Introducing Denotiva DIY speakers!

[Ktacos]

What timing, I just put together this same driver combo but in separate cabinets for the drivers. It sounded pretty good but I ultimately didn't care for the integration between them, I just couldn't get anything I was happy with, my drivers were spaced a fair bit apart compared to yours though. I ended up pairing the woofer with a dx25+wg300 and that seems to be a near perfect pairing.

I wish PE would do more with the nd25fw, it's their best tweeter imo.

 

[RickS]

What timing, I just put together this same driver combo but in separate cabinets for the drivers. It sounded pretty good but I ultimately didn't care for the integration between them, I just couldn't get anything I was happy with, my drivers were spaced a fair bit apart compared to yours though. I ended up pairing the woofer with a dx25+wg300 and that seems to be a near perfect pairing.

I wish PE would do more with the nd25fw, it's their best tweeter imo.

have had some comparable integration challenges but suspect part of it was the cabinet I was using as a measurement fixture. I built new one and am just refining it and will make another run at a crossover.

 

[RickS]

Well I have a first cut but was not near as easy as I thought. The crossover for the bass portion was simple enough but the ND25FW is giving me fits. It is not nearly as linear as its spec sheet would lead one to believe. I ended up swapping one out of a C-Note as I thought it might be defective. Some of the struggle is my own inexperience with designing passive crossovers. Will post some pics tomorrow.

 

[RickS]

So for starters, here is the ND25 FR measure...

As you can see there is 7 dB swing in the critical midrange and it droops pretty badly over 13 kHz. Anyway, since I have it, trying to make the best of it. it took a 3rd order crossover to get close and so here, is the initial on-axis with the SIG180 midwoofer...

Not too bad but was a struggle to get it to meet the SIG1800 at its limits. Not sure about the heat map just yet, as heat map is showing some artifacts that I need to track down. The bass alignment is 5th order as am using the same cabinet as Directiva and the same SB passive radiator. Not sure if I will use the PR, so not dealing with anything under 200 Hz just yet. Will see if I can duplicate using minidsp and see whether this is a keeper or not.

 

[Ktacos]

I have some nd25fw that might be of value, it's between two different ones and I'm not sure which is which, but one seems to look like yours and the other looks like a better sample. I did note that they measured differently, one was just more ragged than the other. I don't think it was audible but haven't used them in awhile.

The top blue line is the raw tweeter on cnote baffle, ignore everything below it. All the data I have on them is from when I was trying to figure out a problem that ended up being load dependency on my cheap amp. I think this was the better of the two tweeters.

This is a cnote I built compared to PE's golden sample I think, or at least one they had on hand I'd have to check the email to remember. They measured them both in house. Red is mine which looks to be a little closer to your measurement. Had to stretch it because PE likes to smoosh things.

I will say I put some 4" pvc pipe round overs on my cnotes and ran active filtering and it was one of the best two ways I've ever heard. I think the nd25fw does really well with 4-5" drivers, but didn't much care for it trying to meet the sig180 but my cabinet also didn't make it easy, drivers were kinda far apart with even diffraction edge distances all around.

Pro tip - Don't make a bunch of seperate cabinets and put little dowel feet that mate with forstner holes, it will rattle and there's nothing you can put between them to stop it. Also tweeters seem to sound a bit better to me with some baffle, seems my assumption to just minimize the baffle as much as possible didn't quite pay off in the way I thought it would.

 

[eddantes]

Wondering if this dip will be audible?

Maybe a low Q filter to pull down this... a wee bit... Probably will make things worse... I have no idea what I'm talking about.

 

[RickS]

Wondering if this dip will be audible?
View attachment 373144
Maybe a low Q filter to pull down this... a wee bit... Probably will make things worse... I have no idea what I'm talking about.

View attachment 373145

When is comes to minor troughs or peaks, find it useful to listen. If find no major objection, it is likely fine. Not one to throw more parts and cost at a design just to get ruler flat response. This one was much uglier at the start and is pretty flat through the midrange now. That said, still working on the power response. In that regard, may also trade nicer on-axis response for better off-axis.

Also should have noted these measurements are not smoothed. Many published ones tend to look better due to smoothing.

 

[RickS]

I have some nd25fw that might be of value, it's between two different ones and I'm not sure which is which, but one seems to look like yours and the other looks like a better sample. I did note that they measured differently, one was just more ragged than the other. I don't think it was audible but haven't used them in awhile.

The top blue line is the raw tweeter on cnote baffle, ignore everything below it. All the data I have on them is from when I was trying to figure out a problem that ended up being load dependency on my cheap amp. I think this was the better of the two tweeters.

View attachment 373134

View attachment 373135

This is a cnote I built compared to PE's golden sample I think, or at least one they had on hand I'd have to check the email to remember. They measured them both in house. Red is mine which looks to be a little closer to your measurement. Had to stretch it because PE likes to smoosh things.

View attachment 373141

I will say I put some 4" pvc pipe round overs on my cnotes and ran active filtering and it was one of the best two ways I've ever heard. I think the nd25fw does really well with 4-5" drivers, but didn't much care for it trying to meet the sig180 but my cabinet also didn't make it easy, drivers were kinda far apart with even diffraction edge distances all around.

View attachment 373137

Pro tip - Don't make a bunch of seperate cabinets and put little dowel feet that mate with forstner holes, it will rattle and there's nothing you can put between them to stop it. Also tweeters seem to sound a bit better to me with some baffle, seems my assumption to just minimize the baffle as much as possible didn't quite pay off in the way I thought it would.

Cool, I had plans to do an 801 Matrix-like speaker too!

I acquired most of the cabinetry and (most of) the drivers but then learned more about baffle shapes and directivity. A more contemporary design is now the target. That is basically what we are doing with Directiva r2.

 

[Ktacos]

A more contemporary design is now the target. That is basically what we are doing with Directiva r2.

Probably a good idea. As neat as some things can look short term, they often don't have long term appeal and sometimes you find yourself sitting there going "ok, this is too much". I've had to pan plenty of projects as they would just look ridiculous. Big fan of your typical bookshelf and tower now.

 

[RickS]

After several crossover iterations seeing a persistent widening of ND25FW directivity just under 4 kHz. For a smaller woofer that reaches higher it is ok, but not a great match for the SiG180. So, if you are ok with wider directivity, this is about as good as this one will get for now. Am expecting Directiva r2 to be here tomorrow and will be devoting my time to it.

Will try to get in a real measurement tomorrow with active crossover to see how well sim matches.

 

[Ktacos]

That dispersion looks better than most speakers, and I generally prefer wider to more narrow. WG300 made a match with the sig180 but in some ways sounded worse to me, my brain seems to be kind of sensitive to putting drivers back into horns, the depth of the nd25fw seems to be kind of the limit for me.

Most annoying part of the nd25fw is what seems to be some sort of transition at around 7k, things get a bit more narrow. Still a better sounding tweeter than most that I've used.

 

[eddantes]

After several crossover iterations seeing a persistent widening of ND25FW directivity just under 4 kHz. For a smaller woofer that reaches higher it is ok, but not a great match for the SiG180. So, if you are ok with wider directivity, this is about as good as this one will get for now. Am expecting Directiva r2 to be here tomorrow and will be devoting my time to it.

View attachment 373581

Will try to get in a real measurement tomorrow with active crossover to see how well sim matches.

TBH looks pretty dang good to me. Parts count is low. Tilt looks right. I think we might be a bit critical considering this is supposed to be a budget build. Looking forward to real life measurements and impressions.

 

[RickS]

Found I could conjure up the parts to build a passive crossover. I did have to unwind a coil down to 1.5 mH. As passive is the target and converting to active would take at least as long, I built it. It passes the initial test in that it matches the sim well...

Dotted is the sim and the solid trace is the actual. A spin is next but need to do some alignment and calibration as yet.

 

[eddantes]

Well that looks alright doesn't it? So cool to see the model align with measurement - must be very satisfying. To my eye the only question is whether this is a thing or a nothing

 

[RickS]

Well that looks alright doesn't it? So cool to see the model align with measurement - must be very satisfying. To my eye the only question is whether this is a thing or a nothing

View attachment 373772

It still needs something. After a quick listen, need to work on the bass. As mentioned earlier, the SIG180 came up quite shy on the promised Fs. A quick listen with the speaker in the middle of my room, makes this clear. Am a tough critic but also have to temper the difference between a more premium design and a budget one too. Stay tuned!

 

[RickS]

Complete spin of the speaker tracks the sim well with a slight droop around the crossover point. Recall Kimmo said something about a sim adjust that I forgot to make. Will address at the same time as I redo the bass tuning. Will need to do some carpentry before I can....



p.s. lowering the woofer shunt cap to 18-20 uF should fix the directivity dip around the crossover point

 

[Ktacos]

Hi Rick, do you have a sensitivity for the sig180 on this baffle? I was considering pairing it with the Tebm46 bmr, assuming 6db of baffle losses, it should be ok on paper but I wanted to get a second opinion before constructing anything.

 

[RickS]

Hi Rick, do you have a sensitivity for the sig180 on this baffle? I was considering pairing it with the Tebm46 bmr, assuming 6db of baffle losses, it should be ok on paper but I wanted to get a second opinion before constructing anything.

I just switched my test rig over to Directiva r2. Once I get in a spin, should be able to do a sensitivity measure on the SIG180.

 

[RickS]

Hi Rick, do you have a sensitivity for the sig180 on this baffle? I was considering pairing it with the Tebm46 bmr, assuming 6db of baffle losses, it should be ok on paper but I wanted to get a second opinion before constructing anything.

Eyeballing it, would say about 89-90 dB.

 

[Ktacos]

Eyeballing it, would say about 89-90 dB.

Oh, that's definitely too high for my desired use. Is that after filtering?

I was under the impression most woofers lose about 6db when tossed on a typical bookshelf baffle but apparently not? I think I'm missing something here.