LSi9 Crossover Modification Project

ben62670

You can do it on any LSI. Any speaker really. I modded an LSIc 2 weeks ago. Same thing. I feel 90% of the SQ increase is in the tweeter. Like DK and I have stated even though it reads the same SPL it seems louder, and clearer. I only charge Polkies $80 for a pair of XO upgrades. You buy the parts. Ship them, and the XO's to me. I ship them back. Done deal. It takes an easy 4 plus hours, and having to go out and ship them, but I do enjoy doing it.
Ben

LessisNevermore

ben, kudos on the project. I've been reading with a lot of interest, and hope at some point, to do this, or have it done to my 9's. Have you done any 7's? Similar results?

Good work.

ben62670

LessisNevermore wrote: »

ben, kudos on the project. I've been reading with a lot of interest, and hope at some point, to do this, or have it done to my 9's. Have you done any 7's? Similar results?

Good work.

Same thing;) a couple different values. The 7's should see the same if not best results(and cheapest) seeing they are 2 ways with no monster caps in them:)
Thanks
Ben
PS this is DK's thread;)

Early B.

DK --

I'm a major proponent of doing mods and I'm glad your project exceeded your expectations. I'm just wondering about the "opportunity cost." Let's say you sold your Lsi 9's for $500 and added the $343 you spent on parts -- would you have been able to buy a higher quality used pair of speakers and achieved a similar or better result?

I'm asking only to help those who wish to undergo this mod to consider a spending threshold. Would you have been willing to spend more money for the mod given the outcome you achieved? If so, how much more?

Face

Early B. wrote: »

I'm a major proponent of doing mods and I'm glad your project exceeded your expectations. I'm just wondering about the "opportunity cost." Let's say you sold your Lsi 9's for $500 and added the $343 you spent on parts -- would you have been able to buy a higher quality used pair of speakers and achieved a similar or better result?

I know it's not aimed towards me, but I wanted to add something. I don't believe you would be able to find something better for the $843 spent due to speakers in that price range will also come with similar quality drivers and mediocre crossover components. Between drivers, cabinets, and crossover components, I've spent about $3k on my Tannoys, and still haven't heard anything in that price range to make me want to replace them.

strider

As always, great write up DK.

Any thoughts/issues about the internal cabinet volume being decreased by the marked increase in the size of the crossover components?

mmadden28

strider wrote: »

As always, great write up DK.

Any thoughts/issues about the internal cabinet volume being decreased by the marked increase in the size of the crossover components?

I was just thinking the same thing. I'm in the middle of doing this same upgrade, (slowly but surely) and have been toying with putting all the big caps on the bottom. In my case I wasn't aware of the custom 18uFs avail from Soniccraft until after my order shipped, so I actually have 3 large caps to tie to the back of the board, instead of 1 - 18uf, I have a 10uF and an 8.2uF. .

mmadden28

While we're on this subject, is it possible to record via Mic, the before and after output? And be able to hear it on playback or see it in a chart after analysis?
Right now I am going to do Left Right comparisons (left unmodded, right modded with mono source using balance controls). But I was wondering if you could do the same by simply recording the output. Or would one have to have some muy expensivo recording studio quality stuff?

ben62670

The actual volume of the caps is not going to make a big enough difference to throw the specs out of whack. If you feel like figuring the volume of the caps out use an online calculator. You can divide square inches by 1728 to get cubic feet.

strider

Nah, I for one am not that concerned. More of a passing thought when I saw the size difference in the components. Seems like every time someone asks about adding bracing to their SDA cabinets, the standard response is don't do it because it will skew the internal volume of the cabinet.

Early B.

Face wrote: »

I know it's not aimed towards me, but I wanted to add something. I don't believe you would be able to find something better for the $843 spent due to speakers in that price range will also come with similar quality drivers and mediocre crossover components.

I meant used price. You can often find speakers that retailed for $2,000 or more on the used market for less than a grand.

LessisNevermore

ben62670 wrote: »

PS this is DK's thread;)

Yes, of course. No slight intended.:o

ben62670

anonymouse wrote: »

How do they sound after burn in?

As close to an active XO as any XO I have modded.
Ben

DarqueKnight

Early B. wrote: »

I'm just wondering about the "opportunity cost." Let's say you sold your Lsi 9's for $500 and added the $343 you spent on parts -- would you have been able to buy a higher quality used pair of speakers and achieved a similar or better result?

With regard to just sound quality, maybe, but there were other considerations. These LSi9's are used in my home theater system as dual center channel speakers, so I had to consider:

1. Timbre matching my SDA SRS front speakers.
2. Timbre matching my LSi15 surround speakers.
3. I wanted to maintain a matching 4 ohm load all around (front-center-surrounds).
4. Fitting into the existing spaces in the equipment cabinet.

Early B. wrote: »

I'm asking only to help those who wish to undergo this mod to consider a spending threshold. Would you have been willing to spend more money for the mod given the outcome you achieved? If so, how much more?

To digress a bit:

I would have never considered this mod if the LSi9's weren't already in my HT system. I had already totally dismissed the entire LSi series for music applications.

I tried a pair of LSi7's and a pair of LSi9's in my office rig at work and I was not satisfied with the sound. The bass was either boomy or not coherent with the midrange and treble. I also found them lacking in clarity. A pair of CRS+'s worked just fine.

I tried a pair of LSi7's and a pair of LSi9's in my home office rig and I was not satisfied with the sound. The bass was either non-existent or not coherent with the midrange and treble. I also found them lacking in clarity. A pair of CRS+'s worked just fine.

I tried a pair of LSi15's in my master bedroom A/V system and I was not satisfied with the sound. I could not blend them properly with my SVS SB12 Plus subwoofer. I also found them lacking in clarity. A pair of Magnepan MG12's worked just fine, and they blended seamlessy with my subwoofer.

Now that I have experienced modified LSi sound, I would consider them for reproducing music, but, as you suggest, I would now have to compare them against other alternatives. It may be the case that there are better speakers for the same as, or even less than, the cost of modded LSi's. Of course, we won't know until someone goes to the trouble of doing a shootout.:)

Now, to answer your question:

Early B. wrote: »

Would you have been willing to spend more money for the mod given the outcome you achieved? If so, how much more?

Let me put it this way, I spent 300 bucks more and received imaging, layering, depth, overall clarity, and bass articulation and tactility that I associate with "bookshelf" speakers in the $2000+ (MSRP) price range...and this was while driving them with the so-so electronics in my HT system.;)

DarqueKnight

anonymouse wrote: »

Thanks. DK - what's your assessment?

I have never even stood next to, nor heard, speakers interfaced with active crossovers.

DarqueKnight

strider wrote: »

Any thoughts/issues about the internal cabinet volume being decreased by the marked increase in the size of the crossover components?

I briefly wondered about that myself, but when I opened the cabinets, I saw that they were already full of that fiberfill stuffing material, so the larger caps are just taking the place of some of that.

Quantitatively, the LSi9 cabinet has an interior volume of 1191 cubic inches. The two large 130 uF Solen Caps plus the two 18 uF Sonicaps take up 32.44 cubic inches of cabinet space. This is only 2.7% of the total volume. I don't think it matters, but someone more experienced in speaker building might want to chime in. The new caps probably take up less space than the drivers and tweeter, but perhaps the designers already accounted for that.

The 1 uF and 12 uF Sonicaps do not impinge upon the LSi9's cabinet interior because they are contained within the volume of the crossover assembly.

McLoki

for my lsic I got aeon caps from sonicraft. Two 100's and 1 60.

They are a solen clone. If the sale was not going on at sonicraft though, I would have gotten daytons.

ben62670

anonymouse wrote: »

So I ordered the Sonicaps for the tweeters.
Wondering what to order for the 260uFs. Any opinion on whether I should go with Solen's or Dayton's? Solen actually makes a 270uF cap http://www.hndme.com/productcart/pc/viewPrd.asp?idcategory=111&idproduct=935 which is probably close enough to 260uF given the tolerance of the original caps. Any opinions on whether that would be a good idea? It would be so much neater to have a single cap for that position, plus I read somewhere that caps in parallel can create artifacts.

If you can get the Solen 270uf it is more than close enough to spec;)

Systems

anonymouse wrote: »

So I ordered the Sonicaps for the tweeters.
Wondering what to order for the 260uFs.

I know everyone on this site will likely disagree but IMO being as that 260uf cap is shunt connected acrosss the woofer it is the least important part as far upgrading is concerned.

It is not in the signal path so it will have very little sonic signature.(but the value needs to be correct). Thererfore I wouldn't spend big bucks on it, as even a paralled combo of a 250uf Bennic electrolytic and a Bennic 10uf poly or mylar should suffice.;)You'll save some money and lots of space.

ben62670

GV#27 wrote: »

I know everyone on this site will likely disagree but IMO being as that 260uf cap is shunt connected acrosss the woofer it is the least important part as far upgrading is concerned.

It is not in the signal path so it will have very little sonic signature.(but the value needs to be correct). Thererfore I wouldn't spend big bucks on it, as even a paralled combo of a 250uf Bennic electrolytic and a Bennic 10uf poly or mylar should suffice.;)You'll save some money and lots of space.

I won't disagree. I wouldn't change it for myself. I do present the option of leaving it alone when people contact me about doing them. The deal is most of the people doing the mods just don't want to have in the back of their minds "should I have changed that?". I have noticed better bass when changing big caps like that, but the mids, and highs are really where it is at with the mod. I know that you do see a lot of comments on the bass actually improving with the mods, but I feel some of this is do to bass extension, and the fact that musical instruments are not just 80hz, but can have a very wide frequency range. When I did just the tweeter mod on my SDA's the whole mid spectrum improved, and my tweeters are crossed over at 2300hz. Just keep in mind about half the parts cost is the big caps, and they are not in the signal path, but used to bleed off at the XO point to make it a 12db slope.
Ben

Systems

ben62670 wrote: »

Just keep in mind about half the parts cost is the big caps, and they are not in the signal path...

And with some of money (and space)saved I would replace the 1.2mh steel laminate inductor with an 16 awg air core type.I would expect this change to improve the mids as the air core will not have a tendancy to saturate at hi levels the way steel laminates can.A coil that is perodically being saturated equals distortion and reduced resolution.

ben62670

GV#27 wrote: »

And with some of money (and space)saved I would replace the 1.2mh steel laminate inductor with an 16 awg air core type.I would expect this change to improve the mids as the air core will not have a tendancy to saturate at hi levels the way steel laminates can.A coil that is perodically being saturated equals distortion and reduced resolution.

Again I agree. I am a fan of using an air core in anything above say 300hz. If I were to do one for myself I would build my own board out of wood, and lay it out point to point like Face has done on some of his work for himself, and like some custom ones I have done for myself. I like point to point much more than a circuit board.
Ben

Systems

ben62670 wrote: »

If I were to do one for myself I would build my own board out of wood, and lay it out point to point....

Yes when parts count and size get on the big side thats what to do.Im just using a piece of masonite for the xovers in a pair of mini monitors Im building.

ben62670

I would love to see a shoot out between LSI9's modded a few different ways.
A: all Dayton no change to the 260uf
B: all Sonicaps no change to the 260uf
C: all Sonicaps no change to the 260uf new inductor
all Sonicaps 260uf worth of Dayton big fatties, and new inductor
The only place I see this possible is the NJ/NY area. I would love to hear the comments from a group of people doing a blind test not knowing what was inside;)

Systems

ben62670 wrote: »

I would love to see a shoot out between LSI9's modded a few different ways.
A: all Dayton no change to the 260uf
B: all Sonicaps no change to the 260uf
C: all Sonicaps no change to the 260uf new inductor
all Sonicaps 260uf worth of Dayton big fatties, and new inductor
The only place I see this possible is the NJ/NY area. I would love to hear the comments from a group of people doing a blind test not knowing what was inside;)

You forgot one .;)

E: Biamped and actively crossed over at 2k with 24db slopes.

All passive parts removed except 9mh coil and 260uf cap on the woofer,and a big poly cap in series with the tweet for DC protection.

ben62670

GV#27 wrote: »

You forgot one .;)

E: Biamped and actively crossed over at 2k with 24db slopes.

All passive parts removed except 9mh coil and 260uf cap on the woofer,and a big poly cap in series with the tweet for DC protection.

Remember our phone conversation about my tweeters in my customs blowing, and you suggested I use a cap to block any sort of DC. Well. This came as a total surprise to me. I had a pair of Dayton 18uf caps laying around so I put them in series with the tweeters. Now I have a 30db slope 24 active, and 6 passive. It sounds much better with the cap in there. I had tried this before with a buddies Sonicaps, but they were brand new. It sounded worse. I attribute that to the Sonicaps needing more time to burn in than the Dayton's. A few things did change when adding the caps. The phase has shifted, and the slope is different. The 18uf cap comes into play at a much lower XO point than the active XO. Now the sonic differences. The highs like cymbals has more sustain. It wasn't artificial sounding. It was actually much more lifelike than before. Before it was a little thin. Also vocals were much more life like, and mellow. The OpAmps used in the Active XO, and the pre were all BB2134's. Not the best, but better than what you find in most gear. I do have some nice inductors laying around to see if they may boost the performance in the <2300hz range. I really doubt it, but I also doubted sticking a fairly cheap cap in the highpass of an active XO system would actually improve the SQ.


DK if we should start a new thread let me know. I feel like this info pertains to the project you posted, but it has taken a side road.
Ben

mmadden28

GV#27 wrote: »

And with some of money (and space)saved I would replace the 1.2mh steel laminate inductor with an 16 awg air core type.I would expect this change to improve the mids as the air core will not have a tendancy to saturate at hi levels the way steel laminates can.A coil that is perodically being saturated equals distortion and reduced resolution.

When I finally get around to finish mine and do my comparisons, I may be inclined to try the inductor replacement out. Any suggestions on a particular inductor or source?

Systems

mmadden28 wrote: »

When I finally get around to finish mine and do my comparisons, I may be inclined to try the inductor replacement out. Any suggestions on a particular inductor or source?

Here is one possibility http://www.madisound.com/catalog/product_info.php?manufacturers_id=139&products_id=983 Ideally you want the inductance values and DC resistance to be the same as the stock coils, but the extra .05mh of inductance and .07 ohms of DC resistance vs stock will not cause response errors.


Here is one from Solen that is 1.2mh witha DCR of .35 ohms.
http://www.solen.ca/pub/cms_nf_catalogue_fiche.php?id=905&recherche=&numRows=&manufacturiers=&niveau1=1&niveau2=4&niveau3=71


And Parts Express http://www.parts-express.com/pe/pshowdetl.cfm?PartNumber=255-424&DID=7


Ben, its a bit OT so I will respond to your post via PM .

ben62670

Which caps did you use on the highs?
BTW I was driving them with Adcom GFA-565's

Edit the 260uf is a shunt cap so bypassing may not be very beneficial. Can you disable the bypass cap for haha's? How much time do you have on them?
Ben

ben62670

It sounds like there may be some sort of a limiting factor like speaker cables, source, or IC's. Your amp certainly is a good amp. What other gear is associated with this test?
Thanks
Ben