Hivi 3.1A DIY Speaker With Sehlin Mod Review

[ayane]

It's also worth noting that an advantage of the bass alignment HiVi chose is that power handling is very good. Trying to extend the bass response will negatively impact that somewhat.

This is a very good point, and it's part of the reason why I'll want to add a sub later on down the line to take over from that mid-bass region and below. I still think it's worth re-tuning the cabinet, because that "bump" at 100 Hz these have before any mods is characteristic of speakers with the wrong cabinet size (i.e. too small of an enclosure) and/or indicative of an improperly configured port.

Extending it down to 50 Hz would require about 3-4 dB of additional power handling, but the 60 watt nominal power rating of the woofer leaves plenty of headroom before the excursion becomes the limiting factor. Needless to say, the THD will become more of a problem before power handling, and the Loxjie A30 chokes up before driving these speakers to distortion anyway.

I'm not sure if it was clear, but what Amir reviewed was "Mod 4", which some still found a little bright.

Yup! I haven't heard mod 4, but mod 5 sounds just fine compared to stock in the mids and treble, maybe with still just a tinge too much energy in that top treble area listening to "Sonic After-6290 Mix" compared to better monitors. Then again, I've got sensitive ears, so maybe that was just me and mod 5 is perfect. I really want to keep that response from the mids up, just want to fix what I'd call "kitchen speaker" bass.

My recommendation would be to set yourself up so that you can revert to Mod 4 to Mod 6 in case the improvement efforts don't go according to plan.

That's the plan (although I chose mod 5). One is re-assembled with mod 5, and the other one is the WIP

Maybe I should buy a second pair so that I can have one mod5 pair and one DIY pair, haha

 

[Billy Budapest]

I'm not sure if it was clear, but what Amir reviewed was "Mod 4", which some still found a little bright. The crossover from the stock kit is a lot brighter. "Shouty and glassy would be consistent with my recollection. My pair started their life out that way.

If you are interested in doing your own design, I would applaud that. I am sure others will be interested in what you come up with. My recommendation would be to set yourself up so that you can revert to Mod 4 to Mod 6 in case the improvement efforts don't go according to plan. It's also worth noting that an advantage of the bass alignment HiVi chose is that power handling is very good. Trying to extend the bass response will negatively impact that somewhat.

I read the thread on the other forum and do not understand what mods 4-6 are. I only saw three mods discussed.

 

[Mudjock]

Mod 4 is the "perfectionist mod" that was in the speaker Amir tested. There were a couple of questions/suggestions in this thread about areas that could be further improved by changing out some of the larger and more expensive components in the crossover, so I revisited my measurement and simulation files and came up with some further modifications one could try. "Mod 5" was discussed on page 4 of this thread Changes were to C5, R2 and R1 to pull the mids and highs down a little more. Here is some info.

Mod 6 is a change to the woofer circuit to slightly bump up the region from 200-400 Hz. The difference is pretty small (shown on page 10), so it isn't clear it will be easily audible. The changes are reducing L4 from 1.5 to 1.2 mH and raising C5 from 68 uF to 100 uF.

 

[Billy Budapest]

Gotcha! No wonder why this thread was confusing me—I did not catch that new mods were being developed as the thread continued.

 

[ayane]

Mod 4 is the "perfectionist mod" that was in the speaker Amir tested. There were a couple of questions/suggestions in this thread about areas that could be further improved by changing out some of the larger and more expensive components in the crossover, so I revisited my measurement and simulation files and came up with some further modifications one could try. "Mod 5" was discussed on page 4 of this thread Changes were to C5, R2 and R1 to pull the mids and highs down a little more. Here is some info.

Mod 6 is a change to the woofer circuit to slightly bump up the region from 200-400 Hz. The difference is pretty small (shown on page 10), so it isn't clear it will be easily audible. The changes are reducing L4 from 1.5 to 1.2 mH and raising C5 from 68 uF to 100 uF.

This is a nitpick, but the polarity of S3 is flipped in all of your schematics. Not important if using HiVi's stock board because they grounded all of the drivers' [-] terminals, but it was bothering me and no one seemed to bring it up, haha.

I think you should also post mods 5 & 6 to your site! Maybe call them "super-perfectionist" and "splitting hairs" mods

 

[Nwickliff]

Hey there everyone. Got everything hooked up but something is going on with the tweeter. I think it might be the last cap in line before the positive tweeter connection. It wasn't working when I got it all put together. I pulled it apart and unsoldered the woofer and tweeter connection because both weren't playing. I think I might have had too little wire when pulling off the covers. Re-soldered but tweeter was intermittent. If I pushed on the board it was kind of coming in and out, which was weird. I took the board out and now nothing.

The tweeter is fine but I think the cap might be gone right before the tweeter. Hopefully, this video explains what I mean. Basically, when I keep the positive wire from the tweeter in its place and connect the negative/ground wire to any of the positive points on the board I get sound out of the tweeter. Does this mean the cap is blown and only acting as a ground instead of a positive connection? I really don't understand crossovers so I'm a bit lost.

https://vimeo.com/667029369/28fed3f291

I'm thinking I'll order another cap and see how that goes. Wonder if I should just do point to point instead of the board. I get all kinds of confused when reading the schematics of the wiring trying to do point to point.

 

[ayane]

Hey there everyone. Got everything hooked up but something is going on with the tweeter. I think it might be the last cap in line before the positive tweeter connection. It wasn't working when I got it all put together. I pulled it apart and unsoldered the woofer and tweeter connection because both weren't playing. I think I might have had too little wire when pulling off the covers. Re-soldered but tweeter was intermittent. If I pushed on the board it was kind of coming in and out, which was weird. I took the board out and now nothing.

The tweeter is fine but I think the cap might be gone right before the tweeter. Hopefully, this video explains what I mean. Basically, when I keep the positive wire from the tweeter in its place and connect the negative/ground wire to any of the positive points on the board I get sound out of the tweeter. Does this mean the cap is blown and only acting as a ground instead of a positive connection? I really don't understand crossovers so I'm a bit lost.

https://vimeo.com/667029369/28fed3f291

Be careful connecting the tweeter across random points on the board! I worry that it's pretty sensitive and can end up getting cooked.

It's hard to tell if your cap is blown, but I don't think C1 is the problem. These foil capacitors tend not to become dead shorts when they fail; their capacitance tends to drop until they become completely open. For that reason, I find it hard to believe that it's acting as ground. (If the cap has failed to be open, then connecting the tweeter to some other point on the circuit shouldn't be able to make it work, especially at audio voltages/frequencies.) In either case, you can check across it with a continuity tester.

If your suspicion is right and your cap has failed to a short, then the voltage across your tweeter is around the same as the voltage across L1 and ground - however, at higher frequencies, there should still be significant voltage across the tweeter, enough to make it sound out assuming your other components are working. That seems to imply that L1 is shorting to ground, too.

I think you should check the hot side of the tweeter's filter circuit; I suspect some of the components there may have failed instead. Alternatively, make sure your connections are nice and connective. The leads may need a bit of sanding before soldering, especially on the inductors.

I'm thinking I'll order another cap and see how that goes. Wonder if I should just do point to point instead of the board. I get all kinds of confused when reading the schematics of the wiring trying to do point to point.

PM me; I have most of HiVi's stock crossover components and can mail them to you without extra cost.

 

[ayane]

If you are interested in doing your own design, I would applaud that. I am sure others will be interested in what you come up with. My recommendation would be to set yourself up so that you can revert to Mod 4 to Mod 6 in case the improvement efforts don't go according to plan. It's also worth noting that an advantage of the bass alignment HiVi chose is that power handling is very good. Trying to extend the bass response will negatively impact that somewhat.

So I ended up putting together the speakers with the back panel upside down, and I replaced the port with a crappy bent PVC pipe pointing up with about 3 cm clearance from the port opening to the top padding. The pipe's own resonance is proving to be a problem (lol, the mids are horrible because of the extra distortion) but I'm working on printing a custom port. I'll post an STL file when I finish. I also have to figure out the best material. Putting this project on the back burner for a few weeks, though.

I also ended up using the foam that came with the kit to pad the speakers. I padded above, below, and behind the terminal box, padded that 1" space under the woofer, and padded the sides, top, and bottom. The only area that's not padded is above the crossover and that vertical cavity between the crossover and top foam.

I don't have measurements, but I was able to keep the response flat from 400 Hz up to that 12.2 kHz peaking. I ended up using 7 ohms for R1 because I really wanted to tame that treble spike, and it added just a touch more flatness compared to mod 5. I used 3.5 ohms instead of 3 ohms for R2 and also used an 80 μF cap for C5 but kept L1 the same to pull down the upper mids a bit more. I'm calling it mod 5.5 because it's somewhat in-between mods 5 and 6.

The other good news is that I'm able to get these to play hilariously low. They're easily doing down to 42 Hz within 2-3 dB! It doesn't sound like a lot compared to 50 Hz, but every little bit counts in those bass frequencies. The peaking at 100 Hz is also gone, and they no longer have a tubby sound. If anything, they sound muddy/boxy/thick because of the port troubles, and I think with a better port material, they'll be perfect. With the right room (like 28 ft by 12 ft with the speaker across the shorter side), who knows - they might even go down below 40 Hz with room gain!

I wish I could measure distortion, but as far as I can tell, there is nothing additional compared to before. The only problem is the port vibrating at lower mids and adding its own distortion. When I come back to this project, I'll get a good measurement mic and post some amateur measurements.

Update: I secured the port to the board in the middle of the cabinet and added a piece of foam at that point. The vibration problem is completely solved, but I'm still going to print a more permanent/reliable rigid solution. They sound incredibly good, truly the best bang for buck at this price point.

 

[erasnick]

I have just built my Hivi 3.1A with the Perfectionist Mod option 4 however I have an amplifier that can bi-amp the speaker and the speaker does come with a binding post which is designed for this. Forgive my ignorance but if I bi-amp the speaker and run cable directly to the woofer in which case my amplifier will act as an active crossover and send only the lows to the woofer and the highs to the mid and tweeter can I leave the crossover as is and what considerations should I make if going for this? Also how would I calculate the impedance of the highs?

I'm surprised this doesn't seem to have been discussed much since it would seem to add more power to the woofer which seems to not be getting enough power in relationship to the mid and ribbon.

 

[ayane]

I have just built my Hivi 3.1A with the Perfectionist Mod option 4 however I have an amplifier that can bi-amp the speaker and the speaker does come with a binding post which is designed for this. Forgive my ignorance but if I bi-amp the speaker and run cable directly to the woofer in which case my amplifier will act as an active crossover and send only the lows to the woofer and the highs to the mid and tweeter can I leave the crossover as is and what considerations should I make if going for this? Also how would I calculate the impedance of the highs?

I'm surprised this doesn't seem to have been discussed much since it would seem to add more power to the woofer which seems to not be getting enough power in relationship to the mid and ribbon.

It hasn't been discussed because the L6-4R has an xmax of only 4.3mm and IEC power handling of 30 watts. Feeding it more power than that will cause it to blow out, and since the woofer of a speaker "gets" most of the power anyway (because it has to make such large excursions compared to the other drivers), it's bottlenecking the midrange and tweeter, not vice-versa. You won't be running it any louder considering the balance with the other drivers even if you bi-amp it.

I'd instead recommend adding a high pass filter at 80 Hz or higher to limit the woofer's excursion, and adding a proper subwoofer to take over from that region. The woofer can play a lot louder when it's not trying to produce output in the lower bass octaves, but you still should stay within 30 Wrms.

If you're going to bi-amp, I recommend disconnecting the lead of L4 from the input on the crossover. Without the woofer connected to the crossover, at about 500 Hz, you'll probably cook R3 by dissipating more power through than it can handle. You can skip this if you're extra careful and use a high pass filter at 1200 Hz, but I wouldn't recommend it.

 

[beagleman]

It hasn't been discussed because the L6-4R has an xmax of only 4.3mm and IEC power handling of 30 watts. Feeding it more power than that will cause it to blow out, and since the woofer of a speaker "gets" most of the power anyway (because it has to make such large excursions compared to the other drivers), it's bottlenecking the midrange and tweeter, not vice-versa. You won't be running it any louder considering the balance with the other drivers even if you bi-amp it.

I'd instead recommend adding a high pass filter at 80 Hz or higher to limit the woofer's excursion, and adding a proper subwoofer to take over from that region. The woofer can play a lot louder when it's not trying to produce output in the lower bass octaves, but you still should stay within 30 Wrms.

If you're going to bi-amp, I recommend disconnecting the lead of L4 from the input on the crossover. Without the woofer connected to the crossover, at about 500 Hz, you'll probably cook R3 by dissipating more power through than it can handle. You can skip this if you're extra careful and use a high pass filter at 1200 Hz, but I wouldn't recommend it.

An Xmax of 4.3 is not what most would consider "only"......Feeding it more than 30 watts, is not going to cause it to "Blow out" as you say, since a woofer is usually getting a music signal, that varies considerably in amplitude.

A Steady loud bass level signal such as a Test Tone may cause issues over time, but I think your fears of this being a limited weak driver, are a bit overstated for sure.

I have in my use, drivers with actual limited Xmax, more in the range of 2.3 mm, with only 25.4mm voice coils, and even they can easily withstand some loud bass at levels going over 30-45 watts with normal music, and they have survived for sure!

The Hivi drivers are more robust than my drivers I mentioned for sure.

 

[erasnick]

It hasn't been discussed because the L6-4R has an xmax of only 4.3mm and IEC power handling of 30 watts. Feeding it more power than that will cause it to blow out, and since the woofer of a speaker "gets" most of the power anyway (because it has to make such large excursions compared to the other drivers), it's bottlenecking the midrange and tweeter, not vice-versa. You won't be running it any louder considering the balance with the other drivers even if you bi-amp it.

I'd instead recommend adding a high pass filter at 80 Hz or higher to limit the woofer's excursion, and adding a proper subwoofer to take over from that region. The woofer can play a lot louder when it's not trying to produce output in the lower bass octaves, but you still should stay within 30 Wrms.

If you're going to bi-amp, I recommend disconnecting the lead of L4 from the input on the crossover. Without the woofer connected to the crossover, at about 500 Hz, you'll probably cook R3 by dissipating more power through than it can handle. You can skip this if you're extra careful and use a high pass filter at 1200 Hz, but I wouldn't recommend it.

Sorry I did not give enough information. I am already running high pass to the Hivi 3.1s at 200hz. My sub is handling everything under 200hz and below 200hz the hivi is not very flat. Also my Amplifier already handles the crossover from the mid to the highs so the woofer should only get from 200hz until whatever frequency I set and the mid and high will handle everything up from the high pass. Does this make more sense?

 

[milehigh]

This thread brought me to ASR. Have the cabinets built and veneered. For those who have built these, did you use the foam for damping? The first pics in Amir's review show some sort of polyfill in the woofer cavity.

 

[erasnick]

So, I just did mod6 to one of my Hivi speakers and compared it to mod4 perfectionist and the difference is dramatic! Sounds much better, before I preferred my Adam studio monitors and now I am leaning towards the Swans. However when I upgraded to mod 6 from mod 4 I achieved this using only the leftover crossover parts that came with the kit so that I could listen to the difference without spending any more money on partsexpress. This means the large 47uF electrolytic capacitor on the woofer is now paired with 3 6.8 Hivi foil capacitors in order to achieve 67.4uF which in my mind was close enough to the 68uF recommended for mod6. This also cleared up the sound on my woofer noticeably which I think is due to the sound quality of a foil capacitor added into the circuit where there was once just a electrolytic one. To achieve the resistor values I just wired them in series with the leftover Hivi resistors.

So now that I have this it has me wondering if any of your smarter than me engineers would know how to address the few remaining issues i see with the freq response graph. The remaining problems I see are as follows:

1. I would like to crossover the sub a little lower. I currently crossover as 200hz since the freq response looks like an EKG in this area I have circled in green. What can I do to get a flatter response from 150-250hz?

2. From 1.8-3k the is a big dip at 2.1k and and spike at 2.9k. Can this be flattened out?

3. From around 12k-20k it seems to drop off a bit. Can this be flatter?

Anyways, as I said the speakers sound great now but my thought was if there was a way to address these few remaining areas before I invest a couple of hundred dollars in crossover parts again that would be great! I am also thinking of going higher end on inductors (waxed foil), capacitors (need suggestions), and possibly resistors (non ceramic?) Reason being is that I definitely noticed the woofer get cleaner when I added the foil capacitors in the circuit so I now imagine upgrading everything will be noticeable.

 

[Mudjock]

So, I just did mod6 to one of my Hivi speakers and compared it to mod4 perfectionist and the difference is dramatic! Sounds much better, before I preferred my Adam studio monitors and now I am leaning towards the Swans. However when I upgraded to mod 6 from mod 4 I achieved this using only the leftover crossover parts that came with the kit so that I could listen to the difference without spending any more money on partsexpress. This means the large 47uF electrolytic capacitor on the woofer is now paired with 3 6.8 Hivi foil capacitors in order to achieve 67.4uF which in my mind was close enough to the 68uF recommended for mod6. This also cleared up the sound on my woofer noticeably which I think is due to the sound quality of a foil capacitor added into the circuit where there was once just a electrolytic one. To achieve the resistor values I just wired them in series with the leftover Hivi resistors.

So now that I have this it has me wondering if any of your smarter than me engineers would know how to address the few remaining issues i see with the freq response graph. The remaining problems I see are as follows:

1. I would like to crossover the sub a little lower. I currently crossover as 200hz since the freq response looks like an EKG in this area I have circled in green. What can I do to get a flatter response from 150-250hz?

2. From 1.8-3k the is a big dip at 2.1k and and spike at 2.9k. Can this be flattened out?

3. From around 12k-20k it seems to drop off a bit. Can this be flatter? View attachment 191879
Anyways, as I said the speakers sound great now but my thought was if there was a way to address these few remaining areas before I invest a couple of hundred dollars in crossover parts again that would be great! I am also thinking of going higher end on inductors (waxed foil), capacitors (need suggestions), and possibly resistors (non ceramic?) Reason being is that I definitely noticed the woofer get cleaner when I added the foil capacitors in the circuit so I now imagine upgrading everything will be noticeable.

I'm glad you are enjoying mod 6.

I'll take a shot at answering your questions, The comparison chart I made, and that you show above was based on my measurements as opposed to Amir's spinorama. That does impact how I approached mods 5 and 6 (which I proposed after Amir's review), and how the above chart should be interpreted.

1. My measurements were gated, so they aren't very accurate at low frequencies - in this case below about 300 Hz. If you look at the spinorama from the Klippel NFS, you see no "EKG" behavior from 150 Hz to 250 Hz. There is a slight dip (a dB or two) between 150 Hz and 350 Hz. Mod 6 is designed to increase output slightly in that area.

2. Amir's on axis data shows a dip just above 2 kHz also, but the predicted in room response doesn't have a dip there. Mod 5 reduces the output from about 500 Hz to just over 4 kHz, which aligns well to the region where the in-room response is trending above the line (see below)

3. It would take a more complex crossover to further flatten out the extreme high frequency region. Most ears are not very sensitive above 12 kHz, so I didn't think it would add enough value to justify making the crossover more difficult to modify. Also, what sounds good in that region will vary depending on individuals and how reflective the listening room is. I would recommend experimenting with eq in that region to see if there is anything to be gained.

As it is now, this speaker with any of the modifications up through 6 should be able to be constructed using the circuit board that comes with the kit and within that limitation is about as well dialed-in as it can be. At some point, one has to keep in mind that there are component-to-component variations both for drivers and crossover parts, so addressing a small squiggle in the response of one speaker might actually miss the mark and do more harm than good in the next copy of that speaker.

I do want to take this opportunity to again express my appreciation for the review and ASR in general. In this case, the more advanced measurements available made it possible to further refine (mods 5 an 6) a modification to the DIY-3.1A kit that already delivered strong performance for the price.

 

[beagleman]

So, I just did mod6 to one of my Hivi speakers and compared it to mod4 perfectionist and the difference is dramatic! Sounds much better, before I preferred my Adam studio monitors and now I am leaning towards the Swans. However when I upgraded to mod 6 from mod 4 I achieved this using only the leftover crossover parts that came with the kit so that I could listen to the difference without spending any more money on partsexpress. This means the large 47uF electrolytic capacitor on the woofer is now paired with 3 6.8 Hivi foil capacitors in order to achieve 67.4uF which in my mind was close enough to the 68uF recommended for mod6. This also cleared up the sound on my woofer noticeably which I think is due to the sound quality of a foil capacitor added into the circuit where there was once just a electrolytic one. To achieve the resistor values I just wired them in series with the leftover Hivi resistors.

So now that I have this it has me wondering if any of your smarter than me engineers would know how to address the few remaining issues i see with the freq response graph. The remaining problems I see are as follows:

1. I would like to crossover the sub a little lower. I currently crossover as 200hz since the freq response looks like an EKG in this area I have circled in green. What can I do to get a flatter response from 150-250hz?

2. From 1.8-3k the is a big dip at 2.1k and and spike at 2.9k. Can this be flattened out?

3. From around 12k-20k it seems to drop off a bit. Can this be flatter? View attachment 191879
Anyways, as I said the speakers sound great now but my thought was if there was a way to address these few remaining areas before I invest a couple of hundred dollars in crossover parts again that would be great! I am also thinking of going higher end on inductors (waxed foil), capacitors (need suggestions), and possibly resistors (non ceramic?) Reason being is that I definitely noticed the woofer get cleaner when I added the foil capacitors in the circuit so I now imagine upgrading everything will be noticeable.

Honestly, the "issues" you still find with the speaker are really quite small, relatively.

The graph you show is only 2 db per division, so your "Big dip" at 2.1 kHz is really only 2 db over a fairly small area.
Likewise, the "Peak" is only 1 db at 2.9 kHz.

Really not important overall, (IMHO) The response from 14or 15-20 kHz is also only a mild drop of at most a couple Db. IMHO< not even something you would hear for sure, at that extremely high range.

 

[Colonel7]

I am also thinking of going higher end on inductors (waxed foil), capacitors (need suggestions), and possibly resistors (non ceramic?) Reason being is that I definitely noticed the woofer get cleaner when I added the foil capacitors in the circuit so I now imagine upgrading everything will be noticeable

Save your money, upgrading crossover parts won't change things. If you hear the difference it would be due to changing from 47uf to 67uf and taming it by a db, not because of foil. I'd ignore above 10k (dminishing returns that high above there) and below 200 which is dominated by your room anyway, and experiment crossing your subs lower from 80 to 120Hz. That's quite nice performance you have already after Mod 6, and the designer has done a great service walking us through the mods and why.

 

[erasnick]

Save your money, upgrading crossover parts won't change things. If you hear the difference it would be due to changing from 47uf to 67uf and taming it by a db, not because of foil. I'd ignore above 10k (dminishing returns that high above there) and below 200 which is dominated by your room anyway, and experiment crossing your subs lower from 80 to 120Hz. That's quite nice performance you have already after Mod 6, and the designer has done a great service walking us through the mods and why.

I hear what you are saying Colonel7 however I could here two differences with the change to the Capacitors. One obvious change was the frequency response change. The other more subtle was the cleaner sound coming from the woofer. Why would I not attribute this cleaner sound to the addition of foil capacitors added?

Reference for foil capacitor improving sound quality:

The main pull to upgrade the capacitor components is that now that I have mod6 and am very happy with it's frequency response and in buying slightly higher quality components to build it I have also noticed that the quality of the crossover components not only improved the frequency response but also improved the clarity and distortion of the speaker as well so much so that I am very tempted to spend almost $300 more for higher quality crossover components to see how much better I can get. For example replacing the iron core inductor with air core inductor should improve sound quality.

 

[erasnick]

Honestly, the "issues" you still find with the speaker are really quite small, relatively.

The graph you show is only 2 db per division, so your "Big dip" at 2.1 kHz is really only 2 db over a fairly small area.
Likewise, the "Peak" is only 1 db at 2.9 kHz.

Really not important overall, (IMHO) The response from 14or 15-20 kHz is also only a mild drop of at most a couple Db. IMHO< not even something you would hear for sure, at that extremely high range.

Good point! These graphs are way more detailed than almost every graph I see posted but speaker manufacturers. My Yamaha receiver has a room correction that might even smooth out the remaining areas as well. For my ears the speakers sound flat and non-fatiguing now with the mod6. Before with the mod4 they were not flat at all and highs were out of control and it bothered my ears. I know I have said this before but the mod6 difference is huge and absolutely worth the cost and time spent.

 

[erasnick]

I'm glad you are enjoying mod 6.

I'll take a shot at answering your questions, The comparison chart I made, and that you show above was based on my measurements as opposed to Amir's spinorama. That does impact how I approached mods 5 and 6 (which I proposed after Amir's review), and how the above chart should be interpreted.

1. My measurements were gated, so they aren't very accurate at low frequencies - in this case below about 300 Hz. If you look at the spinorama from the Klippel NFS, you see no "EKG" behavior from 150 Hz to 250 Hz. There is a slight dip (a dB or two) between 150 Hz and 350 Hz. Mod 6 is designed to increase output slightly in that area.

2. Amir's on axis data shows a dip just above 2 kHz also, but the predicted in room response doesn't have a dip there. Mod 5 reduces the output from about 500 Hz to just over 4 kHz, which aligns well to the region where the in-room response is trending above the line (see below)

3. It would take a more complex crossover to further flatten out the extreme high frequency region. Most ears are not very sensitive above 12 kHz, so I didn't think it would add enough value to justify making the crossover more difficult to modify. Also, what sounds good in that region will vary depending on individuals and how reflective the listening room is. I would recommend experimenting with eq in that region to see if there is anything to be gained.

As it is now, this speaker with any of the modifications up through 6 should be able to be constructed using the circuit board that comes with the kit and within that limitation is about as well dialed-in as it can be. At some point, one has to keep in mind that there are component-to-component variations both for drivers and crossover parts, so addressing a small squiggle in the response of one speaker might actually miss the mark and do more harm than good in the next copy of that speaker.

I do want to take this opportunity to again express my appreciation for the review and ASR in general. In this case, the more advanced measurements available made it possible to further refine (mods 5 an 6) a modification to the DIY-3.1A kit that already delivered strong performance for the price.

Thanks for you response Mudjock. I not only enjoy mod6 I don't think that these speakers should have anything less that this because the overall sound to my ears after hearing them first stock, and then one with mod4, and now one with mod6, and having multiple people blindly decide which they like better, mod6 wins by a huge margin with everyone I have had them listen to. I think the frequency response is plenty good now. I was only being OCD about the way the graph showed a few spikes and dips but you have explained this well to me now.

Only question I have for you now is what are your thoughts on me bi-amping the speaker? My Receiver has this ability and should handle the crossover digitally before the signal goes to the lows and the highs. I would imagine if doing this I would wire the bottom post directly to the woofer and there would be no passive crossover for the woofer in this case since the amp would handle the frequencies being sent to this channel. Then the highs would be sent to the crossover which is now a mod6. Does the crossover need to be changed when powering the speaker this way? Also would you expect better performance?