Alex,
I appreciate your dedication to your 2.2 kits. I recently took the plunge when the price briefly dropped to $167. I notice that your latest take on the crossover went with relatively expensive caps and less expensive inductors. Someone else had mentioned getting good results just using the stock crossover but changing out the caps for higher end versions. I'm wondering if the quality of the caps may be the bigger deal here, at least partly because I was hoping to simply bypass the stock woofer 8.0 with a relatively expensive .22 mfd and maybe use a higher end 4.7 on the tweeter side. I assume that lowering the inductance with the woofer simply lets it play further bit higher which may be giving the speaker a "fuller" sound by providing more overlap with the output of the tweeter.
I appreciate your dedication to your 2.2 kits. I recently took the plunge when the price briefly dropped to $167. I notice that your latest take on the crossover went with relatively expensive caps and less expensive inductors. Someone else had mentioned getting good results just using the stock crossover but changing out the caps for higher end versions. I'm wondering if the quality of the caps may be the bigger deal here, at least partly because I was hoping to simply bypass the stock woofer 8.0 with a relatively expensive .22 mfd and maybe use a higher end 4.7 on the tweeter side. I assume that lowering the inductance with the woofer simply lets it play further bit higher which may be giving the speaker a "fuller" sound by providing more overlap with the output of the tweeter.
Yes, that’s right and it’s like this mostly because I have managed to destroy the coil from the previous crossover while trying to take it out of the board (strong glue!). It was terrible and it was quite expensive inductor, unfortunately. However, when assembling the last crossover I wanted to go budget friendly. I already had the board from the stock crossover and wanted to utilise it. So, you can see on the picture from my previous post that the hefty iron core 2,5 mH coil I found can actually fit there pretty nicely. An air core inductor which is around 2,5mH and less than 0,5ohms would be as big as the whole board (and crazy expensive).
The Mundorf cap (4,7uF) is chosen for the same reason, but I also wanted something of a better quality in series with the tweeter, so this one was a compromise, but I don’t regret it.
Well, my experience with the stock crossover isn’t great, but it’s all about taste, I guess. On top of that, it doesn’t measure very well. Unfortunately, I don’t have a record of the measurement that I took with the stock crossover, because I wasn’t planning to document it and post here, but then I changed my mind and I’m really sorry for not being able to back up my words with some evidence.
In my opinion, changing the cap from 8.0 to 8,2 uF is not an audible issue at all. This is exactly what I did anyway. Anything in series with a driver, though, needs to be of a good quality, if possible. In this case that’s the 4,7uF cap. So, to answer your question, yes - I would recommend using higher quality for this one.
That would give you some more “energy” in the higher midrange and you might even experience some “brightness” in the sound…
I think I mentioned that before, but what I would recommend is for whoever is building these speakers to connect the stock crossover outside of the cabinets (through the bass port, for instance) and see/hear for themselves if it’s OK or not. Then they can proceed with a modification as a second board and do some A/B tests in order to determine which one is better.
I would like to thank everyone here for the work they have done with these fantastic speakers. I just finished a pair of 2.2's and couldn't be happier. I went with Audyn 4.7 uF caps and tested a few different coils for the mid-bass and liked the Jantzen 2.2 mH 18 AWG's best. I guess they are a combo. of the D2.1 SE and D2.1 EX. In a large reflective room the running full range, the bass sounded just right, but in a small bedroom it was rather over-powering (a problem I have never had w/ a standmount before
and I had to attentuate the receiver 4 db. In fact, it's low freq. energy out-does my "Girl From Ipanema's" (a Seas 27TDFC and RS150 design by Paul Carmody) and it's in a transmission-line Tower! The tweeters are pretty simular as one might expect, which is not a bad thing. I didn't have any 5.6uF caps to try, but I have some on order and might try them as there are times, depending on the room and what DAC/amp or receiver are in the mix, that I wish there was a little more treble. Unfortunately the X-overs are mounted , but I suspose I can find a way.
My other pair (2 pairs actually, MT's and MTM's) of "good" speakers are the Aviatrix by Curt Campbell and they are not here, but I know them so well I can compare them in my head. The Aviatrix's ND140 woofers are their weak link and the HiVi's bury them in every way, output, tonality, looks etc. (they are in a sealed box though). But the original Dayton RS28F are still the best tweeter I have ever heard and the reason that the Aviatrices are still being used regularly 10 years after I built them. IMO, the Dayton is a little better than the Seas, which in turn are a little better than the Hivi's.
P.E. still sells the Aviatrix MTM pair in a complete kit for $299 and it's a gd. value, but I am lerry of the new RST28F tweeters. I just priced out what I would cost to build the Ipanemas today and even with their very simple X-overs (very much like the Hivi/Swans) and P.E.'s budget .23 cu/ft boxs it was well over $400. But we know what the Hivi/Swans sell for and it's more speaker. Just a fabulous deal that can't be beat anywhere.
Now the problem of what to pull out of rotation.
I know, build another system
P.S. Concerning "break-in", both woofers needed some time. In fact, the second one gave me a bit of a fright. W/ the first X-over working fine and being so simple, I didn't test the second one and of course that's the woofer that threw a fit. It alternated between making a "whriling, clicking' sound and cutting in and out for a full 15 minutes and I was so alarmed I switched out the gear. But after 30 minutes transpired, it calmed down and was fine.
and I had to attentuate the receiver 4 db. In fact, it's low freq. energy out-does my "Girl From Ipanema's" (a Seas 27TDFC and RS150 design by Paul Carmody) and it's in a transmission-line Tower! The tweeters are pretty simular as one might expect, which is not a bad thing. I didn't have any 5.6uF caps to try, but I have some on order and might try them as there are times, depending on the room and what DAC/amp or receiver are in the mix, that I wish there was a little more treble. Unfortunately the X-overs are mounted , but I suspose I can find a way.My other pair (2 pairs actually, MT's and MTM's) of "good" speakers are the Aviatrix by Curt Campbell and they are not here, but I know them so well I can compare them in my head. The Aviatrix's ND140 woofers are their weak link and the HiVi's bury them in every way, output, tonality, looks etc. (they are in a sealed box though). But the original Dayton RS28F are still the best tweeter I have ever heard and the reason that the Aviatrices are still being used regularly 10 years after I built them. IMO, the Dayton is a little better than the Seas, which in turn are a little better than the Hivi's.
P.E. still sells the Aviatrix MTM pair in a complete kit for $299 and it's a gd. value, but I am lerry of the new RST28F tweeters. I just priced out what I would cost to build the Ipanemas today and even with their very simple X-overs (very much like the Hivi/Swans) and P.E.'s budget .23 cu/ft boxs it was well over $400. But we know what the Hivi/Swans sell for and it's more speaker. Just a fabulous deal that can't be beat anywhere.
Now the problem of what to pull out of rotation.
I know, build another system
P.S. Concerning "break-in", both woofers needed some time. In fact, the second one gave me a bit of a fright. W/ the first X-over working fine and being so simple, I didn't test the second one and of course that's the woofer that threw a fit. It alternated between making a "whriling, clicking' sound and cutting in and out for a full 15 minutes and I was so alarmed I switched out the gear. But after 30 minutes transpired, it calmed down and was fine.
Last edited:
Well, there was a problem. To get the big 2.2mH air core inductors on the small PCB and connect to the traces, I ran both leads thru the pre-drilled hole right in the center of the inductor and when the speaker would vibrate, they would make contact. Since I had get back in there, I went back to the supplied 1.5mH coils.I would like to thank everyone here for the work they have done with these fantastic speakers. I just finished a pair of 2.2's and couldn't be happier. I went with Audyn 4.7 uF caps and tested a few different coils for the mid-bass and liked the Jantzen 2.2 mH 18 AWG's best. I guess they are a combo. of the D2.1 SE and D2.1 EX. In a large reflective room the running full range, the bass sounded just right, but in a small bedroom it was rather over-powering (a problem I have never had w/ a standmount before
My other pair (2 pairs actually, MT's and MTM's) of "good" speakers are the Aviatrix by Curt Campbell and they are not here, but I know them so well I can compare them in my head. The Aviatrix's ND140 woofers are their weak link and the HiVi's bury them in every way, output, tonality, looks etc. (they are in a sealed box though). But the original Dayton RS28F are still the best tweeter I have ever heard and the reason that the Aviatrices are still being used regularly 10 years after I built them. IMO, the Dayton is a little better than the Seas, which in turn are a little better than the Hivi's.
P.E. still sells the Aviatrix MTM pair in a complete kit for $299 and it's a gd. value, but I am lerry of the new RST28F tweeters. I just priced out what I would cost to build the Ipanemas today and even with their very simple X-overs (very much like the Hivi/Swans) and P.E.'s budget .23 cu/ft boxs it was well over $400. But we know what the Hivi/Swans sell for and it's more speaker. Just a fabulous deal that can't be beat anywhere.
Now the problem of what to pull out of rotation.
I know, build another system
P.S. Concerning "break-in", both woofers needed some time. In fact, the second one gave me a bit of a fright. W/ the first X-over working fine and being so simple, I didn't test the second one and of course that's the woofer that threw a fit. It alternated between making a "whriling, clicking' sound and cutting in and out for a full 15 minutes and I was so alarmed I switched out the gear. But after 30 minutes transpired, it calmed down and was fine.
The Hivi's ended up in a bedroom, closer to a wall than optimum and there is plenty of bass in this application. There is also a 10" sub in the vicinity.
After trying different caps on the tweeter, I've decided the 4.7uF are the right ones for me. But even taming the highs, sitting closer than 6 or so feet, gives one that "in your face" impression. Get further back and all is bliss.
BTW, It's not hard to pull the PCB off their nylon mounting tits. Just give them a squeeze w/ a pr. of needle noses and the board will work off fairly easy w/out damaging the mounts.
I've been following this thread from the beginning. In fact, I've built a pair of these and am super happy with them. But I'm a novice with a bunch of wood working experience. So, for me, it was a fun project to do with my teenager and end up with what I consider to be a really nice (and beautiful) pair of speakers (mine are black with walnut sides in the theme of some of the other swans I've seen pics of). But, honestly (as you might guess from my name), I can't quite follow all the tech that is happening here. With that in mind, I think it would be useful for other newbies, if someone would distill all this great work down, and provide a buy list for the 1. best upgrade, and 2. a worthy but inexpensive upgrade. That way, folks could take advantage of all this work y'all have done and end up with excellent speakers too.
It's not a hard assessment as even doing nothing, these speakers sound great. There is the balancing act of adding higher quality parts vs improved sound quality that applies to all DIY speaker designs/builds, so that aspect has to be the listener's judgement call.
Having said that, I think the 4.7uF cap on the tweeter works well.
As for the inductor value, it really seems to be a matter of placement and usage. The Woofer and enclosure's bass out-put is prodigious and there can be "too much of a good thing. I experienced that w/ a 2.2mH and the speakers in a smaller room. I'm waiting to find a final home for mine(as in what system and room) before I try other values." Again, "season to taste".
I found the cross-over PBC easy to remove to make any changes, but squeezing a lg. air-core inductor on it to be challenging. A Laminate steel inductor is easier to work with.
I think the "quality" of the stock parts are fine, and the speakers sound great even with the stock crossover, but there is room for improvement. The problems I noticed the most were, slightly reduced bass, slightly elevated treble, and a suck-out at 3k Hz. I did a couple of measurements, and while they were pretty rough-looking, they confirmed the bass issue. The 3khz issue I didn't notice by ear but seems obvious when looking at the measurement graph. I would suggest doing your own measurements and see if you can get more accuracy than I did; maybe you'll reach different conclusions. Your best approach is to do nothing if you're happy with the bass on the stock crossover! However, you might want to measure and see if it matches your expectations. Indeed, placement makes a huge difference, particularly for bass.
The 3k Hz issue is easily solved by reversing the polarity of the tweeter. That's your cheapest upgrade (free!), however, it may be an intentional part of the original crossover design, so I would be careful about making this change in isolation.
I think the 2.1SE crossover is a relatively cheap and simple upgrade and is the best bang-for-the-buck. Which is what I did. It's similar to stock, except you change the tweeter's cap for a 4.7uf and the woofer's inductor for a 2.6mh one. I had trouble finding a 2.6, so I got this one:
https://www.madisoundspeakerstore.com/sidewinder-16-awg/sidewinder-2.5-mh-air-core-inductor-16-awg/
This is the biggest expense of the modification. At the moment of this writing, these are $33, and you need a pair, so already you're up to $66 not including shipping and the caps.
This mod (2.1SE crossover) doesn't do anything for the elevated treble above, say, 4KHz, Allex had additional components to tame the tweeter. However, this did what I wanted to do -- relatively flat response from the bass to upper-midrange. And the upper treble is no worse than before, as it's unchanged, but it sounds less harsh, now that mids and bass are raised a bit. If you try other values (than the ones in the 2.1SE schematic), you probably should model them in software to see how it changes, or at least measure when done. Differences in DC resistance as well as the induction value will change the output. The stock inductor's resistance of 1.1 ohms seems like a bit much, but maybe it doesn't make as big of a difference as I expect. Model and measure and see for yourself.
However, you might want to measure and see if it matches your expectations. Indeed, placement makes a huge difference, particularly for bass.The 3k Hz issue is easily solved by reversing the polarity of the tweeter. That's your cheapest upgrade (free!), however, it may be an intentional part of the original crossover design, so I would be careful about making this change in isolation.
I think the 2.1SE crossover is a relatively cheap and simple upgrade and is the best bang-for-the-buck. Which is what I did. It's similar to stock, except you change the tweeter's cap for a 4.7uf and the woofer's inductor for a 2.6mh one. I had trouble finding a 2.6, so I got this one:
https://www.madisoundspeakerstore.com/sidewinder-16-awg/sidewinder-2.5-mh-air-core-inductor-16-awg/
This is the biggest expense of the modification. At the moment of this writing, these are $33, and you need a pair, so already you're up to $66 not including shipping and the caps.
This mod (2.1SE crossover) doesn't do anything for the elevated treble above, say, 4KHz, Allex had additional components to tame the tweeter. However, this did what I wanted to do -- relatively flat response from the bass to upper-midrange. And the upper treble is no worse than before, as it's unchanged, but it sounds less harsh, now that mids and bass are raised a bit. If you try other values (than the ones in the 2.1SE schematic), you probably should model them in software to see how it changes, or at least measure when done. Differences in DC resistance as well as the induction value will change the output. The stock inductor's resistance of 1.1 ohms seems like a bit much, but maybe it doesn't make as big of a difference as I expect. Model and measure and see for yourself.
So the 2.2s went down to $191 so I ordered some yesterday. I will be making them using the 2.1se crossover. My biggest question is because I will have them closer(.5 to 1.5ft) to the wall than is ideal I understand that the bass response will be higher than if it were 3-5 feet from the wall, and raising the inductor will increase the bass response as well. I was thinking of only going up to 2.2mH. Now I will have this connected to my AVR with a sub that can be set to cross at 60,80, or100Khz. What bass frequency area does the different inductor affect? The lowest under 100Khz, or the 100-500Khz, or the entire curve? If I'm not concerned with the lowest bass (<60Khz) what inductor do you guys recommend? I am looking at the Jantzen 5280 2.2mH 20 AWG P-Core Inductor with a DCR of .40 ($8.73) or a Dayton Audio IC182-5 2.5mH 18 AWG Laminated Iron Core Inductor with a DCR of .31($7.98)
The capacitor to 4.7uf seems like a universal recommendation and I will use a Jantzen crosscap ($5.55 ea.) For the other cap I will use a Dayton film/foil DFFC-0.22 0.22uF 400V By-Pass cap ($1.28ea) to bring the 8uf up the the 2.1 value. I've heard using a better quality cap in parallel with a cheaper cap brings up the quality overall. Not sure if it will be audible or not but for a dollar I don't think it will hurt.
As for the series 6.8ohm resistors on the Tweeter I might replace that with a Dayton Audio DPR20-7.0 7 Ohm 20 Watt Precision 1% Audio Grade Resistor ($4.98ea) because I heard anything in series with the driver should be of higher quality. Will the .2ohm difference matter? I can't source a 6.8ohm resistor.
I am interested in everyone's opinions as I am still learning all this stuff. Thanks in advance!
Travis
The capacitor to 4.7uf seems like a universal recommendation and I will use a Jantzen crosscap ($5.55 ea.) For the other cap I will use a Dayton film/foil DFFC-0.22 0.22uF 400V By-Pass cap ($1.28ea) to bring the 8uf up the the 2.1 value. I've heard using a better quality cap in parallel with a cheaper cap brings up the quality overall. Not sure if it will be audible or not but for a dollar I don't think it will hurt.
As for the series 6.8ohm resistors on the Tweeter I might replace that with a Dayton Audio DPR20-7.0 7 Ohm 20 Watt Precision 1% Audio Grade Resistor ($4.98ea) because I heard anything in series with the driver should be of higher quality. Will the .2ohm difference matter? I can't source a 6.8ohm resistor.
I am interested in everyone's opinions as I am still learning all this stuff. Thanks in advance!
Travis
Hi Travis,
Changing the value of the inductor for the LP filter would affect only the upper midrange and not the bass response. The bass is basically determined by the properties of the bass driver in combination with the enclosure (volume and port). So, I would stick to the 2.5mH as the best possible value, IMO.
As for the bypass cap, it doesn’t hurt to try and see if there is any benefit, but for me it make sense to use it in parallel to the cap which is in series with the tweeter, namely 4,7 uF and not the 8,2 uF.
It’s a bit strange that you can’t source any 6,8 ohm resistors as this is actually a pretty standard value compared to the 7 ohm you can get. Anyway, it shouldn’t be a big difference.
Changing the value of the inductor for the LP filter would affect only the upper midrange and not the bass response. The bass is basically determined by the properties of the bass driver in combination with the enclosure (volume and port). So, I would stick to the 2.5mH as the best possible value, IMO.
As for the bypass cap, it doesn’t hurt to try and see if there is any benefit, but for me it make sense to use it in parallel to the cap which is in series with the tweeter, namely 4,7 uF and not the 8,2 uF.
It’s a bit strange that you can’t source any 6,8 ohm resistors as this is actually a pretty standard value compared to the 7 ohm you can get. Anyway, it shouldn’t be a big difference.
Thanks Allex, I will go with the 2.5mH inductor and I will use bypass caps on both of the caps. As for the resistor, I'm only shopping on parts express.com to save on shipping and they don't stock the 6.8ohm resistors so a 7ohm it is.
Do you think moving the port to the front of the cabinet will help with placement of the speaker being closer to the wall than recommended? I like the look of the port being on the front of speakers in general so as long as there's no negatives I will do it. What do you think?
Do you think moving the port to the front of the cabinet will help with placement of the speaker being closer to the wall than recommended? I like the look of the port being on the front of speakers in general so as long as there's no negatives I will do it. What do you think?
I tried a 2.2mH and 2.5mh and went back to the stock 1.5mH. I already had a nice 10" sub and was able to intergrate it well below 80Hz. Maybe if I didn't have a sub, I would have used the 2.2mH air core coil. I couldn't detect much difference w/ the 2.5mH steel laminate coil, but it was easier to fit on the small X-over board.
If you will be using a sub, I would suggest trying the supplied coil first. You could leave the X-over outside the box to test, but as I found out, it's not difficult to remove the X-over board once it's installed. Going thru the woofer and terminal openings, simply squeeze the peg/clips w/ needle-nose pliers that mount the board to the enclosure bottom and it lifts right out. Make sure you use spade terminals to connect the wires to the board.
I think moving the port to the frt. would be a mistake, besides, there is not enough room.
You are really going to like these speakers.
If you will be using a sub, I would suggest trying the supplied coil first. You could leave the X-over outside the box to test, but as I found out, it's not difficult to remove the X-over board once it's installed. Going thru the woofer and terminal openings, simply squeeze the peg/clips w/ needle-nose pliers that mount the board to the enclosure bottom and it lifts right out. Make sure you use spade terminals to connect the wires to the board.
I think moving the port to the frt. would be a mistake, besides, there is not enough room.
You are really going to like these speakers.
So Partsconexion.com has a capacitor sale right now so I'll probably get these Clarity CSA 4.7uf capacitors for $8 each. https://www.partsconnexion.com/CLARITY-81524.html
-Any other capacitor brand/type recommendations keeping price in mind?
-I did buy a Dayton microphone so maybe I'll measure it first before adding the new inductor. I just have to learn how to use it first. (Distance/where to center it?)
-Any other capacitor brand/type recommendations keeping price in mind?
-I did buy a Dayton microphone so maybe I'll measure it first before adding the new inductor. I just have to learn how to use it first. (Distance/where to center it?)
Tuning and overall design of the speaker (volume & port dimensions) are generally more critical than where the port is located on the speaker. Placing it in front, though, could be tricky as there’s potential risk of affecting the performance with some cancellations in the frequency response. I’ve seen measurements of speakers with dips in the FR caused by port resonances.
What I can suggest is to connect the crossover components outside the cabinet so you can make changes with different values and decide for yourself which ones suit your needs and taste best. Of course, if you can combine this approach with measurements, that would be even better. Unfortunately, I’ve sold my speakers and cannot be of any help with additional measurements on my part.
What I can suggest is to connect the crossover components outside the cabinet so you can make changes with different values and decide for yourself which ones suit your needs and taste best. Of course, if you can combine this approach with measurements, that would be even better. Unfortunately, I’ve sold my speakers and cannot be of any help with additional measurements on my part.
I was able to build the 2.2s with the SE modification. I found the tweeters very harsh and loud so rewired the crossover to Bi-Amp them, since it already came with the right binding posts. I was able to control the level to the tweeters with a separate amp, but since then, I am using a Carver PM1090 (amazing!) and am using the A + B outputs (A Lo, B Hi) but unable to separately control the gain. Again, the highs were too loud and harsh, so I used my adjustable resistors in series with the tweeter input and was able to quite down the tweeters. I'm sure I changed something with the crossover frequency, but whatever I did made an amazing improvement. I ended up with a 15 ohm resistor to replace the 6.8 ohm one and these little things sound spectacular. I have documented all of my changes in this modified drawing. Happy Listening.
Attachments
15 instead of 6,8 Ohm sounds like way too much, in my opinion. I can hardly imagine these speakers having even remotely decent treble with so much resistance in front of the tweeters. You made me dig for the old files of this project and load them in XSim. Let’s have a look…
OK, now ignore the ramping up frequency response in the top end - it’s a result of a bad amp used for the initial measurements. In reality it’s much better. The red trace is the response with 15 Ohm resistor and the blue one is with the SE modification (6.8 Ohm resistor). The deference appears to be around 4-5 dB!
It’s all about taste, I guess
OK, now ignore the ramping up frequency response in the top end - it’s a result of a bad amp used for the initial measurements. In reality it’s much better. The red trace is the response with 15 Ohm resistor and the blue one is with the SE modification (6.8 Ohm resistor). The deference appears to be around 4-5 dB!
It’s all about taste, I guess
This could have to do with his musical preference, position in the room, and speaker spectral balance preference.
1. He is in a corner and right next to a wall
2. He may listen to compressed rock music
3. He may prefer a polite tweeter with a BBC dip due to the other 2 criteria.
I don't think his response as you modeled it is actually all that terrible.
1. He is in a corner and right next to a wall
2. He may listen to compressed rock music
3. He may prefer a polite tweeter with a BBC dip due to the other 2 criteria.
I don't think his response as you modeled it is actually all that terrible.
Yes, 4-5db is about what I hear, thanks for modelling that. I am playing very high end DSD files ( jazz, contemporary, dire straits, lol) and am in a loud room, right speaker is near a corner. I have burned in the speakers for over 100 hours and treated the floors and walls (not shown in the picture). I do believe that separating the high and low drivers with bi- amping has made a huge difference in the punch and soundstage. I am doing this with my Kef LS50's and they need more gain on the tweeters, again, very subjective to ones taste. The nice thing about bi-amping is it allows the flexibility to do the separate adjustments, either with separate amps or with external resistors.
Last edited:
On xsim I tried the impedance compensation (CR) network, with the 1,5 uF capacitor and the 6,8 ohm resistor across the tweeter that Allex came up with and it seems to do a similar thing to the FR as changing R1 to 15ohms. I was going to change the 6.8ohm for a better quality resistor anyway because it's in the series path with the tweeter, so I only have to buy a 1.5uf for a few bucks.