Ushers had a sealed alignment, even that could have been fixed too, it could also have been fixed if I had opted for a lower and even flatter impedance. But John used that as an excuse to have a go at me. So I stand by my statement, that the difference heard is not due to FR, as the extreme sample was a worst case scenario (270 Ohm) that most, if any, other than they come here and I would gladly demonstrate. I should not be unreasonably blamed for being brutally honest as I presented the worst case scenario and then that data being misinterpreted as normal everyday data.
John is being mischievous, because both he and I know that the final response of any speaker system is ultimately governed by the principle of a voltage divider, so he has no excuse that he didn't get the data right. A voltage divider has two legs, the deviation is the sum difference of those two legs. He knows that.
For the record, I did a less extreme comparison of the Usher below, this time amp with 0.01 Ohm versus typical Single Ended Triode, SET, about 5 Ohm. Now the difference that John honed in on too quickly is rather smaller.
See what I mean?
Attachments
I am your agenda, or you would not try to pull that one on me.
See my post a few seconds ago and that post of yours is now seen for the 'smoke' it was. Two graphs, one extreme as you would have known, and the other was the real world example.
Judge me correctly and I will thank you, judge me unfairly and I will not lie down and have you trash my reputation like that.
Crude and plain wrong!
Did I ever claim that "FR curves with differences of a dB or so in the 200 to 400 Hz octave" are not audible?
NEVER!
I know from which the direction of the 'smoke' is coming from - may I ask that you desist.
Please!
Thanks understand those two traces both has compensation network included, can you show one plot with 0.01 ohm amp with and without compensation network, and another plot with 5 ohm SET amp with and without compensation network.
Last edited:
It just complete misses the whole point.
The new crossover LOCKS IN both mechanical filters and electrical/crossover filter...
... even at 270 Ohm!
That is what the subject was - then the unhealthier aspects of 'social media' contaminates and distorts it.
Like I said earlier, I am not sure if I will ever get used to that. Actually, I hope that I don't - I want to be a sensitive and caring soul. I don't want to become like that.
At least from an engineering point of view, something here has been accomplished: Both mechanical and electrical filters can be made to avoid modulating the FR. That has to be worth something and perhaps asking others to consider using it?
We also have a method/technique for cancelling out DCR in LF inductors. That too has to be of some practical value for somebody.
And finally, I stand by what I have said, that there is an audible improvement to the speaker, one that is related to the speaker directly and can be heard with any decent amplifier. The 'signature' is very different from that of the amplifier, enough people have heard that, to the chagrin of some.
I have offered a design in the spirit of DIY, nobody ever said that in DIY you have to give people money to try it out. IF I was forcing somebody to do it, then fair enough, I should pay. But other than that, I have played by the rules and I am at least pleased. 😀
The new crossover LOCKS IN both mechanical filters and electrical/crossover filter...
... even at 270 Ohm!
That is what the subject was - then the unhealthier aspects of 'social media' contaminates and distorts it.
Like I said earlier, I am not sure if I will ever get used to that. Actually, I hope that I don't - I want to be a sensitive and caring soul. I don't want to become like that.
At least from an engineering point of view, something here has been accomplished: Both mechanical and electrical filters can be made to avoid modulating the FR. That has to be worth something and perhaps asking others to consider using it?
We also have a method/technique for cancelling out DCR in LF inductors. That too has to be of some practical value for somebody.
And finally, I stand by what I have said, that there is an audible improvement to the speaker, one that is related to the speaker directly and can be heard with any decent amplifier. The 'signature' is very different from that of the amplifier, enough people have heard that, to the chagrin of some.
I have offered a design in the spirit of DIY, nobody ever said that in DIY you have to give people money to try it out. IF I was forcing somebody to do it, then fair enough, I should pay. But other than that, I have played by the rules and I am at least pleased. 😀
Yes indeed.
I think I have already done that some way back, but I will redo the graphs, see attachment below.
Green 0.01 Ohm - same with or without compensation.
Red is 5 Ohm (a la SET) - no compensation.
Blue is 270 Ohm - current send - no compensation.
Note that there has to be applied an offset value to the other two plots to match Green 0.01 Ohm - +4dB for Red and +32dB for Blue. That way the comparison is a whole lot easier.
Attachments
Last edited:
Okay do i get right conclusion then that for new XO one can do subjective listening test on 0.01 ohm amp with and without compensation network in search for improvement by compensation network because FR response is exactly same with and without compensation network, but with 5 ohm amp this is not possible because FR will change with and without network and in fact compensation network has to be there for 5 ohm amp to have same FR as 0.01 ohm amp.
That's a funny plot. The 5 ohms Zout should lower SPL output by some 4dB with a 8 ohms load if noting else than a voltage divider. So how can they be exactly on top of each other? Or did you normalise the output to be the same?
Jan
- Status
- Not open for further replies.