Is it good to upgrade the crossovers in your speakers?

A confessed audiophile, threw this Forum I have contracted “Tweakitus”.
QSA fuses, SRA Platforms, Townshend Podiums, NPS Q45T, ad nauseam.

The latest bug in my bonnet is upgrading the crossovers in my speakers.

I asked my speaker designer about part quality. He did mention that caps, for example, can cost as much as $800 each. And that he has gone up to $50 ones.

Like all things “Hi Fi”, cost does not necessarily dictate quality. And I doubt that I would opt for 2 $800 caps. But there must be a sweet spot for crossover components? Any ideas?


Besides the cost, there may be changes in sonic character. I suppose the initial question would be, What do you hope the speakers to do that they're not doing adequately now? What aspect would you hope to improve? After that, I would go searching for some correlations between the parts you would change and other testimonials about what those parts contribute to a speaker's sonic character and performance.

Assuming you already have speakers you are generally comfortable with, I would not recommend that you screw around with the OEM design and OEM crossovers…especially as a tweak.

There is no “set” or “standard” crossover point - different speakers mean bespoke different crossover points.They are already “tuned” at a certain precise point with their bespoke speaker drivers and parts. What you are proposing is to introduce an unknown new component on an informal ad hoc basis that ignores the speaker’s overall design characteristics…. This can - ….and usually will… - be a recipe for a bad outcome.

“ if it ain’t broke, don’t fix it “ rules.

Your are better off to consider other alternatives if your audio presentation is lacking.The philosophy is clear. An affordable speaker (… which includes yours left in its full OEM current state …) matched with a high-end source and amplification makes more sense than an expensive speaker with a cheap amp and source.

Parts cost/quality is so subjective, but my advice to you is this.  If you have an itch to get your hands dirty build a new speaker from a kit.  Then you can upgrade all you want to and be involved in every stage of the process. You'll learn a lot more this way and be more satisfied.

Upgrading parts in an existing speaker may or may not lead you where you want to go unless you've already done it several times.  Rather than play with your perfectly nice existing set, I strongly suggest you get a hold of a good 2 way speaker kit from Meniscus or Madisound or even Parts Express and make it 100% yourself.

You can change everything in a crossover but the point is what do you want to achieve. More detail, different character?

The easiest part are the resistors and internal wiring and most likely what you refer as a sweet spot. Capacitors are costly, can bring a very good improvement but can also expose drive units. With wrong inductors it can get a real mess.

I agree with all posts above, leave them in factory form unless you have good experience in parts and building crossovers as not upsetting speakers' voicing.




It’s only a good idea if you really know what you are doing. Or just want to tinker and maybe learn and see what happens. Otherwise, if you have the right good quality speakers for you to start with, there is no need. The crossover was already designed and integrated by an expert. Also It is most likely that any changes one makes on any well respected speaker model will limit the # of potential buyers if you decide to sell, assuming one is ethical and properly transparent about such things

I'll be contrary here and say go for it. As long as you're not changing values of components you're not going to change voicing to a huge degree. And the best quality parts are only going to improve timbre, natural sound quality. You can always go back to stock if you didn't like changes. Understand exactly what sound qualities you're looking to improve, do your research on parts you plan to upgrade. Get it right and you're speakers will be much improved.

I agree with sns, you can always go backward if a change is not in the right direction.  There are a lot of popular choices of current production parts that are not necessarily the way to go; vintage parts should not be overlooked even if the common wisdom is that they go bad with time.  Some of the best recent builds I’ve heard utilized such vintage caps as Western Electric paper-in-oil capacitors, Cornell Dublier caps, and Jupiter caps (not a secret, as many of these parts are ultra expensive).  Internal wire can also be quite important and the right choice can be hard to determine; a builder I know basically chooses between various types of Audio Note speaker wire, including crazy expensive Sogon wire.

I have rebuilt crossovers and wiring in two different speakers, and for the improvement made, looking back, I'm not sure that it was really worth it dollar for dollar. I spent the most on capacitors of course. Looking back, I think with what I know now I would change wiring and leave the rest alone unless there was a problem. My current two pair of speakers are both wired with what appears to be lamp cord, and I know that high quality wire with better insulation will give a better sonic result, so that's something I'm planning on doing soon.

Unless the caps are known to be inferior, I’d leave them alone.  You might stumble into what you perceive as an improvement, but you could also cause some damage, or end up  not liking the change as much as the originals.  

Ten years ago I had my vintage (1978) Large Advents and B&W 805 Matrixes (1997 or so) upgraded by John Van Leishout of Van L Speakerworks in Chicago. I pretty much gave him carte blanche within reason to upgrade everything. I know the caps, inductors, and internal wiring were replaced. The sound improvement was substantial in both sets of speakers. I recall that at some point he called me about various cap options for the B&W's ranging from $200 to $1500 per cap. I am pretty sure I told him to go with something in the $200-$300 range and that my "carte blanche" was not really blank. 

My recommendation is that if you are considering this truly for sound and not also for the fun of the exercise, than have a pro do the work. 

For those advising against diy or any crossover upgrades. Did you have negative experience with crossover upgrades? I've been doing diy crossover upgrades for many years, these have never changed the essential character of speaker. I've only experienced improved resolution, more natural timbre with the parts I've chosen. Duelund, Jupiter, Audyn caps, Path, Texas Components TX2575 resistors, Jantzen inductors, Duelund, Furutech internal wiring, never has a single one of these parts not been an improvement over original parts.


There is no magic in most original crossovers, parts often chosen because of cost or space constraints. Crossover upgrades can one of the most cost and/or overall effective upgrades one can make.

@sns et al

I have been considering the replacement of the capacitors in my speakers which are approaching 20 yrs old. I’ve always heard that 20 yrs is the life expectancy of caps, thus my concerns. Beside that is that I believe there is a slight loss of transparency over the years or expectation bias at work?, Even so, I like the overall sound signature I have and want to keep what I have, if for no other reason, the $12-15k replacement cost. I was told that if I replace with the same values, it should be fine

So, the question is it a good thing to do as a maintenance project? Then too, which capacitors? Nichicon made a big improvement to my 20+ yr old preamp. Are they appropriate for a speaker (original Silverline Sonata)?

Instead of buying overly-overpriced and overly-overrated audiophile caps, makes far more sense purchasing higher performance speaker.

Other than that, replacement premium quality caps can be found on generic electronic sites such as or They have Nichion made caps as well.

I replaced *everything* in my Thiel CS2.4s except the cabinets and drivers. Sounds much better than original. But it was a lot of work and expense. And fraught with the possibility that I would screw it up. Important to match values of original parts (well, ok to increase voltage of caps or wattage of resistors) and make sure that total series resistance is close as possible to original circuit.

As for capacitor life, electrolytics start to degrade after, maybe, 10 years and should probably be replaced by, say, the 20 year mark. But film caps are very stable, essentially lifetime caps.




It’s only a good idea if you really know what you are doing. Or just want to tinker and maybe learn and see what happens. Otherwise, if you have the right good quality speakers for you to start with, there is no need. The crossover was already designed and integrated by an expert. Also It is most likely that any changes one makes on any well respected speaker model will limit the # of potential buyers if you decide to sell, assuming one is ethical and properly transparent about such things



Good advice.

Even if they were vintage speakers I'd still keep the values exactly the same.

Things like crossover points and the related dispersion patterns are difficult enough for experts with good measuring equipment and software.


I have $2300 worth of parts on order now for crossover upgrades. Obviously I’m a believer. I also obviously disagree with @czarivey . But it depends on what speakers you have and what their capability is. You don’t want to spend a lot of money on an upgrade to a speaker with poor drivers, for example.

@sns said it well:  As long as you're not changing values of components you're not going to change voicing to a huge degree. And the best quality parts are only going to improve timbre, natural sound quality.


@artemus_5  As mentioned above 20 year old capacitors getting long in tooth, should probably replace. As far as I'm concerned electrolytics don't belong in speaker crossovers, the issue is whether film caps will fit, large values mean less likely. Now, I"ve heard very good things about the VH Audio ODAM,  not as costly as others and relatively smaller size.  Sonic caps are much lower price alternative to the caps I spoke about in previous post.


I'll relate an interesting interaction I had with Bobby Palkovich (RIP) many years ago. Bobby's life work was development of his Merlin line of speakers, constant refinement of original design. Well, naivety lead me to call various manufacturers, designers in those days to report on various mods/substitutions of parts I made to their equipment. So one day I decide to call Bobby and report on Duelund cap mod I had done to my Merlin VSM-MM bam module and internal speaker crossover. Bobby was not too happy with me ( as were other manufacturers I contacted in those days), but showed some interest in spite of this negativity. Lo and behold, perhaps a year later Bobby was offering the Duelund upgrades to his beloved Merlins. It is to Bobby's credit he was open minded enough to hear me through and try my mods on his design.  Bobby was already aware of boutique caps prior to my conversation as  he was usiing Hovlands at that time, he discovered, as I had the Duelund VSF were an improvement. Many OEM's not as open minded, parts are parts to them, they'll try to convince you the same.

I've upgraded crossovers on a dozen pairs of speakers.  Never spent more than $200 total for caps and wiring, sometimes Mortite packing.  The $200 went into a pair of Epos 14.  Every speaker was a success and fun to do.  Never changed any of the cap values.

The experts above don’t even mention insertion values they say you are fine as long as you use the same values but those same values have different levels of insertion loss. This is a reason upgraders hear differences since insertion loss can change the output levels of your transducers. In my opinion, most modding networks really don’t understand what they are doing. And tend to consider cost more than anything else when selecting parts used in the upgrade. Lots of confirmation bias involved in crossover mods as those who don’t understand but spent much and did a thing now hear an amazing difference mostly described as an improvement you don’t hear them saying what they ever did was not as good as stock its always better since they spent much and did a thing. I see most not comparing upgraded networks to stock networks. These are all newbie mistakes but I do know you all will still do it because logic is not involved. And of course, it will sound amazing right? like the guy above {Every speaker was a success and fun to do} really every? hmm

IMHO you are pissing money away on items that in the end make minimal difference. Save up and buy better equipment particularly speakers. If you want to "tweak" get a measurement microphone ($300 at Parts Express) and work on your room. 


Yes, very good points raised.  This is like people raving about tweaks they employ.  Either they do, or they don't change the character of the sound, and if they do, why does it always seem to be for the better?  I suspect that bias is involved--people expect (or hope for) an improvement so that is what they hear.  

If one goes into these types of projects with an open mind, and the willingness to accept that a component swap may not work out for the better, then it is part of the hobby and I say go for it and GOOD LUCK. 

A local dealer who builds his own in-house speakers and electronics frequently encounters "improved" gear utilizing "better" parts that sound like crap.  Just because the parts are expensive and people rave about them does not mean that they will improve a specific piece of gear or will sound better based on the taste of the particular listener.  I heard one of his amps that had been modified by another company that was shockingly bad sounding (to my taste) even though the parts changes were of the same value (e.g., use of "better" Blackgate coupling capacitor).  In this particular dealer's speakers, Mundorf and Duelund capacitors are a complete bust, but, that is not to say that they won't work out in other designs; parts choices are very specific to the particular speaker an so one's chances of improving on the original design are not good unless one tries a lot of different alternatives.  

I found a pair of Sonics AS331 for 20 bucks. I looked around and found the usual audiophile snobbery about them online but decided since the cabinets and drivers were in awesome shape that I would give them a once over. I went in and found they are very well made, plywood, screws, and an actual crossover instead of some cheapo soldered in run of the mill caps and inductors. These were very nice! I replaced the caps with new/modern film caps, updated the inductors too. Tossed out the wiring and replaced it with some high quality 12ga wires and soldered them to the drivers. I sealed the enclosure and gave it a teak oil finish that made them look very nice. I was even pleased with the fact the lattice grill was real wood. I then hooked them up and cranked "It Aint Me" (CCR) and took a trip back to the 70's. I will be keeping these and no matter what the snobs say they sound great to me, look great and did not cost me a fortune. Go ahead and do your speakers if you want, some of us do not have the means to drop thousands on equipment, but if it sounds good to you then that is what matters. Besides it will give you a connection to your equipment that most will never have. 

I've always been satisfied upgrading crossover capacitors, resistors, inductors without changing the design/values.  Generally you get what you pay for, that's my experience.  

It's a tricky business no doubt because confirmation bias is a thing.

However, I was once asked to solder a friend's Celestion 44 cable connectors where they met the drive units inside the cabinet.

The existing tag connectors looked pitted and dull and after hardwiring the cables (after stripping and cleaning) the sound noticeably changed.

The tonal signature was still the same, but the pace and dynamics seemed to have been given a shot in the arm.

Even 6 months later, when listening to the the Beach Boys Sunflower double CD I could not escape the impression.

Thankfully, my friend Mitch was pleased with the results but I daresay there might be some who would prefer the previous more laidback nature of the speakers.

Just how they sounded when new, early 1970s would have been anyone's guess.

Anyway, I'd like to think all decent loudspeakers nowadays are either hardwired or use non tarnishing tags inside.

It can’t hurt you can always put old caps back in. Have you ever seen the crap parts in some of these expensive speakers. I’ve only done it to a couple pair of speaker and was a very nice improvement. New wire improves sound and outboard crossover made it easy if caps are bigger than original. Also makes it easy to play around with caps. I wonder if crossover not rattling around if cabinet helped. Modding both pair of speakers improved sound for me. It wasn’t subtle. 

Arguments about issues with sensory perception is problematic here, confirmation bias exists with all equipment, whether oem or modded,  it can always exist. To make blanket claim that all crossover modifications resulting in one hearing sound quality improvements are faulty because of confirmation bias is total nonsense. One could make that argument with every single listening impression posted on this or any forum. This would conform to argument that more expensive equipment is only superior due to confirmation bias, little or no correlation to better sound quality vs. the less expensive component. This is argument for not trusting one's own sensory perception. I could make the same argument in regard to the cheaper component, the value seeking consumer is biased towards the cheaper component.


Individual audio parts make their contribution to holistic sound quality of individual pieces of equipment. The savvy engineer or designer is aware of this and chooses his/her parts accordingly, you do see increasing use of specialized or upgraded parts in audio equipment. The naysayer will state its for marketing purposes, the listener will listen. The experienced listener remains ever mindful of bias, can also control for bias by return to previous configuration.

@mglik: A viewing of Danny Richie’s latest GR Research Tech Talk video on YouTube, in which he goes through his appraisal of the pair of Focal Aria 906 a customer sent him, will be informative.

It’s easy to find: go to YouTube, do a search for GR Research, and a row of his many videos will appear. The Focal Aria video is the most recent, and will be first in line.

Danny shows the Aria’s drivers and enclosure (he has already disassembled the speaker), it’s measured on-axis and off-axis frequency response, its spectral decay characteristics, and examines and describes the speaker’s crossover. He then explains the reasons for the Arias performance failings, and how he addresses them. He not only modifies the crossover to improve the Aria’s measured performance, but uses better parts in doing so. He explains the benefits in sound quality the better parts afford.

As with all Danny’s Tech Talk videos, it is a short (13:21) and sweet look into loudspeaker design and build. You will be VERY glad you watched it!


I agree with you. I love watching his videos. Very informative, and he backs up what he does by showing the improvements in the measured frequency response and spectral decay.