What's Best in Crossover Capacitors

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
As if this form wasn't technical enough, it's about to get worse... So here's the deal; I am going to punt on buying a new pair of Summit X's and explore upgrading the crossovers in my ML Odyssey's. The sound is so beautiful that I can't justify the X purchase... yet.

So I have the schematic from ML, and there are two crossovers really, one for the panel and another for the woofers. First the woofers - simple crossover per woofer: one inductor in series as you'd expect, one capacitor in parallel with the woofer. The capacitors are 100uF (forward firing) and 170uF (backward firing); these will be replaced by Mundorf MKP, with 100pF and 150pF respectively, and it's hard to find better caps with such high values. Question #1: what will the 150uF cap do to the rear firing woofer, as opposed to the original 170uF, and by extension, what will a 220uF cap do?

Then comes the panel - here we have four capacitors: 10uF, 13uF, 51uF (all polypropelene) and one 47uF polyester. The question is then - what are the best capacitors for this type of application? For the three polypropelenes I can only find the very expensive but super-performing Deulund VSF's, for a total of $1500 per speaker! I can't find yet a *worthy* replacement for the 47uF . Next, what happens if you replace a polyester with a polypropelene?

There are some incredible caps out there, from the Mundorf Supremes, to ClarityCaps, VCaps (some, not all), all the way to the Deulunds, and some interesting comparison blogs, like audiocircle, humblehomemadehifi, and laventure.

Whatever I chose must not alter the overall character - for example, I read that the Mundorf Supreme Silver in Oil et al tip up the treble at the expense of the bass, while the Deulunds are allegedly completely neutral.

What do the more knowledgeable have to suggest?

PS: I'll start another thread on coils and resistors later on; for resistors at least, I am going with Mundorf M-Resist...
 

MylesBAstor

Well-Known Member
Apr 20, 2010
11,236
81
1,725
New York City
As if this form wasn't technical enough, it's about to get worse... So here's the deal; I am going to punt on buying a new pair of Summit X's and explore upgrading the crossovers in my ML Odyssey's. The sound is so beautiful that I can't justify the X purchase... yet.

So I have the schematic from ML, and there are two crossovers really, one for the panel and another for the woofers. First the woofers - simple crossover per woofer: one inductor in series as you'd expect, one capacitor in parallel with the woofer. The capacitors are 100uF (forward firing) and 170uF (backward firing); these will be replaced by Mundorf MKP, with 100pF and 150pF respectively, and it's hard to find better caps with such high values. Question #1: what will the 150uF cap do to the rear firing woofer, as opposed to the original 170uF, and by extension, what will a 220uF cap do?

Can't you parallel and build a bank of caps totalling 170 uF? That's what I did with my old Maggies to get the correct value. Is it a space issue? As I remember reading, it is preferable to use say 10x10 uF to one 100 uF cap.

Then comes the panel - here we have four capacitors: 10uF, 13uF, 51uF (all polypropelene) and one 47uF polyester. The question is then - what are the best capacitors for this type of application? For the three polypropelenes I can only find the very expensive but super-performing Deulund VSF's, for a total of $1500 per speaker! I can't find yet a *worthy* replacement for the 47uF . Next, what happens if you replace a polyester with a polypropelene?

Didn't do this with the ML but with the 12 uF polyester that was used on the Maggie 3A ribbon. And it was not trivial. More spatial information, less blurring of transient attack, not to mention intranote silence as well as greater transparency. On the midrange, that "Maggie" coloration, that carboardy sound, was vastly reduced and the midrange and bass was much tighter and quicker.

There are some incredible caps out there, from the Mundorf Supremes, to ClarityCaps, VCaps (some, not all), all the way to the Deulunds, and some interesting comparison blogs, like audiocircle, humblehomemadehifi, and laventure.

Whatever I chose must not alter the overall character - for example, I read that the Mundorf Supreme Silver in Oil et al tip up the treble at the expense of the bass, while the Deulunds are allegedly completely neutral.

What do the more knowledgeable have to suggest?

PS: I'll start another thread on coils and resistors later on; for resistors at least, I am going with Mundorf M-Resist...

Only used the Mundorf Au/Ag in oil caps in my tape preamp. and have been very happy. Compared to some other caps, they were smoother, less grainy, greater transparency and much better imaging. In addition, not as harsh as some lesser caps. As far as the Duelands go, am getting some in next week to try in my tape preamp. Am going to burn them in on the Cable Cooker for 300 hrs and then install them. Let you know in 6-8 weeks how they sound compared to the Mundorfs.
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
Can't you parallel and build a bank of caps totalling 170 uF? That's what I did with my old Maggies to get the correct value. Is it a space issue? As I remember reading, it is preferable to use say 10x10 uF to one 100 uF cap.

I hope this is NOT the case - by the time one cap charges the other may discharge and a huge mess ensues. I have tried this on a tweeter before with 3 or 4 caps and the sound was just dead.

Didn't do this with the ML but with the 12 uF polyester that was used on the Maggie 3A ribbon. And it was not trivial. More spatial information, less blurring of transient attack, not to mention intranote silence as well as greater transparency. On the midrange, that "Maggie" coloration, that carboardy sound, was vastly reduced and the midrange and bass was much tighter and quicker.

So polypropelenes are better than polyester? Good.

As far as the Duelands go, am getting some in next week to try in my tape preamp. Am going to burn them in on the Cable Cooker for 300 hrs and then install them. Let you know in 6-8 weeks how they sound compared to the Mundorfs.

I'd hate if I had to spend $3000 on six capacitors, and still be two short... Cable Cooker huh? I think I may need one too.

Thanks for the fast response.
 

MylesBAstor

Well-Known Member
Apr 20, 2010
11,236
81
1,725
New York City
I hope this is NOT the case - by the time one cap charges the other may discharge and a huge mess ensues. I have tried this on a tweeter before with 3 or 4 caps and the sound was just dead.

Hmmm....don't quite understand that? Sound more like in series than paralleled. Never heard that in my system. When I did it, I used Rel PPs.

On another note, could also experiment with bypassing some of the polypropes with small value, say teflon (since don't think could find any polystryrenes) caps.
 

fas42

Addicted To Best
Jan 8, 2011
3,973
3
0
NSW Australia
Playing with capacitors like this is far from trivial, if you use the wrong combination of values paralleled you will end up with a mini disaster, as far as sound quality is concerned. Ben Duncan did an excellent series of articles on this over 25 years ago, unfortunately not on the net as far as I know.

A rule of thumb if trying film caps, those with names starting with "poly" is that either you combine values that are very close together or very far apart, use a 50 times or more ratio with the latter. Doing other will allow a resonance, an impedance peak will occur, and you don't want this!!

Paralleling lots of caps of same value is generally good, depending on what you're after ...

Frank
 
Last edited:

MylesBAstor

Well-Known Member
Apr 20, 2010
11,236
81
1,725
New York City
Playing with capacitors like this is far from trivial, if you use the wrong combination of values paralleled you will end up with a mini disaster, as far as sound quality is concerned. Ben Duncan did an excellent series of articles on this over 25 years ago, unfortunately not on the net as far as I know.

A rule of thumb if trying film caps, those with names starting with "poly" is that either you either combine values that are very close together or very far apart, use a 50 times or more ratio with the latter. Doing other will allow a resonance, an impedance peak will occur, and you don't want this!!

Paralleling lots of caps of same value is generally good, depending on what you're after ...

Frank

I'm not sure they still do it but the old MIT Multi-caps were essentially multiple paralleled caps all rolled up into one.
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
So the results after the first change - the cap to the rear woofer - are very encouraging... I can't believe people still use electrolytics in the crossover. Testing with warble tones after finishing just one speaker, the difference is subtle but audible if you put your ear next to the woofer - the modified is clearly clearer, the other muddier.

Next up, the front woofer electrolytic caps and the Solen panel caps (Class B apparently, popular in power supplies because of the high values and high voltage ratings, but not so good in crossovers).
 

MylesBAstor

Well-Known Member
Apr 20, 2010
11,236
81
1,725
New York City
So the results after the first change - the cap to the rear woofer - are very encouraging... I can't believe people still use electrolytics in the crossover. Testing with warble tones after finishing just one speaker, the difference is subtle but audible if you put your ear next to the woofer - the modified is clearly clearer, the other muddier.

Next up, the front woofer electrolytic caps and the Solen panel caps (Class B apparently, popular in power supplies because of the high values and high voltage ratings, but not so good in crossovers).

Sounds like you're making the right decision!

It'll take a few hours for the caps to sound good--not to mention, once you shut down the estats, they take 24 hrs to recharge. Hard to judge sound til then :(
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
Playing with capacitors like this is far from trivial, if you use the wrong combination of values paralleled you will end up with a mini disaster, as far as sound quality is concerned. Ben Duncan did an excellent series of articles on this over 25 years ago, unfortunately not on the net as far as I know.

A rule of thumb if trying film caps, those with names starting with "poly" is that either you combine values that are very close together or very far apart, use a 50 times or more ratio with the latter. Doing other will allow a resonance, an impedance peak will occur, and you don't want this!!

Paralleling lots of caps of same value is generally good, depending on what you're after ...

Frank

Frank, sage advice, thanks! So I am having trouble breaking down a 47uF and a 51uF capacitor - how would you do it? There are extremely expensive 50uF Duelund, or 17uF (3x) ClarityCaps etc., but I don't want to spend thousands of dollars for these two values. I can do two 22uF Mundorfs in place of the 47 (because the value is within the tolerance range), and something similar for the 51 - would that be acceptable?
 

Speedskater

Well-Known Member
Sep 30, 2010
941
15
368
Cleveland Ohio
We have a nasty typo here:

The capacitors are 100uF (forward firing) and 170uF (backward firing); these will be replaced by Mundorf MKP, with 100pF and 150pF respectively, and it's hard to find better caps with such high values.

Somehow the uF became pF and that just won't work well.
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
Nasty indeed, thanks for catching these! Yes I eventually got and installed the 100uF Mundorf MKP, and for the 170uF, I am will install three 56uF's - that's close enough, although I could throw in a couple of 1pF's to get 170 exactly.

Now - I keep reading that I should bypass the Mundorf Supremes - slated for the panel crossover - with the VISHAY-RODERSTEIN MKP-1837 10nF! Why? I picked up the following from partsconnexion.com:

Sound: I was tipped by Klaus Witte of Germany to try this capacitor as a bypass cap for the Mundorf M-CAP SUPREME. I tried them as a bypass for the tweeter series caps in my Progress speaker and I must say I am very impressed! To get straight to the point they don't change a Supreme into a Supreme Silver-Oil but they really do clear things up. I must admit I was skeptical at first as the value is only 10nF (0,01uF) - and the caps in the Progress are 12,6uF. The difference is most noticeable with classical music but also good quality recordings of jazz and fusion benefit: No change in sound stage width or depth but there is more "concert hall acoustics" that let you get into the recording more. Not as liquid as silver/oil but they take away the "grainy" edge from the Supreme's. A gain in clarity and transparency making instruments better separable from each other, the violins in an orchestra are a group of individual violins instead of one mass. Jazz drum brushes sound more like a brush than a "shush".
Thanks
 

DonH50

Member Sponsor & WBF Technical Expert
Jun 22, 2010
3,947
306
1,670
Monument, CO
Capacitors* are actually LCR networks and above their self-resonant frequency they act more like inductors than capacitors. Bigger caps have lower self-resonant frequencies, so adding smaller caps in parallel helps create a better broadband circuit than a single cap. It's also one reason for using several smaller caps in parallel instead of one large one. The trick is to find a combination of values that not only provides the desired capacitance for the crossover but also provides broadband capacitive reactance without a lot of "wiggles" in the FR.

Frankly, I tend to think inductors make a bigger difference in crossovers, as most do not use real heavy gauge wire and some of the iron-core (vs. air-core) ones have lower power handling than desired. It is really annoying to measure a speaker and find the crossover L is contributing to the distortion (it rarely if ever dominates but I have seen it climb into significance).

IMO - Don

p.s. Actually, so are resistors and inductors...
 

fas42

Addicted To Best
Jan 8, 2011
3,973
3
0
NSW Australia
So I am having trouble breaking down a 47uF and a 51uF capacitor - how would you do it? There are extremely expensive 50uF Duelund, or 17uF (3x) ClarityCaps etc., but I don't want to spend thousands of dollars for these two values. I can do two 22uF Mundorfs in place of the 47 (because the value is within the tolerance range), and something similar for the 51 - would that be acceptable?
Yes, two of the same half of value would be perfectly acceptable, in fact they would be better than the single value unit! Why, because using 2 or more reduces the inductance, as Don mentions above, and also the equivalent series resistance, the ESR, would have a better curve with respect to frequency.

The most important thing is to mount them extremely "firmly", not touching something else loosely, they should be locked down in position so they can't move, or vibrate against anything. Best is if when done they feel like they been cemented into place ...

Frank
 

garylkoh

WBF Technical Expert (Speakers & Audio Equipment)
Sep 6, 2010
5,599
225
1,190
Seattle, WA
www.genesisloudspeakers.com
Capacitors* are actually LCR networks and above their self-resonant frequency they act more like inductors than capacitors. Bigger caps have lower self-resonant frequencies, so adding smaller caps in parallel helps create a better broadband circuit than a single cap. It's also one reason for using several smaller caps in parallel instead of one large one. The trick is to find a combination of values that not only provides the desired capacitance for the crossover but also provides broadband capacitive reactance without a lot of "wiggles" in the FR.

Frankly, I tend to think inductors make a bigger difference in crossovers, as most do not use real heavy gauge wire and some of the iron-core (vs. air-core) ones have lower power handling than desired. It is really annoying to measure a speaker and find the crossover L is contributing to the distortion (it rarely if ever dominates but I have seen it climb into significance).

IMO - Don

p.s. Actually, so are resistors and inductors...

Don, you are giving away all the secrets!

I addition to Don's suggestions, I would use two different values to make a larger one instead of two of the same value. Different values would sound different even if they total the same. If you can, model the capacitance, ESL and ESR to reduce the "wiggles" in the FR. That may have been why when you tried this on a tweeter with 3 or 4 caps, it sounded dead. The 3 or 4 you picked "clashed". Don't be afraid to mix brands, types and/or series within a brand. As with all components, more expensive doesn't necessarily equate to better.

Something that no one else may have considered: Ampohm paper-in-oil capacitors. Unfortunately, the company closed down. So, manufacturers like me can't use them - what to do when they run out? - but for DIY you don't have that restriction. I loved the sound of their paper-in-oil tin foil capacitors.

The MKP-1837 is a good suggestion. I am surprised that more loudspeaker designers haven't discovered that yet. It can make a cheap cap sound like a very good one, and make a very good cap sound absolutely unbeatable. Problem is that it looks really cheap :eek: Manufacturers can usually hide one under the circuit board. Another to try is a good ceramic C0G. Ordinary ceramic capacitors are the worst for crossovers, but the C0G ones are one of the best for bypassing a large electrolytic.

I agree with Don that inductors can make all the difference. On your woofer, since the inductor is in series, 100% of the signal is going to go through the inductor before it gets to your woofer.

I like the Erse SuperQ and the Alpha Core Foil, but again, neither are a panacea. They are also not as expensive as some of the more highly regarded ones, but in the right application, I think that they sound better. I've got a Alpha Core inside my Maggie 3.6's.

For the resistors, nothing beats the Vishay metal FOIL resistors.
 

MylesBAstor

Well-Known Member
Apr 20, 2010
11,236
81
1,725
New York City
Don, you are giving away all the secrets!

I addition to Don's suggestions, I would use two different values to make a larger one instead of two of the same value. Different values would sound different even if they total the same. If you can, model the capacitance, ESL and ESR to reduce the "wiggles" in the FR. That may have been why when you tried this on a tweeter with 3 or 4 caps, it sounded dead. The 3 or 4 you picked "clashed". Don't be afraid to mix brands, types and/or series within a brand. As with all components, more expensive doesn't necessarily equate to better.

Something that no one else may have considered: Ampohm paper-in-oil capacitors. Unfortunately, the company closed down. So, manufacturers like me can't use them - what to do when they run out? - but for DIY you don't have that restriction. I loved the sound of their paper-in-oil tin foil capacitors.

The MKP-1837 is a good suggestion. I am surprised that more loudspeaker designers haven't discovered that yet. It can make a cheap cap sound like a very good one, and make a very good cap sound absolutely unbeatable. Problem is that it looks really cheap :eek: Manufacturers can usually hide one under the circuit board. Another to try is a good ceramic C0G. Ordinary ceramic capacitors are the worst for crossovers, but the C0G ones are one of the best for bypassing a large electrolytic.

I agree with Don that inductors can make all the difference. On your woofer, since the inductor is in series, 100% of the signal is going to go through the inductor before it gets to your woofer.

I like the Erse SuperQ and the Alpha Core Foil, but again, neither are a panacea. They are also not as expensive as some of the more highly regarded ones, but in the right application, I think that they sound better. I've got a Alpha Core inside my Maggie 3.6's.

For the resistors, nothing beats the Vishay metal FOIL resistors.

Hardly a secret :) The benefit of using multiple paralleled caps was I believe talked about in Audio Amateur many, many years ago. When I redid my Maggie MGIIIa xover, used 16x10 uF instead of 1 x 160 uF cap for the bass shunt (yes it made a difference even on the shunt!). Did the same for the midrange too_Of course, using multiple paralled caps adds to the overall cost. I think that in raw parts (chokes, wiring, binding posts, caps, building a xover box, etc.), I spent close to $700 (that was with some discount). Oh, and I used back then the beautifully wound Solen chokes of 10 ga for the bass and either 12 or 14 for the mids/tweeter. The bass chokes weighed over 30 lbs each ;) So even if a manufacturer would go this route (and nowadays there's far more exotic and expensive caps available than the Rels polystyrenes and small teflons that I used), that would add close to $3000 to the cost of the speaker.
 

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
Next question... if I were to add a 2.8uF or 3.3uF to the two 22's above, in order to approach the presumed optimal 47uF, would they not be far enough from the 22's? Am I better off perhaps with three 1uF instead?

Thanks
 

garylkoh

WBF Technical Expert (Speakers & Audio Equipment)
Sep 6, 2010
5,599
225
1,190
Seattle, WA
www.genesisloudspeakers.com
Hardly a secret :) The benefit of using multiple paralleled caps was I believe talked about in Audio Amateur many, many years ago. When I redid my Maggie MGIIIa xover, used 16x10 uF instead of 1 x 160 uF cap for the bass shunt (yes it made a difference even on the shunt!). Did the same for the midrange too_Of course, using multiple paralled caps adds to the overall cost. I think that in raw parts (chokes, wiring, binding posts, caps, building a xover box, etc.), I spent close to $700 (that was with some discount). Oh, and I used back then the beautifully wound Solen chokes of 10 ga for the bass and either 12 or 14 for the mids/tweeter. The bass chokes weighed over 30 lbs each ;) So even if a manufacturer would go this route (and nowadays there's far more exotic and expensive caps available than the Rels polystyrenes and small teflons that I used), that would add close to $3000 to the cost of the speaker.

Yeah - before I turned my hobby into a business, I also modified my Maggie 3.5's. I think that a few hundred dollars in caps and inductors resulted in far more benefits then than a few thousand in a better amplifier.
 

garylkoh

WBF Technical Expert (Speakers & Audio Equipment)
Sep 6, 2010
5,599
225
1,190
Seattle, WA
www.genesisloudspeakers.com
Next question... if I were to add a 2.8uF or 3.3uF to the two 22's above, in order to approach the presumed optimal 47uF, would they not be far enough from the 22's? Am I better off perhaps with three 1uF instead?

Thanks

Best thing to do is to buy a bunch and find the combination that rocks your boat. Instead of two 22uF's try combinations of different values.
 

DonH50

Member Sponsor & WBF Technical Expert
Jun 22, 2010
3,947
306
1,670
Monument, CO
What Gary said. In the RF world, and in this case I think it applies, the rule of thumb is to go in no more than decade'ish steps, an order of magnitude (factor of ten) spacing between values. In practice, you still need to measure (or listen) for what responds best. (Sorry for giving away secrets, Gary, but I think you'll survive on your own merits -- giving out the textbook basics does not directly translate to the many hours of effort required to realize a good design!)

Myles, shunt caps matter too! If they "fall apart" at high frequencies where you think it shouldn't matter, they are potentially letting HF energy into a speaker not designed to handle it. At best it will reduce dynamic range since unwanted signals are moving the cone, and at worst the HF energy will be converted to heat and reduce the driver's life. In between, the HF energy can modulate the cone and increase distortion.
 
Last edited:

ack

VIP/Donor & WBF Founding Member
May 6, 2010
6,774
1,198
580
Boston, MA
I am beginning to get it - thanks to all :eek: So I will split the 51uF into 5x10uF's + 1uF, and the 47 probably into the two 22's plus 1uF's as necessary
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing