Thank you for taking your time, your estimation isn't that off, but it's more like 160x350, quite a narrow baffle...
It's indeed for baffle compensation. The LR was too broad for this application, it was lowering the spl below 500hz and I didn't want that. If you look at my raw measurement for the woofer I think you'll see why:
I am almost certain it is not an artifact. There was another driver installed in this cabinet which showcased a similar response around that area. For reference, I measured other speakers at that exact spot, and I couldn't see anything similar.
This is the FR measurement of the driver made by hificompass.
The woofer drops 3dB's from 1k till 700hz, combined with the baffle loss, I think it kinda makes sense why my measurement looks like this.
As for the near field measurements, I didn't really feel the need since at lower frequencies the room becomes more important than the speaker as you'll probably agree, and I use some room eq in almost every circumstance. But of course, it would be nice to see the overall picture and give a better idea about what's happening. All in all, I think I am just gonna trust my measurements for now. I believe I can live with a few dB's of deviation
A couple of things about breakup. One is that it usually peaks on axis but that's not representative of the overall response. Don't rely on an on axis measurement to tell you the story, and don't listen on axis...
Secondly, you shouldn't be using the speaker through breakup. It should be fully in the stopband of the crossover. I understand it gets tight when you use a waveguide.
This situation should be dictating to you the needed slope. The original circuit in this thread just happened to be one way to get a sharp turnover of the response.
"Barely noticeable" doesn't mean that improvement isn't possible, however subtle.
In a proper passive XO, any component change or addition has an inexorable affect upon other components & sound.
Thank you all for your opinions. Many people particularly suggested using a 3rd order filter on the woofer, so I have revised the design and implemented a 3rd order lowpass. The high pass filter had to be changed from 3rd to 2nd order (which I guess is even better since I am not going to use expensive caps)
However!
In order to get a decent FR and phase, I needed to push the xo point down to 1.65k. The tweeter's Fs is 900hz and I am using a waveguide which theoretically should make it work. (According to my measurements, the WG boosts about 5-6 dB's around that region) The thing is though, the filter is only 2nd order and I don't know if that's gonna be alright. How would you interpret this? Has anybody here crossed that low? (in a waveguide or as is) what are the pros and cons?
You can check the tweeter's specs from this link if you like: https://www.parts-express.com/pedocs/specs/277-010-morel-mdt-29-specifications.pdf
However!
In order to get a decent FR and phase, I needed to push the xo point down to 1.65k. The tweeter's Fs is 900hz and I am using a waveguide which theoretically should make it work. (According to my measurements, the WG boosts about 5-6 dB's around that region) The thing is though, the filter is only 2nd order and I don't know if that's gonna be alright. How would you interpret this? Has anybody here crossed that low? (in a waveguide or as is) what are the pros and cons?
You can check the tweeter's specs from this link if you like: https://www.parts-express.com/pedocs/specs/277-010-morel-mdt-29-specifications.pdf
Last edited:
Thanks for letting me know, I actually rarely exceed moderate levels but I believe it's good to be on the safe side regarding the dynamics, headroom and what not.
I think I've exhausted my options, neither 3rd order high pass nor 4th just doesn't work when the woofer is also 3rd order.
How did you do your measurements of the drivers and did you move the microphone for each driver measurement?
If so you will have subtle phase changes that maybe not helping you. If the tweeter is some cm back in the horn relative to the woofer cone sometimes it helps sometimes it doesn't.
One relatively easy way of avoiding this is and to save the entering of x, y, and z, co ordinate for each driver is to do the measurements of both drivers with one fixed microphone position, usually the listening axis of the tweeter.
I thought it was worth mentioning just in case it could help.
If so you will have subtle phase changes that maybe not helping you. If the tweeter is some cm back in the horn relative to the woofer cone sometimes it helps sometimes it doesn't.
One relatively easy way of avoiding this is and to save the entering of x, y, and z, co ordinate for each driver is to do the measurements of both drivers with one fixed microphone position, usually the listening axis of the tweeter.
I thought it was worth mentioning just in case it could help.
I've taken 3 measurements from the same spot without moving the microphone or the speaker (1 meter away from just below tweeter axis). First I measured the drivers individually, and then I hooked them up in parallel and take a combined measurement. Then I've loaded the .frd files in xsim and tweaked the mod delay until the xsim's summed up response matches my combined measurement. (I think it's called inference method)
Assuming my measurements are reliable, what would you say about crossing this tweeter at 1.65k in a waveguide? https://www.parts-express.com/pedocs/specs/277-010-morel-mdt-29-specifications.pdf
thanks
That is a tricky question as my experience with waveguides is zero. Consequently I have no measure on how much margin they give you over the normal dome.
I think as you and others have pointed out you have several dBs in hand due to the gain of the horn.
The spec sheet for your tweeter gives the resonance at 1Khz you may have to deal with that to get the roll off to protect your tweeter. If you don't intend to listen loud it could work within those constraints.
You could build the intended tweeter crossover and do measurements looking at response and distortion as you gently increase the volume levels.
Back to the C6 question.
I have seen the responses in the Solid state forum. I think one of the responses suggested that with correctly designed commercial amps with the correct output circuits they should be immune. Somebody also mentioned a link to a D Self Book and page reference where i imagine the idea of how to protect a solid state output stage with the correct zobel characteristics.
There are many compromises in loudspeaker design you have a couple of solutions that can work. Once fully assembled i am sure you will adjust one or two elements before you are happy.
If you blew a tweeter or both of them can you afford to easily repair or replace them?
Another pain in the modern world is that some manufacturers are removing certain models or no longer supporting the DIY element. It may be sensible to see items are not close to end of life in terms of future availability.
I think as you and others have pointed out you have several dBs in hand due to the gain of the horn.
The spec sheet for your tweeter gives the resonance at 1Khz you may have to deal with that to get the roll off to protect your tweeter. If you don't intend to listen loud it could work within those constraints.
You could build the intended tweeter crossover and do measurements looking at response and distortion as you gently increase the volume levels.
Back to the C6 question.
I have seen the responses in the Solid state forum. I think one of the responses suggested that with correctly designed commercial amps with the correct output circuits they should be immune. Somebody also mentioned a link to a D Self Book and page reference where i imagine the idea of how to protect a solid state output stage with the correct zobel characteristics.
There are many compromises in loudspeaker design you have a couple of solutions that can work. Once fully assembled i am sure you will adjust one or two elements before you are happy.
If you blew a tweeter or both of them can you afford to easily repair or replace them?
Another pain in the modern world is that some manufacturers are removing certain models or no longer supporting the DIY element. It may be sensible to see items are not close to end of life in terms of future availability.
fair enough
also fair enough : )
Yeah, my first proper build and couldn't agree more, it's all about compromises... If I blew a tweeter that would be really bad because even if I can somehow afford it, diy speaker building isn't really a thing in my country, which makes it kinda impossible to find quality drivers and other components. I ask friends living abroad to bring me drivers and what not when they are visiting.
That's reassuring, thanks. I am also curious about how it's going to sound because I have never heard a tweeter crossed that low.
I think you will have seen post 37 by ianbo over in the associated solid state questions Problems associated with capacitive load. Here is the link he pointed to.
https://purifi-audio.com/blog/app-notes-2/low-distortion-filter-for-ptt6-5x04-naa-11
It also highlights a very useful feature of VituixCad and some clever thinking to enable a better understanding of network behaviour by moving the source to the driver end. Very smart.
I think Xsim will allow you to model these filter combinations you could try the "series notch" made up of parallel components to remove your low impedance worry and maybe the phase will behave for you which would be a double win. Good luck.
https://purifi-audio.com/blog/app-notes-2/low-distortion-filter-for-ptt6-5x04-naa-11
It also highlights a very useful feature of VituixCad and some clever thinking to enable a better understanding of network behaviour by moving the source to the driver end. Very smart.
I think Xsim will allow you to model these filter combinations you could try the "series notch" made up of parallel components to remove your low impedance worry and maybe the phase will behave for you which would be a double win. Good luck.
A lot of people are afraid of crossing tweeters low. Most of the time the distortion profile is the determining factor, not the possible damage to the tweeter. If I’m right, the Morel does fine around 1,5k.
Yes, it was indeed very helpful. I ultimately ended up building a design implementing that parallel/series notch thing. I used a 2nd order filter on the woofer, which gave me the option of crossing at about 1.95K with good results.
I don't think crossing at 1.65k LR2 in a waveguide was going to create problems (especially at my preferred listening levels). If I was using a bigger woofer like 7 or 8inch, I'd definitely cross it there, but I believe it's the bare minimum this tweeter can handle. Considering the fact that the tweeter begins to show some signs of ringing around 1.5k, I think 1.95k 3rd order is the better choice.
Just wanted to let you know that I built a prototype crossover and it measures pretty good I guess. Thank you all for your feedback.
By the way, does anybody have a solution in mind regarding that drop around 12k? (courtesy of the waveguide) Like maybe a phase plug or something?
By the way, does anybody have a solution in mind regarding that drop around 12k? (courtesy of the waveguide) Like maybe a phase plug or something?
Last edited: