Modifying the Odyssey crossover - my adventure

[spectral]

Great job there, excellent write up. It appears that swapping the caps has had a substantial effect on the signal, (for the better), which is good to hear. Out of the caps replaced, which one's are in series with the load?

All the panel ones are; the woofers' are in parallel.

As i mentioned on Agon, i have placed an order for Clarity MR for my Spires, and have divided the capacitances equally between the originals, ie, there are 2 x 47uf for the mid band, so i'll use 4 x 23.5uf. Clarity will make these to order, but will be around 5 weeks, then i'll put them on my Audiodharma for a few days.

Excellent choice! And probably much better than Mundorf. May I ask what the price for these is? I suspect about $200 per cap...

Let us know how things alter during the first 200 hrs with the Mundorf's, it sounds like it wont take long to get there..

You got that one right

Another question is how do you measure the FR, do you use software/mic, or sig gen etc. I will want to check mine before removing any part to keep as a reference, but will need to get a test method together. All very exciting, for us at least!

I just use warble tones, the RadioShack SPL meter and the Rives-corrected response curve of said meter (but not their CD, which contains pure sinewaves)...

BTW, I am looking at my pictures again and the inductor values, and I am thinking I can easily fit better Mundorf inductors in there...

 

[spectral]

So the reason I went with the Mundorfs in the end was their construction... I've been told that a capacitor is really an unavoidable RLC network - I scraped the following pseudo diagram from the ClarityCap paper I mentioned earlier:



Therefore, what sold me was their "special" induction-free winding technology: "Two capacitor windings are interleaved so that their inductances effectively cancel each other out. These two windings are connected in series."

 

[spectral]

Update 10/14/11: I will be installing the remainder of the caps (front woofer and bypass) this weekend. And I have taken the plunge and ordered replacement Mundorf Copper Foil inductors for the panel, for $560, to arrive in about three weeks...

 

[Ps68]

Well, whatever the reasoning behind the caps design it sounds like it has worked well in your O's, of course by paralleling 4 or 5 caps will reduce the resistance considerably, would that mean that inductance would increase, albeit to a small amount?

The Clarity MR's were not cheap, the 47uf, for which i'll use 2 x 23.5uf were about £155, or $245 each, so 500 USD per 47uf cap, and i need 4.. Scary, i hope this works, there's 16 caps, 4 inductors, which are actually not expensive at all, and 8 Cast resistors @ £25 ($39) each:stop:....

I'm hoping that the Mundorf copper foils will suit the panels well, have you ordered 17mm width? That gives the higher dcr, as ML use 18 or 20 gauge wire inductors, the 17mm width equals to 16 gauge.

For me and i'm sure for you Peter, that these speakers are for long term use, i'd like to think i'll still be enjoying them for another 15 yrs. The only other model i really dont want to hear are the MBL, 111f, 116 etc. I'm sure Gordon will reassure me that they sound hifi..

Good luck with this weekend's tweakery, as usual, keep us up to date on progress..

 

[Ps68]

My replies seem to take forever to get posted, unless I use my phone, so there should be a response soon..

 

[spectral]

All caps are in, and there is no longer easily detectable bass overhang. In fact, the bass is now fast, taught, and punchy on Reference Recordings and some Telarcs that I tried - actually, it's phenomenal! So much so, that I stopped looking for replacement bass drivers (I was going to try the Accuton 8" - similar electrical and mechanical characteristics with what I got from ML).

On the bypass caps - again, I first installed them on one speaker for an easy A/B - not successful, because in the first hour, as they were breaking in, the sound was a little veiled on that speaker; basically, a 10nF cap is capable of inflicting its own [detrimental] sound. After an hour, things improved; I didn't want to wait any longer, so both speakers now have the bypass caps and been playing constantly.

I'll just mention a very characteristic case here which was the primary driver for this modification - Stravinsky's Rite of Spring on Reference Recordings... I happened to have auditioned the Summit X with the exact same Spectral electronics and Berkeley Alpha DAC right before I was about to purchase - the sound got harsh on full orchestral climaxes especially with mass winds, and in fact one of the amps shut down. Well, at home now and after the mods, I just played the same piece as loud as back in the dealer room - the sound was just simply effortless, horns blaring and all. Granted the O's impedance goes as low as 1 ohm while the X's will go down to 0.8, but I don't think impedance or the amps' ability to drive the X's was a factor - I hear much less distortion, more vividness, a wider and deeper soundtsage, and larger dynamic contrasts in the new O's...

Finally... I've never heard drums sound so crisp, real and annoying before - on par with what I heard from the Magico Q3.

What I can easily say at this stage is that this has been a phenomenally successful modification, and is highly recommended... I am not sure when I'll replace the inductors yet... I no longer feel ashamed coming back home from Symphony Hall to my stereo

 

[Ps68]

Fantastic Peter, well done, i bet it is like a second honeymoon with your odysseys..

How many bypass caps did you use in the end? While most cap rollers will suggest that the caps make the substantial difference in SQ, some do believe that the inductors are as significant or maybe even more so, it is certainly true that they can mess things up just as easily, but selecting correct parameters should control this. I expect you are feeling that it sounds pretty damn good now, and you are happy with that so may not feel the need to fit inductors just yet.

I am itching to get going on mine but have a 5 week wait still. Oh well, it still sounds terrific with the stock boards.
Keep us posted on burn in and improvements!

Cheers,
Paul.

 

[miljac]

I am really glad that it worked out for you just as great as it did for me.

I did not want to comment on Summit X vs Ascent sound deliberately not to sound preposterous
but as you did it I say a big +1 from my side.

Actually, after hearing the Summit X - which is beyond any doubt a magnificent speaker - we
concluded that the sound at my den does not lag behind it at all, in some aspects being even
more natural. I think this coincides pretty well with your findings.

 

[spectral]

Thank you both. Yes I did not mean to be critical of the X's, but facts are facts when it comes to driving them with my electronics. Let it also be known that I have praised the X's in the past in this forum as well, since they do a lot of things really well. It now feels like most ML speakers should have their crossovers upgraded in order to get the max out of them; same goes for the X's perhaps.

Ps68 - yes, some do claim that inductors do make more of a difference, and this is why I decided to upgrade them too, hopefully by the end of October. I intend to do one speaker at a time and just panels at the moment, and see where we go. I'll be sure to report asap.

Finally, I used just one bypass cap per capacitor bank (i.e. per capacitor replaced). From your own Spire pictures, I think I see the same horrible yellow Bennic caps???

 

[spectral]

A couple of pics of one of the bypass caps...

 

[Ps68]

Yes, they could be Bennics, they are badged ML, no other name.
Update when you get the inductors, the caps should be well run in by then..

 

[MPS]

Thank you Spectral for your informative and enthusiastic report. Reading your comments leaves me no choice but to modify my xovers too.. Being the cheap barstand that I am, I will go for a more budget oriented approach. I will surely report my findings in this forum for the public demand

I can't see anyway how basic component modification could fail unless one is absolutely 100% satisfied with his speakers. ML is using the industry standard Solen which is fine but far from the best. I have always found these basic MKPs somewhat dull and soft as opposed to crisp and open. Electrolytics in the low pass section is really a shame but nothing exceptional from a manufacturer. They lack in sonics and will not hold as well over the time as foil types.

I have set an order for 7 (14) capacitors to replace 5 (10) capacitors in xover(s). Order value is 310€ which equals to one half decent interconnect
More to follow..

 

[spectral]

MPS - let us know...

The Mundorf coils will take another 3 weeks to arrive, so we are talking ~ Thanksgiving installation...

 

[spectral]

The Mundorf panel coils are finally arriving this Wednesday 11/30, due to enormous delays at Mundorf and Fedex (shipment was sitting in Fedex/DE for 9 days due to "bulked out aircraft" - meaning, higher priority items were pushing it out)... But at least the caps have now settled down, and I'll say again - WOW, how did I live w/o them for so long. I've been listening mostly to CDs that I haven't touched for over a year at least, and I am extremely impressed... last one was a soprano singing Haydn - truly incredible dynamics and overall rendering.

The decision to go with banks of capacitors for the larger values was a risky move - the caps may have not ended up charging and discharging at the same time, causing disaster - but in the end I was convinced by others to do it. The end result is much purer capacitance by reducing in-series resistance and inductance (obviously because they were connected in parallel). Although this move did also push the price higher, it was probably worth it.

The reason for replacing the panel coils next is to put in as pure cooper with less loss as possible, considering the hundreds of feet of wire involved. We'll see how that plays out, this weekend.

 

[Ps68]

It's great to hear how things are developing Spectral, be sure to post your findings with the inductor replacements..

Some of my parts are here, still waiting for the Duelund resistors, not sure why so long though.. MR caps are big, though not an issue for me due to ext cases. Anyway, I'm really looking forward to getting going with the build, the caps and foils have been on my burner for 5 days so should just need a bit of time to settle/form. Did your Mundorfs settle down quickly?

Best Regards
Ps

 

[spectral]

Did your Mundorfs settle down quickly?

Took them about 200 hours, as expected, with wild variations during that period... I will post thoughts on the coils hopefully this weekend

 

[spectral]

So the panel coils are in... Quick measurements: the larger coil measures 0.1 ohms resistance vs 1.5 in the stock. I am not sure what the new lower-resistance, 14awg (vs 16-18 in the stock) coils have done to the impedance, but initial impressions are that there is a very slight improvement in terms of dynamic headroom, perhaps less than half a dB.

I did upgrade one speaker first again, and did the channel switching thing using the revox, and heard only slight differences, certainly nothing like the night-and-day with the caps. We'll see how things change over time, but one thing that particularly pleases me is how crisp and dynamic timpani sound, with no overhang.

I am not sure if I want to replace the woofer coils at this point, or experiment with the Accuton drivers (which are similar but perhaps not close enough), or just enjoy music... Will post pictures next.

 

[spectral]

And here are some pictures - thankfully there was plenty of room to work in...

 

[Ps68]

Nice work Peter, your first pic could have been from my camera, our coils are the same. Have you upgraded both speakers or just the one so far? Not sure how much burn in they need but mine will come off the burner tonight, 5 days on.
How, from memory, would you compare your sound now to Summit x when you heard them? I know some time has passed, though any thoughts would be of interest.

 

[spectral]

Nice work Peter, your first pic could have been from my camera, our coils are the same. Have you upgraded both speakers or just the one so far? Not sure how much burn in they need but mine will come off the burner tonight, 5 days on.
How, from memory, would you compare your sound now to Summit x when you heard them? I know some time has passed, though any thoughts would be of interest.

Hi there, both speakers' panels are updated, and I think I will now upgrade the woofers' too, after taking space measurements last night and asserting that I can fit the monsters that I expect them to be.

Copper will burn in but I don't know for how long. What I have noticed within a day is that this upgrade has significantly reduced distortion, and has now removed the last bit of strain I was able to detect with some very demanding instruments, like piano fortes (I use a Chesky/Earl Wild/Chopin Etudes CD primarily for this), and I have reached the point of wanting to turn the volume up to enjoy these extreme phenomena.

Last night my jaw dropped listening to Bruch's Scottish Fantasy with Oistrakh, both on XRCD and LP. Extreme pinpointing of the violin; layers and layers of soundstage behind it; powerful, effortless and very realistic rendering of the violin right there in front of me - complete you-are-there emotion; total disbelief. It's great to be able to hear what a Stradivarius might really sound like... shivering...

What this entire crossover update has achieved (and you'll see for yourself) is bring out the emotion in music reproduction. You will be stunned by how realistically strings are rendered (assuming upstream components are up to the task) - you will feel them vibrating (as opposed to making just a sound at some frequency); this is what the speed of an electrostat is all about. The speaker currently still significantly falls short in shear size of the orchestra, something it will never achieve, though still spectacular with great recordings, and of course the bass's speed can still be improved.

WRT Summit X, yes it's been a little while since I last auditioned it (~7 weeks), but I am pretty certain this new Odyssey smokes it in just about every aspect: larger and faster panel (due to crossover updates), no rattling woofer enclosure (the X is really bad in that respect), higher level of realism than the X, less distortion; and I still think the tonality of the Summit series is tipped up higher and meant to be matched to the room by tilting, but up to a point - I like the vertical panels in the older speakers and the CLX. I intend to go back and re-listen to the X and the CLX, with the same music and driven by the same Spectral electronics. I think the question is not whether the new O's are better than the X, but rather how do I get my hands on a Monolith III to upgrade now that would be the speaker to reproduce the size of the orchestra more realistically