I have assumed this. Excellent idea and engineering and all sincere appreciations from me.
Especially if the paint it is what I guessed, and if the coating is indeed the simple urethane spray...
It is Genially therefore simple.I understand: source must contain details. The reproduction chain must maintain the details levels.
On TOPIC: did you tested your custom drivers with Joe's transamp too? Or only with the "magic knob" solution?
By any chance, could you be kind and ask customers living Karlsruhe or 50miles radius if they could demonstrate it to me, if possible at all? IF they have a current source.. even better and more interesting for me!
Ionmw
Last edited:
I've removed OT posts that attempted to re-open a closed thread. Don't do that. No further warning.What is speaker's DDR? Care for a definition? Never heard of it. R="resolution", right?
I come from another industry, found time for extreme audio hobby just now, so please be a bit more patient with me.
Ionmw
Especially if the paint it is what I guessed, and if the coating is indeed the simple urethane spray...
Paint is high solid content railroad model paint. Conformal coating is gloss acrylic. (more infoin the EnABL threads). Pre-treatment, if needed, includes PVA (ModPodge), ZIG 2-way glue, and damar.
We haven't built Joe's amp, which is essentially a subset of the Variable TransAmp.
Closest would be about 180 km as the crow flies in Heuchelheim (at least from info at the time of shipping)
dave
Downward Dynamic Range -- a term coined by Joe's partner Allen Wright (RIP) to describe how well a device is able toreproduce very small detail/information even in the precence of a much larger signal.
dave
I wonder if EnABL affects the higher modes of cone behaviour. Below a certain frequency the cone moves as a piston in unison with the Voice Coil. Above a certain frequency related to each diameter, typically an 8" driver is pistonic up to 1KHz, a 4" up to 2KHz and 2" up to 4KHz and 1" (typical dome tweeter) 8KHz. These numbers are only a guide. If there is to be output above what Ted Jordan called the "knee" the cone needs to become non-rigid. If the cone was to be totally rigid, then the response would roll off at 6dB (first order) above the knee. I seem to recall Roy Allison did this with a rigid glass cone and it was effectively part of the Crossover handing over to the Tweeter.
It is clear that the cone in our usual cone drivers need to flex in a very controlled manner. The way it flexes are described as higher modes of cone behaviour. But effectively what is accomplished is that the radiating area diminishes as frequency rises, in effect some parts of the cone becomes (in theory) virtually stationary relative to other parts of the cone, so that parts of the cone is "+" and other parts "-". Flexing could be radial or concentric or a combination - I am speculating. This is not to be confused with so-called cone break-up which is basically uncontrolled or undesirable behaviour and resonances that store energy at discrete frequencies causing outright distortion.
I do believe that this is an area of which there is yet a complete understanding - not totally in the dark, but that the air close to the cone itself can have bunches of localised vortexes and stuff like that - I mean, it is only in relative recent times that aerodynamic engineers realised that they could reduce large amounts of vortex drag by lifting or adding appendages at the ends of wings that we see in all modern jetliners. The less disturbed air brought benefits - and also reducing fuel costs.
So while this is speculation, it is not without some foundation.
It is clear that the cone in our usual cone drivers need to flex in a very controlled manner. The way it flexes are described as higher modes of cone behaviour. But effectively what is accomplished is that the radiating area diminishes as frequency rises, in effect some parts of the cone becomes (in theory) virtually stationary relative to other parts of the cone, so that parts of the cone is "+" and other parts "-". Flexing could be radial or concentric or a combination - I am speculating. This is not to be confused with so-called cone break-up which is basically uncontrolled or undesirable behaviour and resonances that store energy at discrete frequencies causing outright distortion.
I do believe that this is an area of which there is yet a complete understanding - not totally in the dark, but that the air close to the cone itself can have bunches of localised vortexes and stuff like that - I mean, it is only in relative recent times that aerodynamic engineers realised that they could reduce large amounts of vortex drag by lifting or adding appendages at the ends of wings that we see in all modern jetliners. The less disturbed air brought benefits - and also reducing fuel costs.
So while this is speculation, it is not without some foundation.
Yes Joe, the holy grail of a full range driver is that the effective radiating radius decreases as the frequency increases, the outer parts remaining fixed (at those frequencies) and the inside radiating all the energy.
The best units are getting better but there is still a long ways to go.
dave
The best units are getting better but there is still a long ways to go.
dave
Not just full range I would think.
But is the benefit of EnABL acoustic or mechanical? I suspect the former - that it may actually have a 'quietening' effect and hence lower noise floor? This would certainly help explaining improved DDR and spatial/image improvement. Being full range already means that they avoid the phase noise of using crossovers - as I do crossovers that have a low noise floor, this also results in better DDR and image/3D scale etc. Which means of course that EnABLE can improve speakers of the type I do as well, not just full range, right?
It's about noise IMO.
Dave, as you maybe already expect, I would ask if you measured the density of your paint as-dried.
About DDR, is translated in "my language" as sensitivity and purity. My colleagues would call it maybe lownoise&precision linear AMP (or whatever). So I agree with DDR in principle, if all accept this is a system-wide property, spanning all components and also that it is a subjective matter: spanning capacity of focus, day to day or hour to hour one's mood, not last the kind of music and kind of auditioning (dance, talk, casual or "serious listening"). Etc., etc., etc. Excluding the subjective metrics then it will help if all realize that it is also hard enough though for hardware (electronics) and producing/reproducing to achieve over 120dB S/N and dynamics (both voltage defined). I come from a world of 20% bandwidth, 99.99% precision, -165dBm@100KHz and about 90db dynamics: I know how hard are these specs, especially considering a 3 decade bandwidth, audio signal being characterized by amplitude, phase, frequency and time information too. Comparing resources and high end audio gear selling prices, I very much doubt that for hardware and for production/reproduction steps the high-end audio will ever go better than 50% precision, -90dBm@10KHz and 70dB dynamics (both voltage, divide by 2 for SPL), top values. Even then it looks like miracles, not understood by 90% of makers, or perhaps 99.99% of customers.
Ok, now if you can imagine that we have a perfect system, with exception of the drivers. We know these are 5% efficiency things, maybe. So now, arise a simple question to me: if you want to reproduce a 1KHz signal, can you estimate how many periods of -70dB incoming signal are needed until you gather sufficient energy to drive the local mode in the membrane in synch (with all dampers)? Realize the point? For 3 decades bandwidth it comes to distorsions, much like a LP filter (Trebles) or THD effects (High Mids) or TID (Low mids), no other chance. I have excluded the Low and mid-Bass because of Loudness.
Q1: in your opinion the Joe's Transamp would help (or your variable TIA forexample)? What parameters or details are most most important; what settings are best? Best PS maybe? Best cabling? ... Or the choice of OP-AMP? What parameters are of interest for DDR? Best Input stages? Best Output? Bias currents? Detail if necessary, other points, if not mentioned.
Q2: in your opinion your full ranges obtain a better DDR than separated and specialized Mid/Trebles drivers? Why, if no bass is needed for DDR, use a full ranger instead of dedicated Mid+Treble? What parameters make the difference in your opinion? Detail if necessary, other points, if not mentioned.
The gain in DDR is exactly the feature that drives me, or most of us here I guess, to build Joe's amp. But I think it is due to understand what AMP parameters, focused on Joe's Tranamp solution, are more important for DDR, so we can do educated experiments before and after its realization.
I would much appreciate an "engineer focused" answer.
Thanks
Ionmw
About DDR, is translated in "my language" as sensitivity and purity. My colleagues would call it maybe lownoise&precision linear AMP (or whatever). So I agree with DDR in principle, if all accept this is a system-wide property, spanning all components and also that it is a subjective matter: spanning capacity of focus, day to day or hour to hour one's mood, not last the kind of music and kind of auditioning (dance, talk, casual or "serious listening"). Etc., etc., etc. Excluding the subjective metrics then it will help if all realize that it is also hard enough though for hardware (electronics) and producing/reproducing to achieve over 120dB S/N and dynamics (both voltage defined). I come from a world of 20% bandwidth, 99.99% precision, -165dBm@100KHz and about 90db dynamics: I know how hard are these specs, especially considering a 3 decade bandwidth, audio signal being characterized by amplitude, phase, frequency and time information too. Comparing resources and high end audio gear selling prices, I very much doubt that for hardware and for production/reproduction steps the high-end audio will ever go better than 50% precision, -90dBm@10KHz and 70dB dynamics (both voltage, divide by 2 for SPL), top values. Even then it looks like miracles, not understood by 90% of makers, or perhaps 99.99% of customers.
Ok, now if you can imagine that we have a perfect system, with exception of the drivers. We know these are 5% efficiency things, maybe. So now, arise a simple question to me: if you want to reproduce a 1KHz signal, can you estimate how many periods of -70dB incoming signal are needed until you gather sufficient energy to drive the local mode in the membrane in synch (with all dampers)? Realize the point? For 3 decades bandwidth it comes to distorsions, much like a LP filter (Trebles) or THD effects (High Mids) or TID (Low mids), no other chance. I have excluded the Low and mid-Bass because of Loudness.
Q1: in your opinion the Joe's Transamp would help (or your variable TIA forexample)? What parameters or details are most most important; what settings are best? Best PS maybe? Best cabling? ... Or the choice of OP-AMP? What parameters are of interest for DDR? Best Input stages? Best Output? Bias currents? Detail if necessary, other points, if not mentioned.
Q2: in your opinion your full ranges obtain a better DDR than separated and specialized Mid/Trebles drivers? Why, if no bass is needed for DDR, use a full ranger instead of dedicated Mid+Treble? What parameters make the difference in your opinion? Detail if necessary, other points, if not mentioned.
The gain in DDR is exactly the feature that drives me, or most of us here I guess, to build Joe's amp. But I think it is due to understand what AMP parameters, focused on Joe's Tranamp solution, are more important for DDR, so we can do educated experiments before and after its realization.
I would much appreciate an "engineer focused" answer.
Thanks
Ionmw
Last edited:
I don't have enuff experience to say. I will note that the discreet transamp Daniel built outperformed the chip amp version.
My concentration is on FRs and FASTs (no XOs where driver c-c is > ¼ wavelength -- XOs are also simple & 1st order). Our best speakers so far have been FASTs, but some of the FRs come close and cost a lot less.
EnABL was originally applied to multi-ways. XOs are such a problem that they still often are the limit on how well a multiway can perform.
DDR was 1st used by Allen to help describe what he was working toward with his electronics.
dave
To wash away, again, the distrust of others capabilities: I know at least two companies (to focus only in Germany) that are doing combined FEM+Flow analysis, even including chemistry of molecules in vortexes. I also know many more scientists who order and can critically analyze their results. Most of their jobs are with aero industry, but these FEM+Flow guys will accept tasks from other industries for a price about 200K per year/study.
In Japan is the Fostex who does this analysis, day to day business. I am sure many others do. Maybe only Eminence, NOT , but it might be just a joke
What I want to say is that they (more than 1 person in the world, for sure) would analyze the effect of Dave's smart paints lines and re-coating very frugal, from 9am to 12 and then go for lunch warmed up for Boeing... or Airbus... or others.
I personally leave some place for due credit to many poor genius-engineers or genius-scientists.
Ionmw
Last edited:
Thanks!
I still wish to see a schematic of your variable transamp. Let me know when is published. Because DDR topic is so much about noise, it is pointless to refer to results if we do not see the schematic.
...
EnABL on multi-ways give problems with XOs: also with Transamp or with Voltage? Not enough clear to me, so I must ask. I intend to mildly re-coat my 3-way (bass and mid) drivers for Joe's transamp.
Ionmw
The amp is Daniel's not mine… i poked him in the hope he would post it. Other than the variable part it differs little from Joe's, which Daniel vetted before Joe posted it.
That is an over-simplification. It has been shown that humans can hear into the noise.
While the actual circuit is important, one cannot discount the importance of implementation.
That is not what i meant. XOs are a problem no matter the system. There are ways to minimize their problems
dave
Thanks!
Of course humans can discern into "noise", mind the limits of the problem though. Example: X and Y can have a discussion in a disco and this is proved ever day. So, some people (most of engineers, I agree) should define noise in a smarter way.
Allow me to push you for more details on what problems and what solution regarding DDR related problems with XOs, with coatings. Why XOs if you test them on Transamps? Did you have active XOs or passive?
Last but not least. By how much Joe's Thansamp was predated? Difference improved eventually for other not-so-perfect speakers?
Ionmw
Of course humans can discern into "noise", mind the limits of the problem though. Example: X and Y can have a discussion in a disco and this is proved ever day. So, some people (most of engineers, I agree) should define noise in a smarter way.
Allow me to push you for more details on what problems and what solution regarding DDR related problems with XOs, with coatings. Why XOs if you test them on Transamps? Did you have active XOs or passive?
Last but not least. By how much Joe's Thansamp was predated? Difference improved eventually for other not-so-perfect speakers?
Ionmw
Last edited:
Treatment that increases the ability of a driver to produce lower level information will give more room for an XO to remove it.
XOs are a problem with or without treatment. Often placed where the driver-driver spacing is greater than a ¼ wavelength the separation of the drivers cause an issue. XOs of higher order screw up the phase. That is one of the areas where FRs often gain (some FRs have a mechanical XO). XOs often create very uneven impedance.
We don't have a tran-amp here. I dohave Daniel building me one, but he is pushing the tech and it is stalled.
More often than not a PLLXO, but i am currently using an active XO (an old Ace Audio) and we have done some series XOs. The passive XOs are much harder to get right and more expensive.
The 1st TransAmps were probably built in the 20s or 30s.
Joe's amp is likely not that different than the one published at Elliot Sound or the chip amp version in ESA's book.
dave
Please, help with this design. The attached image is a modifocation for work in current of the scheme in this page, can be with a TDA2002 or a TDA2003:
Amplificador de audio LA4425, y similitud con TDA2002
Have a old German radio with the two outputs burned, no spares. Mi intention is use two of this low power amps in current for have the sonud of valves.
Also, I have in mind build a small 'boombox' with two low power amps and two cheap fullrange speakers for connect a Sansa Fuze, so I have a 'radio reproductor' with sound of valves, but whitout the inconvenients of the valves. Other people can like made some as this.
Please, is correct, any suggestions?
Amplificador de audio LA4425, y similitud con TDA2002
Have a old German radio with the two outputs burned, no spares. Mi intention is use two of this low power amps in current for have the sonud of valves.
Also, I have in mind build a small 'boombox' with two low power amps and two cheap fullrange speakers for connect a Sansa Fuze, so I have a 'radio reproductor' with sound of valves, but whitout the inconvenients of the valves. Other people can like made some as this.
Please, is correct, any suggestions?
Attachments
Hi
I would like to build similar simple one
but unfortunately had no valid model for TDA2003.
I am too weak in LTSpice to make the model by myself and nobody would like to create the model, mod, sub file instead of me
http://www.diyaudio.com/forums/soft...mulate-transcond-tda2003-amp.html#post4426984