I know voice coil inductance is a point of interest on this thread. When I saw this dual voice coil vs only one of the coils response plot, I noticed the top end rolling off. Posting in case it may be interesting to consider.
https://www.diyaudio.com/community/threads/dats-v3-measurement.421254/post-7874896
https://www.diyaudio.com/community/threads/dats-v3-measurement.421254/post-7874896
I saw that ad: for reasonable, substitute excellent. Unfortunately, we don't need a large pair of speakers but it looks like a great deal.
Not sure if the XO parts are included?
Geoff
My first post in this thread after a few pm's regarding some designs. And after many hours of reading
I'm going to build the Elsinore ULD, after owning a Vifa/ScanSpeak tmm for the last 16 years or so.
The waveguides from Joe are on their way from Australia to The Netherlands and just placed the order at Purifi!
So now i have a couple of months to gather wood/MDF, other parts etc, because production is fully booked at Purifi till march
I'm going to build the Elsinore ULD, after owning a Vifa/ScanSpeak tmm for the last 16 years or so.
The waveguides from Joe are on their way from Australia to The Netherlands and just placed the order at Purifi!
So now i have a couple of months to gather wood/MDF, other parts etc, because production is fully booked at Purifi till march
As a word of encouragement during the wait on drivers from Purify, I am very much enjoying my ULD build, and it was well worth the effort. It is a great strategy to plan your build, you can get moving on a lot ahead of time, assembling crossovers, securing your materials, even cutting out the panels, but always wait on the drivers before routing out the driver holes and rebates and only assemble afterwards. I actually used this project as an excuse to upgrade my shop, which meant trial runs to see how the new tools worked and achieve the precision I was striving for. And I cannot recommend enough getting plenty of clamps, you will need more than you can imagine.
On another note, I am finding these speakers really like a good Class A MOSFET amp. I have been a tube amp guy for ages, both PP and big (211 Class A2) SETs. But after completing a dual mono DIY PASS FW F4 driven by a 12SN7 Aikido tube linestage, I am finding it is a special combination.
David
On another note, I am finding these speakers really like a good Class A MOSFET amp. I have been a tube amp guy for ages, both PP and big (211 Class A2) SETs. But after completing a dual mono DIY PASS FW F4 driven by a 12SN7 Aikido tube linestage, I am finding it is a special combination.
David
David,
Good to hear from you.
The ULD build is a decent excuse to get the wood working shop upgraded. 🙂
I am not surprised that you are finding the Class Amplifier + Elsinore ULD combination to your liking.
Good to hear from you.
The ULD build is a decent excuse to get the wood working shop upgraded. 🙂
I am not surprised that you are finding the Class Amplifier + Elsinore ULD combination to your liking.
Hi David
Congratulations. The "ULD" version is the most easy to drive speaker that I have done. I agree that a good class A amplifier should go well with them.
Hope your new work shop gets used for other good things. 😊
Congratulations. The "ULD" version is the most easy to drive speaker that I have done. I agree that a good class A amplifier should go well with them.
Hope your new work shop gets used for other good things. 😊
Noted. But to be honest, not had much experience with dual voice coil drivers. If that is the behaviour, then there has to be some logical explanation, that we can be sure of. Do you have an idea? In the instance of a single coil, was the other just left unconnected? The last picture he posted seems to indicate that?
Come to think of it I'd like to have seen it out of the cabinet, and maybe there's not much to see here. In any case leaving the unused winding open is the way to make it like it isn't there.
Looks like purify is finally releasing their tweeter, however so far I can't actually find any real info on it apart from the two blog posts and the fact that they've shipped some initial samples.
So I’ve been doing some amp upgrades lately and decided to purchase some Dueland silver 0.01uf bypass caps for my TCMs as this was a bit more economical than the miflex’s and they could be used elsewhere in the future (especially if the purify tweeter can be used in a future revision).
So I did some quick listening after installing them and was immediately surprised at how clear cymbals and other high frequency instruments came through. One particular song I’ve been listening to repeatedly through the upgrades had a background cymbal sound that jumped right out mid song that I had never heard at all before.. I’ve gotta keep listening but so far this was quite worthwhile with respect to high frequency resolution..
These are the ones I used:
https://partsconnexion.com/product_images/downloads/DUELUND-JDM-AG-CAPS.pdf
Yes, I can believe that. But we are comparing apples and oranges.
Using 0,01uF in amplifiers can make quite a difference. This is because the impedance of the circuit is much higher than speakers.
But using this small value in parallel with 1.8uF feeding a low impedance (all speakers are low impedance and high current), I can only see that it would make a very small difference. But 0.1uF as by-pass in speakers, that may make a difference.
Rule: The lower the impedance, the by-pass value must be higher. Take a look at the termination impedance that the capacitor sees. That is how I see it, if others disagree, I will let them make their case.
Hey - I should’ve been clearer, my apologies. I bypassed the 1.8uf TCMs in the crossovers of my ULD elsinores and saw this improvement - I had been doing other work which prompted me to revisit all of this.
OK on that. If it works, then so be it, but it is surprising that it does. If somebody can explain it, I would be pleased to know.
Nominally the tweeter plus series resistor is roughly 10 Ohm.
A 0.01uF capacitor has a 'reactance' of 5300 Ohm at the crossover frequency 3KHz. The termination impedance is circa 10 Ohm. The 'reactance' of the 1.8uF is 29 Ohm. That is surprisingly high, but it is achieved by the Waveguide having a gain of nearly +6dB @ 3Khz and hence we can make the capacitor here about 1/3rd value. Yes, normally we would need 5-6uF here and this would be a huge capacitor in the Miflex or Duelund ranges. So the Waveguide means that the Miflex 1.8uF/250V cap can actually just fit the crossover PCB. Yes, the Waveguide also saves us a LOT of money.
If anybody would dispute the above numbers, please feel free to do so. But if it makes a difference to the sound, what could be the reason? Just asking. In speakers I would go for 0.1uF as by-pass value and that making a difference... maybe I can see that.
Nominally the tweeter plus series resistor is roughly 10 Ohm.
A 0.01uF capacitor has a 'reactance' of 5300 Ohm at the crossover frequency 3KHz. The termination impedance is circa 10 Ohm. The 'reactance' of the 1.8uF is 29 Ohm. That is surprisingly high, but it is achieved by the Waveguide having a gain of nearly +6dB @ 3Khz and hence we can make the capacitor here about 1/3rd value. Yes, normally we would need 5-6uF here and this would be a huge capacitor in the Miflex or Duelund ranges. So the Waveguide means that the Miflex 1.8uF/250V cap can actually just fit the crossover PCB. Yes, the Waveguide also saves us a LOT of money.
If anybody would dispute the above numbers, please feel free to do so. But if it makes a difference to the sound, what could be the reason? Just asking. In speakers I would go for 0.1uF as by-pass value and that making a difference... maybe I can see that.
So I’ve been recently thinking about analog audio waveforms.. obviously you have audible signals to 20khz.. BUT you have multiple instruments’ waveforms superimposed. So those transitions between the component waveforms may require a higher frequency bandpass?
I at one point was doing mods on a CRT projector with analog video signals in the ~20-100mhz range from recollection.. and I remember at that point reading that the square waves required to render an individual pixels’ signal required the third order harmonic to be passed to form the square wave shoulder.. so you needed to reliably pass at least 3 times the base signal frequency in order to render the square wave. In fact to closer approximate a true square wave you needed to pass higher order harmonics too - third order was just to cause a semi sharp shoulder.
Is it not possible that some of the difference in audibly detectable differences between caps is actually due to band pass outside of the 20khz range as that is the area that form shoulders/edges between different high frequency components of a signal? Forgive me if this is super off base.
I at one point was doing mods on a CRT projector with analog video signals in the ~20-100mhz range from recollection.. and I remember at that point reading that the square waves required to render an individual pixels’ signal required the third order harmonic to be passed to form the square wave shoulder.. so you needed to reliably pass at least 3 times the base signal frequency in order to render the square wave. In fact to closer approximate a true square wave you needed to pass higher order harmonics too - third order was just to cause a semi sharp shoulder.
Is it not possible that some of the difference in audibly detectable differences between caps is actually due to band pass outside of the 20khz range as that is the area that form shoulders/edges between different high frequency components of a signal? Forgive me if this is super off base.
I am thinking... the reactance of a 0.01uF cap (10nF) is 795 Ohm and tweeter is 8 Ohm. In a tube amp the termination impedance into output tubes is around 100,000 Ohm, that I can see that making a difference from 150 Hertz and upwards. Here even a quite small capacitor can have significant payoff.
I am not trying to throw cold water on your theory, but if I was to use a bypass cap in speakers, I wouldn't use less than 0.1uF and maybe even a bit higher.
BTW, it is not hard to find capacitance calculators online: https://www.omnicalculator.com/physics/capacitive-reactance
You can even do it easily with a calculator: 1/(6.283*freq*uF) - make sure the 'freq' is in megaHerz.
I am not trying to throw cold water on your theory, but if I was to use a bypass cap in speakers, I wouldn't use less than 0.1uF and maybe even a bit higher.
BTW, it is not hard to find capacitance calculators online: https://www.omnicalculator.com/physics/capacitive-reactance
You can even do it easily with a calculator: 1/(6.283*freq*uF) - make sure the 'freq' is in megaHerz.
When you mix two waveforms? It doesn't. What could fit before will, and what couldn't will still be cut off.
So if you had two 12khz sine wave sources (representing two instruments etc being played at once) offset by 90 degrees on the what would your bandwidth requirement be? If you then threw in a third 15khz low volume signal from another instrument overlaid as well? With three together on the analog waveform wouldn’t you have two sine waves in closer proximity to each other than each component alone plus another smaller volume signal superimposed? Maybe I’m misunderstanding the base waveform but I would have to think the above considerations are why digital music sampling rates in high resolution audio are getting higher - to maintain the complexity of the waveform.