I have not taken any polars as of yet. I just wanted to see what things looked like. Knowing I have not taken polars, here is a modeled response using four components:
I have not assembled components and tested that, yet.
Some distortion measurements, the tweeter did really well I think.
That looks really good, but I'm curious about the off axis, looking forward to the polars. I firmly believe that Dynavox woofers are under rated. If I didn't state it earlier, the cabinet shape is excellent.
I'll take polars tomorrow and do some serious crossover work at that time. In the meantime, I have been playing with the on-axis files.
Here is something with 11 components:
If the directivity match is good enough at 1500hz, this one would likely warrant exploring. Six of the components are for two notch filters - one at ~4k on the woofer and one to tailor the tail on the bottom end of the tweeter. Not an overly simple crossover, but nothing too strange either.
Given the sheer amount of components I have accumulated over the years, I no longer have fear of throwing components at a design. Just heave shit at the breadboard until something works. Just go nuts. Kind of liberating, to be honest.
That is a prototype polartakerizer I built a while back. I keep meaning to refine it to be part of an overall flexible measurement speakerholderizer at some point. Scrap wood, a cheap ball bearing lazy susan thing from McMaster-Carr, and some time spent finding center points. Takes just a few seconds to ensure the speaker is on-center and off I go.
Anyways, tweeter at 0-45 in 5 degree increments:
If you imagine normalizing it in your head, it looks like it stays pretty constant up to ~2500. Other than not extending the bottom end as far as I would have hoped on the Seas tweeter, it appears to be a pretty good match. This is the H8801-06 model, built in 2009. I had to overdrill the four mounting holes on the tweeter faceplate by about 1mm, but it is otherwise just bolted right on to the waveguide.
Woofer (on tweeter axis) 0-45 in 5 degree increments:
Again, with the power of the mind, normalize and it looks pretty constant up to about 1K. The off-axis is otherwise normal - no surprises really.
This is going to lead to a slight mismatch in directivity, the large diameter waveguide, the relatively small cone to flange area of the woofer both lead to a pretty suboptimal CTC spacing.
However, I believe I can utilize the lower order slopes to help mitigate some of that.
With a small form factor midrange driver, say a typical 3", crossed to a 7 or 8" woofer at ~300Hz one could expect more or less CD to 2500Hz with this tweeter in this waveguide. In my opinion, that would look pretty funky, but I guess a guy could throw a grill over it.
Anyways. Polartakerizer is going back on the shelf and I am going to assemble the notch filters I modeled and see what happens.
Oh, and I am fairly impressed with the woofer response. The facets help with that somewhat, I imagine. The only real concern is a slight off-axis bunching centered at 2.1K or so. The mild breakup behaves as breakups do - when I first measured off-axis breakups years ago I stopped trusting pistonic modeling in crossover software. Breakups are rarely pistonic, I guess that is why they are called breakups. What the hell do I know.
Actually, looking again - if I am willing to tolerate a slight off-axis depression centered at 1.5K, the bunching at 2.1k emulates CD on the woofer. 2K would be as good of a crossover point as any other. I'll check it out.
Don't let the young DIYers see that picture! They'll all start putting scissors in their speaker cabinets and claim the vocals really cut through the mix.
I know I have measured quite a few, and have yet to have one fall outside of ~1% tolerance. I otherwise don't really express opinions on sound quality.
Out of curiosity, I'd be interested in what you might, or might not find, subbing something else in there. I wonder if some caps work better, say in a crossover, than a higher voltage output cap?
Anyhow, I like the way the angled sides look, and surely it helps.
@Kornbread said:
Out of curiosity, I'd be interested in what you might, or might not find, subbing something else in there. I wonder if some caps work better, say in a crossover, than a higher voltage output cap?
Anyhow, I like the way the angled sides look, and surely it helps.
You are probably on to something on the voltage thing. I'm really not sure, I have detected subtle differences in caps in the past - but I suspect that ability has either declined with age or I have gotten better at nailing a crossover sooner in the process and my motivation to roll components has diminished. As I said, I'm not really sure. What I do know, is subtle differences in phase can be detected - as well as minor dips/peaks in response if they are in sensitive regions (such as where many speakers are crossed over). Variations in component value affect the crossover region in both phase alignment and linear distortion. At the end of the day, I would argue component type/material is secondary to decisions/compromises made by the designer on crossover topology and driver choices.
Been rolling these again this weekend, I had soldered up what I thought was the crossover - but I mislanded one lead on each board and swapped a couple components around. Long story short, they didn't work. It has been a very long time since I messed up a crossover assembly process.
As a result, I just kind of decided to start over and came up with a 4 element crossover (2nd electrical on woofer, 1st electrical plus padding resistor on tweeter). If I get motivated today, I will post measurements etc. They are not super flat in this configuration (by DIY standards - but they are still within +/-3db), but they have excellent directivity.
Anyways, with all the talk about the honeycomb Dynavox with Indy coming up, thought I would resurrect this thread.
I have to say that I quite like these woofers. I've been messing with the Missing Links for a bit now, and the mids really are pretty good, treated properly.
So, before posting a couple measurements etc, some explanation. I found a significant spectral tilt was needed to balance these out. It so happens this was easy to do for a number of reasons.
The bass is pretty lean - about a 70Hz Fb. I forgot to save the measurement, but the nearfield confirms a bottom end in the 70-80Hz range. These will definitely require augmentation of some kind. Subjectively, the bass that is there is pretty nice.
Anyways, buttoned them up and they are done. These will serve short term duty in the living room after next weekend. We are hosting a couple Sat night, and the little towers I brought to Iowa will be going home with them. The Taiga are moving downstairs and these will be setup in the living room. I may build a quickie subwoofer at some point for upstairs but I dunno. Taiga very well could end up there long term depending on what I ever decide on for my basement.
I digress.
On-axis and impedance. Impedance is a little swingy, and although it doesn't show on this picture the phase pushes the limits in the treble region. Not terrible, but it might prove confusing for tubes and cheaper class D amps. I use neither.
Distortion. In the passband the measurement can be more or less trusted, it doesn't exceed -40db. I did not do thermal compression since I was tired of listening to test tones at this point.
Here are some horizontals, 5 degree increments out to 45 degrees. I only take them in one direction since I only build symmetrical baffles these days. You would think seeing how it behaves above 8K that there would seem to be a lack of "sparkle" but to my ears at least, it is just about right. They pass the sibilance tests with flying colors. Very natural sounding. The Seas is no joke as a tweeter. The only complaint I have about it is the bottom end is anemic - without the boost from the waveguide it has a pretty weird rolloff below 2K.
The crossover, four components. Hand soldered, fully point to point, premium components to minimize insertion losses and phase fuckery. Sorry - had to indulge some marketing speak
All in, these can be built for around $300-350. There are better speakers for that money, to be honest. Not that these do anything particularly poorly - in fact, they sound very nice it's just you have to ask yourself how many 6-7" 2-ways for this kind of money will perform better on the bottom end. PE currently has the prices for the Dynavox jacked to $70. That is pretty premium territory in my opinion. The RS180, the Peerless HDS, and even the aluminum cone Dayton Designer are all more or less in this price class and offer (imho) better performance. That being said, this is a solid design and with a little help on the bottom end doesn't really give up a whole lot to designs using the aforementioned drivers.
If I get really ambitious, I'll buy one of these and build a (kind of) matching subwoofer:
Steve has the amps as well, and if I can find a breakdown on the cabinet design KRK used I can build a pretty solid subwoofer to help these out. I would bet dollars to donuts that KRK has some EQ built into the amp.
Comments
I guess it was actually not enough toe in.
I assume that's your toe.
Ron
Some work done today...
I have not taken any polars as of yet. I just wanted to see what things looked like. Knowing I have not taken polars, here is a modeled response using four components:
I have not assembled components and tested that, yet.
Some distortion measurements, the tweeter did really well I think.
At this volume level, the woofer is pretty clean on the bottom end.
That looks really good, but I'm curious about the off axis, looking forward to the polars. I firmly believe that Dynavox woofers are under rated. If I didn't state it earlier, the cabinet shape is excellent.
I'll take polars tomorrow and do some serious crossover work at that time. In the meantime, I have been playing with the on-axis files.
Here is something with 11 components:
If the directivity match is good enough at 1500hz, this one would likely warrant exploring. Six of the components are for two notch filters - one at ~4k on the woofer and one to tailor the tail on the bottom end of the tweeter. Not an overly simple crossover, but nothing too strange either.
Given the sheer amount of components I have accumulated over the years, I no longer have fear of throwing components at a design. Just heave shit at the breadboard until something works. Just go nuts. Kind of liberating, to be honest.
Time to make the whoop whoops.
That is a prototype polartakerizer I built a while back. I keep meaning to refine it to be part of an overall flexible measurement speakerholderizer at some point. Scrap wood, a cheap ball bearing lazy susan thing from McMaster-Carr, and some time spent finding center points. Takes just a few seconds to ensure the speaker is on-center and off I go.
Anyways, tweeter at 0-45 in 5 degree increments:
If you imagine normalizing it in your head, it looks like it stays pretty constant up to ~2500. Other than not extending the bottom end as far as I would have hoped on the Seas tweeter, it appears to be a pretty good match. This is the H8801-06 model, built in 2009. I had to overdrill the four mounting holes on the tweeter faceplate by about 1mm, but it is otherwise just bolted right on to the waveguide.
Woofer (on tweeter axis) 0-45 in 5 degree increments:
Again, with the power of the mind, normalize and it looks pretty constant up to about 1K. The off-axis is otherwise normal - no surprises really.
This is going to lead to a slight mismatch in directivity, the large diameter waveguide, the relatively small cone to flange area of the woofer both lead to a pretty suboptimal CTC spacing.
However, I believe I can utilize the lower order slopes to help mitigate some of that.
With a small form factor midrange driver, say a typical 3", crossed to a 7 or 8" woofer at ~300Hz one could expect more or less CD to 2500Hz with this tweeter in this waveguide. In my opinion, that would look pretty funky, but I guess a guy could throw a grill over it.
Anyways. Polartakerizer is going back on the shelf and I am going to assemble the notch filters I modeled and see what happens.
Oh, and I am fairly impressed with the woofer response. The facets help with that somewhat, I imagine. The only real concern is a slight off-axis bunching centered at 2.1K or so. The mild breakup behaves as breakups do - when I first measured off-axis breakups years ago I stopped trusting pistonic modeling in crossover software. Breakups are rarely pistonic, I guess that is why they are called breakups. What the hell do I know.
Actually, looking again - if I am willing to tolerate a slight off-axis depression centered at 1.5K, the bunching at 2.1k emulates CD on the woofer. 2K would be as good of a crossover point as any other. I'll check it out.
H0881-06, not H8801-06.
Well, the kid staked a claim to these.
Good choice I think.
Those should be solid performers.
Some shit in the cabinet that needs to go downstairs, but just wanted to show lack of seam broadcasting (so far).
And the top:
Don't let the young DIYers see that picture! They'll all start putting scissors in their speaker cabinets and claim the vocals really cut through the mix.
Got the crossovers soldered up today. Will do final testing and assembly by Sunday.
I liked those audiophiler caps until I used some in the 4s preamp as output caps. Have you noticed a difference between them and, say, a q4?
I know I have measured quite a few, and have yet to have one fall outside of ~1% tolerance. I otherwise don't really express opinions on sound quality.
Every one of them I measured was spot-on.
I've used them in a few budget builds - this is the first build using decent drivers that I am using them in. I think they should be OK.
Out of curiosity, I'd be interested in what you might, or might not find, subbing something else in there. I wonder if some caps work better, say in a crossover, than a higher voltage output cap?
Anyhow, I like the way the angled sides look, and surely it helps.
Parked for a few days, waiting for a part. Just liked this picture.
Whatever will I do in the meantime?
I could give you a few ideas
There's always AA. Audio anonymous.
You are probably on to something on the voltage thing. I'm really not sure, I have detected subtle differences in caps in the past - but I suspect that ability has either declined with age or I have gotten better at nailing a crossover sooner in the process and my motivation to roll components has diminished. As I said, I'm not really sure. What I do know, is subtle differences in phase can be detected - as well as minor dips/peaks in response if they are in sensitive regions (such as where many speakers are crossed over). Variations in component value affect the crossover region in both phase alignment and linear distortion. At the end of the day, I would argue component type/material is secondary to decisions/compromises made by the designer on crossover topology and driver choices.
In any event, red components look pretty cool.
Been rolling these again this weekend, I had soldered up what I thought was the crossover - but I mislanded one lead on each board and swapped a couple components around. Long story short, they didn't work. It has been a very long time since I messed up a crossover assembly process.
As a result, I just kind of decided to start over and came up with a 4 element crossover (2nd electrical on woofer, 1st electrical plus padding resistor on tweeter). If I get motivated today, I will post measurements etc. They are not super flat in this configuration (by DIY standards - but they are still within +/-3db), but they have excellent directivity.
Anyways, with all the talk about the honeycomb Dynavox with Indy coming up, thought I would resurrect this thread.
I have to say that I quite like these woofers. I've been messing with the Missing Links for a bit now, and the mids really are pretty good, treated properly.
InDIYana Event Website
So, before posting a couple measurements etc, some explanation. I found a significant spectral tilt was needed to balance these out. It so happens this was easy to do for a number of reasons.
The bass is pretty lean - about a 70Hz Fb. I forgot to save the measurement, but the nearfield confirms a bottom end in the 70-80Hz range. These will definitely require augmentation of some kind. Subjectively, the bass that is there is pretty nice.
Anyways, buttoned them up and they are done. These will serve short term duty in the living room after next weekend. We are hosting a couple Sat night, and the little towers I brought to Iowa will be going home with them. The Taiga are moving downstairs and these will be setup in the living room. I may build a quickie subwoofer at some point for upstairs but I dunno. Taiga very well could end up there long term depending on what I ever decide on for my basement.
I digress.
On-axis and impedance. Impedance is a little swingy, and although it doesn't show on this picture the phase pushes the limits in the treble region. Not terrible, but it might prove confusing for tubes and cheaper class D amps. I use neither.
Distortion. In the passband the measurement can be more or less trusted, it doesn't exceed -40db. I did not do thermal compression since I was tired of listening to test tones at this point.
Here are some horizontals, 5 degree increments out to 45 degrees. I only take them in one direction since I only build symmetrical baffles these days. You would think seeing how it behaves above 8K that there would seem to be a lack of "sparkle" but to my ears at least, it is just about right. They pass the sibilance tests with flying colors. Very natural sounding. The Seas is no joke as a tweeter. The only complaint I have about it is the bottom end is anemic - without the boost from the waveguide it has a pretty weird rolloff below 2K.
The crossover, four components. Hand soldered, fully point to point, premium components to minimize insertion losses and phase fuckery. Sorry - had to indulge some marketing speak
All in, these can be built for around $300-350. There are better speakers for that money, to be honest. Not that these do anything particularly poorly - in fact, they sound very nice it's just you have to ask yourself how many 6-7" 2-ways for this kind of money will perform better on the bottom end. PE currently has the prices for the Dynavox jacked to $70. That is pretty premium territory in my opinion. The RS180, the Peerless HDS, and even the aluminum cone Dayton Designer are all more or less in this price class and offer (imho) better performance. That being said, this is a solid design and with a little help on the bottom end doesn't really give up a whole lot to designs using the aforementioned drivers.
If I get really ambitious, I'll buy one of these and build a (kind of) matching subwoofer:
Steve has the amps as well, and if I can find a breakdown on the cabinet design KRK used I can build a pretty solid subwoofer to help these out. I would bet dollars to donuts that KRK has some EQ built into the amp.
Anyways, time to cook dinner.
So, with or without toe-in? Nice minimalist crossover JR.