Crossover upgrade in a Jeff Bagby Auricle

Jump to Latest
B

Bryan S.

Member
Joined 2022
#1
Hello, I'm new to this forum, although I've read many posts here over the past several months and learned a lot. Having built most of my power cables and interconnects, I decided last fall to build a pair of Jeff Bagby designed Auricles, through Meniscus. At the time I wasn't very knowledgeable about crossover components, and just went with the highest quality offerings Meniscus listed. Since then, and being an obsessive optimizer, I've been reading up on caps and inductors, in order to gain a more intuitive grasp of their functions. Wow, rabbit hole. Now I'm seriously considering swapping out the caps and coils on these speakers. I know I"m just jumping in hard here, but I'll describe the current configuration so anyone who is interested has the info.

The Auricles use RAAL 70-20 tweeters and Satori 6.5" midwoofers. I've built the boxes quite robustly, with a 1.5" baffle and thick damping material on the inside walls of the cabinets, which are front ported. Corssovers are as follows: Tweeter - 6.8mF Mundorf Supreme (no oil) cap, .68mH 20 ga basic wire coil inductor, and a couple of mundorf resistors to pad the level a bit. Woofer uses a Clarity Cap 4.7mF cap, and 3mH 14 ga wire coil. Pretty simple crossovers as these drivers integrate very smoothly without too much fuss. Crossover point is 1800 hz.

I'm wondering first of all, if it would be worth messing with the Satori board, and if so whether to address the inductor, the cap, or both? I'm pretty confident swapping out the parts on the RAAL board will open things up even further, but not sure about the woofer.

Right now the candidates for tweeter caps are VCap ODAM, or either Mundorf Supreme silver/oil or Mundorf Supreme Silver/Gold/Oil (EVO or Classic). The inductor would be a Mundorf copper ribbon. If I mess with the woofer, I'd probably do something similar.

If you ask me what my design goals are, I can't answer in the way I'd like to as I have nothing to compare these speakers to in my system except the last pair which were ELAC UB52s, and the Auricles are in a different league. They sound absolutely fabulous, but I really think I left a lot on the table when I just went with Meniscus' default components. I've seen what something as small as a binding post swap can do, and my system is decent enough to hear the rewards of upgrades, to a point. Right now I'm using a Schiit Aegir in stereo but in the near future I'll either add another and go monoblocks, or swap out for another class A amp altogether. I really think these Auricles are something special, and that it's worth putting the extra shine to them if it will make a difference.

Any ideas would be greatly appreciated, thank you!
 
#2
For the tweeter inductor (assuming it is a shunt component) you could probably upgrade it with a litz inductor with the closest DCR to the stock inductor. For the series tweeter capacitor, you could try going with foil capacitors. Russian polystyrenes are your most cost effective option. For the woofer inductor you may not be able to improve it if it is already an air-core inductor. If it is not, a 10 gauge Solen air-core inductor with a DCR less than or equal to the stock component is what you are after. If the DCR is greater, you will actually be attenuating the bass.

Do you know which version of the Raal 70-20 you are using? The 70-20XR uses an amorphous core transformer which is a premium type material. That could also be an upgrade.
 
Last edited:
#8
Don't put Litzes on tweeters. The capacitance of the multi-strand can actually be detrimental in higher frequencies.
Click to expand...

I don’t think those frequencies are in the audio band, just like the technical benefits of using foil inductors does not exist in the audio band.

See this chart below comparing the Q of various types of inductors, all 14AWG and of the same total inductance:
FA7C3393-8DB8-4216-B1E1-24DABB02E015.jpeg


A lower Q indicates greater losses in the filter response and conversely, a higher Q corresponds to less losses.

For filter target frequencies between 1.6K-40KHz, litz wire is the higher Q inductor material. Below 1.6KHz, solid core wire edges the rest out.
 
#9
Brinkman said:
For the tweeter inductor (assuming it is a shunt component) you could probably upgrade it with a litz inductor with the closest DCR to the stock inductor. For the series tweeter capacitor, you could try going with foil capacitors. Russian polystyrenes are your most cost effective option. For the woofer inductor you may not be able to improve it if it is already an air-core inductor. If it is not, a 10 gauge Solen air-core inductor with a DCR less than or equal to the stock component is what you are after. If the DCR is greater, you will actually be attenuating the bass.

Do you know which version of the Raal 70-20 you are using? The 70-20XR uses an amorphous core transformer which is a premium type material. That could also be an upgrade.
Click to expand...
Thanks Brinkman. First, yes it's the 70-20XR, which I think is the most up to date version, although it may be tweaked specifically for Meniscus and this speaker and it's big brother, the Testarossa. I seem to remember a post from a few years ago where Jeff mentioned something about that. I'm really wary of using a Litz inductor on the tweeter, but am definitely intrigued by Mundorf's foil inductors, and matching or slightly reducing the DCR should be no problem there. As for capacitors, I'm really, really interested in the Vcap ODAM, but I could also move up the line with Mundorfs. The whole system right now is extremely rich and musical, sometimes shockingly so, but there's a slight lack of realism I'm chasing. I may not get all the way there with my current front end, but I think I can get closer with these tweaks, and save up and deal with the amplification later.
 
#10
FWIW, I've been responding here but my posts aren't showing up because I'm a newbie. Your Q chart is very interesting Brinkman, especially given the 1.8 khz crossover point of the drivers. I'm not that knowledgeable on some of these finer points, but I wonder how general those plots are? In other words, could the best foil have a more satisfactory Q curve than a mediocre litz coil? And as wolf teeth mentioned, I also am concerned about the capacitance resulting from all that dialectric, though I may be really misunderstanding. I had a set of Litz speaker cables when I first set up the Aegir, and it was an absolute no-go, the amp went into protection mode no matter what I tried. Other cables, no problem. My only wild guess at the time was the higher capacitance of those cables was freaking the amp out. They worked like a charm on my previous 100w AB amp, which was cooler in sound and pretty high contrast, and the 16 core Litz wire warmed it up. I figured it was all about capacitance, and attenuating something in the higher frequencies. I'm still learning, so I have to go with the best conclusions I have!
 
#11
The above attached chart is for a three-way horn speaker where (as noted on the plot) the mid-horn crosses to the tweeter horn at 6kHz where, as can be seen on the chart, the Q of the litz construction inductor is much greater than of the other type constructions. For a crossover at 1.8kHz I don’t think the technical benefits warrant changing out inductors. I personally think foil inductors would be a step in the wrong direction however. Other types of HF drivers, such as compression drivers typically require various forms of EQ correction in the crossover and on those occasions where the center frequency for the EQ may benefit from the increased Q a litz type inductor might require.

Capacitors are a can on worms I won’t get into but I would strongly suggest looking at Russian polystyrene foils before jumping straight into the Vcap price point. I would also consider experimenting with charge-coupling inexpensive Solen film capacitors before going there.

I’ve swapped out powder core inductors for air core in a woofer network and in a subjective sense sensed a night-and-day change in the tactile feel of the low frequencies but that’s a high-contrast comparison. Simply swapping and air core for an air core of lower DCR will be less subtle but is technically an improvement. I bought a pair from North Creek back when they were in business and they didn’t meet their posted specs. Solen is much better at this and less costly.
 
Last edited:
#12
It is interesting to note that on the 4-ohm MTM model of your loudspeaker, The Testarossa, Jeff opted for a litz woofer inductor. According to the above chart, litz does technically edge out the alternatives here though that wouldn’t be the case any much lower in crossover frequency.
 
#13
Funny that- it was suggested to Jeff to take advantage of the situation that could use a Litz for a premium upgrade. It was not something Jeff would have done without being asked, but he allowed the opportunity in the case of Kairos and Testarossa.

I tested Litz's vs others on the tweeters in my Cecropia Supreme project, and the Litz, which was unwound to value and had close DCR with resistance for match, caused a suckout in the response of my tweeters that I could not account for. I still really don't know what the exact cause was, but the fact that all of the other types did not do the same inferred the type of coil was to blame. When I brought this to the attention of Meniscus guys Mark and Joel, Joel had stated that he had the same experience with Litzes on tweeters in the past. He chalked it up to the high capacitive nature of the Litz coil.
 
#14
That's enough to scare me off of Litz coils for my tweeter board. I think some of the issues about foil stated in the North Creek article have been remedied, and the Mundorfs look like great quality. How can air reach the copper when it's encapsulated in dialectric material? And they aren't using elves to wind the foil, I'm guessing that in the last 15 years they've paid some attention to the weak areas of production. But I don't build speakers for a living, and know enough to make some bad things happen to my system. My original thought was that I'd have a better result with foil inductors in the tweeter board, but maybe just a higher quality solid core is about as good as it's gonna get? I just don't want to give up on the foil idea, higher Q in the upper range and stability are two things worth investigating.
As for the Vcaps, I have yet to hear anyone say anything negative about them, but I just don't know if they or the Supreme S/G/O will provide an audible step up from the Mcap Supremes already in there. They do sound lovely, musical, and I would say a bit lush even, but my system is already pretty warm and I wouldn't mind a bit more sparkle up top. The reason I can define the sound of the Supremes is I listened to them burn in, and the process was not subtle. Stunning for 2--3 hours, then flat and dull with bad center image for a looong time, up and down, it was depressing.....then one day I was listening to Leonard Cohen "you want it darker" and holy crap his big head was right there in front of me and it was spooky! All this to say, I'm not very knowledgeable, but I'm attentive and really do appreciate sonic differences, so if a couple hundred bucks and some solder can make me choke on my beer once or twice, it's worth it.
 
#15
wolf_teeth said:
I tested Litz's vs others on the tweeters in my Cecropia Supreme project, and the Litz, which was unwound to value and had close DCR with resistance for match, caused a suckout in the response of my tweeters that I could not account for. I still really don't know what the exact cause was, but the fact that all of the other types did not do the same inferred the type of coil was to blame. When I brought this to the attention of Meniscus guys Mark and Joel, Joel had stated that he had the same experience with Litzes on tweeters in the past. He chalked it up to the high capacitive nature of the Litz coil.
Click to expand...
Wolf, this is interesting. Do you have a setup where you could have put this theory to test (ie testing for the inductor’s self-resonant frequency then rearranging the the self-resonant frequency equation to solve for C)?
 
#16
That's enough to scare me off of Litz coils for my tweeter board. I think some of the issues about foil stated in the North Creek article have been remedied, and the Mundorfs look like great quality. How can air reach the copper when it's encapsulated in dialectric material? And they aren't using elves to wind the foil, I'm guessing that in the last 15 years they've paid some attention to the weak areas of production. But I don't build speakers for a living, and know enough to make some bad things happen to my system. My original thought was that I'd have a better result with foil inductors in the tweeter board, but maybe just a higher quality solid core is about as good as it's gonna get? I just don't want to give up on the foil idea, higher Q in the upper range and stability are two things worth investigating.
Click to expand...
did you ever stop and think about the construction of a foil inductor and why such a thing might appeal to a capacitor manufacturer?
 
#18
The parasitic capacitance of an inductor is electrically in parallel with the inductance; you could look at it as a parallel resonant circuit.

First you would measure the inductance (L) of your coil at the standard 1kHz.

Two different ways to find the resonant frequency (Fs):

1) impedance vs frequency sweep of the inductor. the point of highest impedance is your Fs. Or,

2) frequency sweep of the inductor with a series resistor. Attach a volt meter across the inductor and the frequency at which the voltage across the inductor is at a minimum is your Fs.

[DATS can do method 1 above but the hi-frequency limit is 20KHz so only within the audio band]

Since the formula for a self-resonant frequency is:

Freqency self-resonance = 1/(2pi(sq rootL*C))

and you know the L (measured above) and you know the frequency (measured using method 1 or 2 above), you can rearrange the equation to solve for the C, the parasitic capacitance:

C = 1/(L(2pi*Fsr)^2)

What you will find is that the Fsr is well outside of the audio range. In fact, Al Klappenberger (who produced the graph I posted above) has said as much and given his extensive background in microwave filter design (RF work) I wouldn’t try wasting my time trying to prove otherwise.

I looked at your Cecropia crossover and I have two initial guesses that might explain your measurements with the litz: 1) inductive coupling with nearby inductors and/or 2) the zip-ties on the litz coil interfering with the coil on a mechanical level (too much pressure?) and creating a frequency response aberration. Potting/impregnating the litz coil in paraffin wax might solve that; I believe they do that to foil inductors to keep them from singing.
 
#19
No, there is no cross-talk between active coils as not all are running at the same time. I checked this with meters. I made them in position to not affect each other when active.

No- the zip-ties binding the coil will not affect the value if the coil had already been dipped and the coil was measured to be accurate. In fact- I re-zipped them and re-dipped them. There is no reason to think that the tying would influence this. The more rigid a coil is, the better.

I have seen wax on wax-foils and Litzes from Jantzen, but only on Jantzens. Foils don't normally get dipped in parafin. Most of them are tape-lined on both sides and the tape edges are then melted into a hard mass after winding to encase the windings.
 
You must log in or register to reply here.
Top Bottom