LSi25 Subwoofer Upgrade

sk88

I replaced my LSi25 subwoofer with Polk MM1040UM and listened to it for the first time last night. I can say that I am very happy with the results. I listened to The Battle of Kerak and I felt vibration from the floor and sofa where I sit. Then I listen to some pop beats and it really filled where the upper drivers left off with solid bass.

This is just my first impression and it is only from one speaker as I've not done the other one. In my plan I have prepared that I might have to modify the stock port to 4" x 26" (more data about this at a later time) but it seems I may just be satisfied without that extra work.

I need to do more listening as it requires 20 to 30 hours of break-in according to Polk and re-adjust my settings when I have my other speaker done. There are other detail information I will provide but just wanted to share this quick news for now.

ZLTFUL

Interested in your findings.
I was thinking of doing the DB1040 swap as the 15s take the DB840. The DB requires little port modification from what I have read.

Anyway, hope to read more as you finish the second one and port mods.

sk88

Hi zltful,
I am away from home for about 2 weeks. Once I am back I can provide the driver and enclosure measurements so you can check if a DB1040 can physically fit or not. In my case, MM1040UM can fit but not without enclosure mod. I also simulated driver performance to make sure it would work sonically. Again, I can run a simulation for DB1040 if you want me to when I am back home.

ZLTFUL

That would be great. They seem to run around the same price so that's not a concern.
Thanks.

sk88

Zltful, the interior Lsi25 width is 5.5" but DB1040 has a 5 7/16" top mount depth which is not going to fit well.

sk88

I finished the sub upgrade on my 2nd speaker. This upgrade takes about 2 hours for the work but I can say the difference is significant and totally worthwhile. The bass becomes much tighter, stronger and integrates well sonically with the other drivers. I chose not to run additional subs in the past but for some type of music I wished I did. With this driver upgrade, I don't think I'll need additional sub anymore.

Old vs. New

Stock subwoofer
diameter 10 1/16"
cutout hole diameter 9 3/16"
depth (bottom mount) 5 3/16"
4 screws

MM1040UM diameter 10 7/16"
cutout hole diameter 9"
depth (bottom mount) 5 3/16"
8 screws

You can see that the two sub have the same mounting depth which is very important due to the way the sub needs to be mounted on the side of the speaker. Any other sub with longer depth would either not fit or have issues of not allowing adequate air flow. The existing cutout hole is a tiny bigger than what's required by MM1040UM which works well without any change. However, the diameter of MM1040UM is a little bigger than the stock - 3/8" to be exact. For this small difference, the recess ring where the sub would sit on needs to be widened. To do this, you'll have to use a router, and a plunge router would work better than a fixed router in this case. You will also need a circular router jig to make a circular cut.

I made my own circular jig with a piece of acrylic. I can drill a hole anywhere on the jig for the radius I need. You can also buy one with many pre-drilled holes. To widen the recess ring for 3/8" more in diameter, I used a 1/4" router bit with a radius of 5 3/32"

To use the jig, you must have a center hole where you can pin down the router. With the hole already cut on the speaker, this is a problem. So, what I did was to glue a temporary board on the back of the hole to close the hole and waited for it to dry. I used silicone adhesive so that the board and the adhesive can be removed easily and cleanly afterwards. You don't want to use Gorila Glue or alike for this. Then I made a 3/4" thick and 9 1/4" diameter round MDF (using router) to sit in the hole. Now I have a center point. Make sure it sits tightly without room to move (the center point). If it does, you can tape around the edge to make it bigger until it sits tightly. I then glue another layer of 3/4" MDF square to sit on the top. It's very important to ensure the center point is maintained on the square. Now this 2nd layer is flush with the speaker cabinet and ready for the router work.

To prevent the cabinet getting scratched by the jig, I taped the surrounding areas. But, I still made some scratches. Oh, well, I am not selling this speaker so - so be it. It's better to cover up all the surrounding area.

Now that the recess is done, the temporary board can be removed and start to mound the new sub. Surprisingly or not, the 4 screw holes from the stock sub fits precisely to the new sub, so you'll just drill 4 new holes. It's that simple. I'd suggest to tighten all 8 screws because this new sub is powerful and really can move.

sk88

About this sub replacement, without any port modification, the frequency response is tuned to 42.5Hz. This is a enjoyable frequency that most speakers do for home theater and general music listening which has more thump. There is no data I can find for the original stock subwoofer, but I wouldn't be surprised if it is tuned to the same result by knowing how the rest of the LSi drivers sound.

However, if a true flat bass response is desired for critical listening, then the port would need to be modified. Changing the port diameter would be very difficult due to the existing Power Port at the bottom of the speaker and that the interior space is only 5.5" wide. My recommendation would be just to extend the length of the port tube and keeping the same diameter. According to WinISD simulation, extend the tube to 20inches would result a tuned 30Hz frequency for a fairly flat response above that. Although MM1040 has a frequency response down to 24Hz based on the spec, 30Hz would be a good realistic frequency. Rarely music really go down to that low anyway. There is a bit of concern in port air velocity as normally it should be kept under 55ft/sec, and making the port wider (such as 4" diameter) would decrease that concern. However, it should be fine as long as you are not listening at a very loud volume (who can hear that "chuffing" when the overall volume is at that level). If anyone tries to modify the port length or diameter, please share your result with us.

Initially, I didn't think it is possible to extend the tube as the sub section of the cabinet is only 26" tall. Then totally by accident I looked up inside the cabinet, I found that there is a long and narrow chamber that goes from the sub section all the way up to the top of the cabinet behind the MTW section (see picture). So it can be extended very long as long as you want to make the effort. I am sure there is somewhere that sells 3.5" PVC pipes and fittings. I've not done this myself because I am satisfied without modifying the port. I can only imagine that it can be done by cutting down the existing tube somewhat and then use different angle elbows to extend the tube around the subwoofer (because the tube is right underneath it) and then goes up straight. One must also check into that there is enough space between the amp and the subwoofer.

Although it may appear to be a concern that MM1040 is capable of a 350W continuous rating and the LSi25 amp output is only 150W (300W peak). In reality, it's only a 3dB difference. For most ears, 3dB is the threshold of human hearing, so I don't think there is going to be a noticeable difference in volume if you want to play it loud. We must also consider the coherence to the other drivers, so I don't see a need nor recommend to change anything to the amp section if that's even possible.

LSi25 subwoofer enclosure
lower chamber 13.5" D x 5.5" W x 26" H (volume 1.12 ft3 including port)
upper chamber 4.75" D x 5.5" W x 18,5" H (volume 0.28 ft3)
round port 3.5" diameter x 8" length (volume 0.045 ft3)
Amp 150W RMS, 300W Dynamic, 4 Ohm

Polk MM1040UM SVC
Fs 35Hz
Xmas 1"
LE 3.0 mH
RE 3.25 ohm
SD 54.25 ft2
Vas 0.65 ft3
Qts 0.45
Qes 0.529
Qms 3.0
Power Handling 350W Continuous, 700W Peak
Nominal Impedance 4 ohms
Efficiency 91 dB

In the pictures, green lines are w/o the tube extension (8"), white lines are with tube extension (20").

mpitogo

Nice job. Did you ever extend the tube?

sk88

Wow, it's been almost 4 years. No, I haven't extended the tube. Wish I did as now I doubt I have the same energy that I used to have.

pineomt

Good info on this thread. I thought I'd share what I know.

First, some background. I've been watching these speakers on Ebay for years, and just finally pulled the trigger on (what I thought was) a mint condition set. However, once I got it setup, I realized one of the internal amps is was blown.

I was on the phone with Polk customer service twice today. The first time I was told that I was essentially SOL. So I started thinking about potential workarounds. The most obvious (and simplest) is to bypass the internal amp, via an external amp with signal from the LFE output on the back of my AVR. To do this, you need to disconnect the woofer from the internal amp, disconnect the internal amp from the bottom binding posts, splice-in connectors on the woofer that are compatible with the binding posts, and connect the woofer directly to the binding posts. Next I'd need to connect a 4-ohm stable amp (as the woofer is 4-ohm nominal) to the lower binding posts on the back of the speaker. The amp would be fed information from the LFE out on the back of the amp. This would result is similar be to the 3rd connection method as described in the instruction manual, whereby the the passive drivers are left/right channels and the woofer is setup as a true subwoofer.

The positives of this solution include no need to actively manage a crossover as AVR/Audessey will take care of this via bass management setup. From my opinion the negative is you no longer have the option of running a full range speaker (matters for 2-channel listening, especially if you use PureDirect or a comparable 2-channel only listening mode that turns off all crossovers and bass management).

The second option is essentially the same as the first, but instead of rely on the LFE out to send information to the amp, I'd incorporate a low pass filter via an active crossover. The advantages would be that you could still run the speaker as full range, which I think is optimal for 2-channel listening, and the you don't have to incorporate 2 additional subwoofers into your room. It's worth noting that most modern mid to high level AVRs include 2 LFE outs and each is EQ'd independently for room correction purposes. When you add 4 subwoofers into the mix that becomes more difficult (especially considering you have mismatched subs - e.g. multiple subs of different make).

All that being said - I called Polk back a second time, and spoke with a customer service rep with strong knowledge of these speakers. He said that although the amps are no longer made, Polk would still try to repair for me, so I'm sending the amp in to see what, if anything, Polk can do.

While on the phone with the rep, I took the time to ask about the details of the speaker that I thought I'd share with you:

- The passive drivers (midrange and tweeter) have a 90hz high pass filter.
- The passive drivers are not effected by any tweaking to the crossover on the back of the amp - meaning the high pass filter is static - does not change - stuck at 90hz.
- The crossover on the back of the amp is actually better described as a variable low pass filter. However, since the passive drivers roll off at 90hz, it really doesn't make sense to turn that knob off the factor setting (which the rep confirmed was 90hz). Lowering that knob below 90hz will cause loss of information between the active and passive speakers, and raising above 90hz is redundant, and results in a a "boomy" sound.

sk88

I haven't look at the crossover for a long time, but the amp setting on the back of lsi25 has nothing to do with the three drivers on the front. It's only for the subwoofer on the side.

pineomt

That's correct. The passive drivers are on a HPF that rolls off the information at about 90hz. The crossover "factory setting" on the amp is 90hz. In fact, calling it a crossover is actually sort of misleading; it's best described as a variable low pass filter.

I don't think it makes sense to run these speakers as full range and rolling off the entire speaker to a dedicated subwoofer at 80hz (THX recommended) - you'd barely utilize the woofer. Given that, I see two viable options for integration:

1. Put woofer on LFE track - which requires some additional subwoofer EQ if you're like me and have a dedicated sub. I suppose this is probably the way t go if you're mostly watching movies, where the audio mixer developed a separate LFE channel.

2. Use left/right pre-outs and run the speakers full range (set to "Large" in AVR). Probably the better way to go if you like to listen to two channel music in Pure Direct (on Denon AVRs), or another comparable audio setting that optimizes two channel music. For my setup, I'd have to run a low pass filter before sending the signal to the external amps, so this will take some extra integration with an SPL meter, mic and a way to implement the EQ/crossover.

I envision wanting to use the speakers to play full-range 2 channel audio as well as surround sound mixes, so I'm going to integrate an A/B switch to hop back and forth between LFE and L/R signals easily.

pineomt

Do you have any idea what the woofer's LPF's specs are (slope, etc.)?

K_M

Not sure what you are explaining exactly, but to clarify, are you planning on using the Lsi25 AND another sub or two, or JUST the Lsi25's?

pineomt

No - have two dedicated subs, I will run those off the LFE output from AVR as a true sub.

Here's what I'm trying to do: The amp on one of my 25s is shot. I spoke with Polk, and they "might be able to repair it." However, the cost, just to ship to and from the repair site is almost 170 dollars, and again, they may not even be able to repair (no longer support/make parts for speaker). Considering that these amps have a reputation of being delicate and prone to failure I started thinking about potential workarounds.

A good reliable external amp makes a lot of sense. However, if I wanted to run these speakers as full range (as opposed to a tower with a built in true passive-subwoofer), I need to recreate the stock amp's integrated low pass filter by using an active crossover (Minidsp.com, for example). To do so, and to keep these things as close to the stock sound as possible, it would be good to know the details of the woofer's integrated variable low pass filter.

Alternatively, I could just manually EQ this thing with a mic, some room correct software to a point where the transition between the passive drivers and woofer is smooth.

I welcome any thoughts the community has as I'm still trying to work all this out in my head.

K_M

Gotcha!
So one still works fine as intended, you could use that as a reference point. Pretty sure the built in amps employ EQ boost.

Dayton makes several stand alone amps, that have EQ built in or variable. Most likely there is a 6db boost at 35 hz in the built in Lsi25 amps or something very similar.

The one odd thing, your "dedicated subs" will be in essence crossing over to another amplified subwoofer at some frequency, and then again the built in subs, to the front mounted speakers again.

What are the dedicated subs you are using?

I only ask because the built in subs in the Lsi25 are fairly decent as they are. Not great, but it may be a case of similar enough to not really need one or the other.

pineomt

So I was thinking of going with a Emotiva BasX A-300 to power the woofer, and eventually getting the same for the passive drivers on the top.

With regard to the dedicated sub question - I was thinking of running the 25's on LARGE setting within my AVR, so would not be employing any bass management on those or rolling anything off to the dedicated subs.

I have two Klipsch RW-12d, which are fed the LFE pre-out from the AVR.

pineomt

I guess fundamentally, I'm asking is for advice on (quasi) bi-amping and tuning these speakers.

My thought is I could use a MiniDSP 2x4 unit to EQ/integrate the woofer back into the broken speaker. Using a program like REW, I could use the EQ curve of the good speaker to help me model the LPH on the broken speaker (once I am happy with the sub integration on the broken speaker, I'd make the same upgrade to the good speaker).

My current thinking as to how to do this:

1. setting crossover - passive drivers have HPF at 80hz at 12db. I was thinking of rolling off to the sub at 18db and seeing how that looks on the REW graph compared to the stock sub. That said, this all gets a little funky when you start thinking about how these speakers interact with bass management from an AVR. The passive portion of the speaker is connected to the FL/FR speaker level out. Audyssey recommends setting these to SMALL and correctly sets the filter to 80hz. But in doing so, there are now two 80hz HPFs on these speakers; that cannot be optimal. I think better is to set to passive portion of the system to LARGE and allow passive HPF to roll off the signal. Also, consider further that if you connect the LSi 25 woofer to the LFE out on the AVR, you're bypassing whatever LPF Polk designed for the woofer...And so I start wondering then how sacred can the specs of the LPF really be if Polk intended the user to bypass the LPF altogether (in the manual, one of the suggested setups include using AVR's LFE signal to subwoofer). So I think it will come down to tinkering with the crossover in the MiniDSP application until I get a reasonably flat 80hz crossover. That said, I'm happy to hear suggestions about how to set that LPF. Again, ideally I'd like to run these speakers as full range using Left/Right signal as opposed to using LFE signal.

2. setting gain - Polk allows the user to set the gain on the sub (though they have no suggested factory setting), so presumably Polk had no hard and fast rule about subwoofer gain. I think I can dial in the gain using an SPL or some EQ software (like REW). Having never done this manually, I'm looking for advice on how do practically do this. My thought would be to use an SPL meter and to push signals at various frequencies through the crossover range. However, I'm trying to understand how to identify modes/nodes vs gain mismatch. I suppose if I set the mic a few feet away from the speaker (as opposed to seating position) I might mitigate potential modes/nodes, but curious to hear other's thoughts on this.

3. delay - this one is really tricky in my mind because I have no way of knowing if Polk setup some sort of delay in the stock circuitry. What little reading I've done here says it's more important for car audio where tweeter and woofer can be really far apart, but I've also read how speaker designers may create an angled front baffle such that the tweeter is further away from the listener than the mids, and further still from the woofer. Not even sure it's material, so I guess I won't attempt to adjust this unless I can see a compelling reason to do so. Even then I have no idea how to measure/account for any adjustment here.

Thanks to the forum for acting as a sound board for my ideas.