The Joudas project

[Chris Roemer]

I think you know what you're doing. Our diffs are prob. due to diffraction / BSC / and room gain.
Unless you're lifting those monsters up off the floor, you shouldn't need any BSC.

[trevordj]

I think you know what you're doing. Our diffs are prob. due to diffraction / BSC / and room gain. Unless you're lifting those monsters up off the floor, you shouldn't need any BSC.

I appreciate the vote of confidence. I’ll report back after I make some actual frd measurements and hopefully I can get something usable. The measurements are still intimidating for some reason, just something new but it will be an opportunity for learning. These things are definitely staying on the floor! These boxes weigh almost as much as my sub boxes 😐 (without the HST-18s of course) -Trevor

[rpb]

I used the files in post 20 and 21. Assumed zero offset, and crossed about where you did. (I don't know how low it's safe to cross this tweeter.)



These drivers seem to be easy to work with. You should be able to target most any response needed if you feel the need to alter tonal balance. My tweeter filter was second order with a little shaping. It could be simpler, but resistors don't cost much, and they help tilt the response. My sim is at 92dB. I see yours is at 95dB. When I looked at the woofer response, and a baffle diffraction sim, I saw a 2dB rise around 400hz. Other than that, most of the baffle effect was below the range of this x-over anyway. I did not alter the files. I just looked at a generic sim with the same dimensions as your boxes. The woofer response follows a LR 4th at 1.7k, and needed only a 3rd order filter.

[rpb]

Some advice. Since you are new to this..... To protect your tweeters specifically. I think I heard that ribbons are fragile! (I never used a ribbon.)

Stuff can happen when you measure. To stay on the safe side, I'd recommend never sweeping the tweeter without an estimated xo that includes padding.

In addition, I'd practice on some inexpensive tweeters first. And remember... Your tweeter will be much louder, so turn the amp down very low.

[trevordj]

Some advice. Since you are new to this..... To protect your tweeters specifically.

Stuff can happen when you measure. To stay on the safe side, I'd recommend never sweeping the tweeter without an estimated xo that includes padding. Start with one crossed much higher than needed, and approach the target in small steps. Now, I know that you can't make a new frd file to sim with from your measurements, but you can put the measurement in your software, and compare it to a target curve. With a few adjustments, you should get close pretty quick.

In addition, I'd practice on some inexpensive tweeters first.

Right on, thank you for all of that information. It's good to see we are both on a similar track with the initial sims. That's great advice regarding the tweeters. I have some Vifa XT25s that I can put in a baffle and play with. I am actually going to be potentially using those in a SEOs 8 waveguide crossed with some AE TD8Ms for my surrounds. I already have an adapter I made to fit the tweeter flush to the waveguide and I am going to need to measure them anyway to see if they will play nice in the waveguide. I could potentially kill two birds with one stone by measuring those first so I don't blow up a $400 tweeter.

[trevordj]

Some advice. Since you are new to this..... To protect your tweeters specifically. Stuff can happen when you measure. To stay on the safe side, I'd recommend never sweeping the tweeter without an estimated xo that includes padding. Start with one crossed much higher than needed, and approach the target in small steps. Now, I know that you can't make a new frd file to sim with from your measurements, but you can put the measurement in your software, and compare it to a target curve. With a few adjustments, you should get close pretty quick. In addition, I'd practice on some inexpensive tweeters first.

The tweeters, adapter, and waveguide:

-Trevor

[trevordj]

While I am waiting to get my act together to take some measurements I am working on checking another box off of the to do list, designing PCBs for the crossovers. I am still working with practice values, obviously, but the learning curve is steep. I work with Fusion 360 for all of my cabinet renderings, CAD/CAM for my CNC, and recently they integrated many of the functions of Eagle into Fusion 360 directly. The thing that's taking the longest is doing the 3D designs for all the passives (really its the inductors) and building my local parts library. There are a lot of nuances that aren't obvious with the electronics plugin but I think I'm getting it. This is not a complete crossover or schematic, just proof of concept that I can do what I need to do to get the crossover PCBs made.

[trevordj]

I am going to start measurements of the LCRs. I got the drivers mounted in the box today and took and impedance sweep which I’ll post from my computer in a minute. It looks like I have some impedance spikes that I think are from box resonance. I’m going to take the driver out and mount some OC703 to see if I can improve them. I didn’t have time to do any frequency sweeps so I am going to try that tomorrow. -Trevor

[trevordj]

Impedance sweeps:

[rpb]

Did you practice with the cheaper drivers?....... Be careful!

This is the fun part of speaker design. I enjoy xo work much more than wood work.

[trevordj]

Did you practice with the cheaper drivers?....... Be careful! This is the fun part of speaker design. I enjoy xo work much more than wood work.

I have a baffle built for the cheaper tweeter. These were just impedance sweeps with DATS which I had already done free air. I’m going to start messing with some sweeps on the cheaper tweeter and the woofer today. If all goes well I will take a measurement with the Beyma too. -Trevor

[trevordj]

I ended up having to do a project for a family member this weekend so I didn't get much time with these. I figured I would share my test rig I came up with in the mean time.

I designed a simple baffle in Fusion 360:



I got everything cut and I just need to add some roundovers and cut a few rabbets so I can glue it up. It should only take a few minutes. I'll post pictures of the part from my phone in a minute.

[trevordj]

I had to cut this a couple times to get the fit correct. I am using a Vifa XT25TG30-04 with a custom made adapter bolted to a SEOS8 waveguide. I am also using a Vifa NE180w-08 woofer i have had kicking around before it was renamed the Peerless NE180w.

https://www.parts-express.com/peerless-by-tymphany-xt25tg30-04-1-dual-ring-radiator-tweeter--264-1016

https://www.parts-express.com/peerless-by-tymphany-ne180w-08-6-1-2-fiber-cone-woofer-speaker--264-1134

-Trevor

[trevordj]

Alright, the testing rig is ready to go. I’m getting ready to dive in with some test measurements. Wish me luck. -Trevor

[trevordj]

Ok, I got my first measurements with the test rig today. I think I am going to approach this as if I was making a crossover. I need to answer the question (for myself): what all do I need to do to get accurate measurements for crossover design? Hopefully I can do this without blowing anything up, so far so good. I am making measurements in room at 1 meter and approximately 1 watt of power. I limited the tweeter sweep 100hz-20,000hz and the woofer I ran full range. I did all measurements without changing any input or output settings. My microphone is calibrated for SPL so the sensitivity should be both relatively and absolutely accurate.

My measurement gear includes a Steinberg UR22C sound card, an Audix TM1 microphone, and an Emotiva UPA-2 amplifier. I have calibrated my soundcard in REW and am using a loopback timing reference using the right channel to input 2. Input 1 is my microphone which was calibrated by CSL. I have calibrated my SPL measurements in REW using a galaxy CM-170 SPL meter which was also calibrated within the last month.

I am going to go through the Vituixcad user manual tomorrow which has a really nice guide to getting measurements in REW. I tried HOLMImpulse but I don't see a way to do a loopback timing reference with that program (which seems to be important based on what I recall) so I will be sticking with REW. Other than that I think my goal for tomorrow will be to get polar measurements. Does that sound reasonable? Does my measurement strategy sound reasonable up to now? I am still not really sure if I should be doing measurements in room at the seating position vs. 1m vs. nearfield vs. 4 pi space outside.

Each of these measurements were done 4 times, they were not averaged. Each of them have 1/12 octave smoothing.

Woofer:



Tweeter: