Hello all!
I'm designing a 3-way as a spare time project, in a WMTM alignment. The baffle is 30cm wide, 45cm high and the drivers I have at hand are:
Peerless 830868: https://www.samodelka.ru/pictures/peerless/peerless_830868.pdf
VIfa TC9FD: https://doc.soundimports.nl/pdf/brands/Peerless by Tymphany/TC9FD18-08/TC9FD18-08.pdf
Seas 27TFFNC/CG: https://www.costruireaudio.com/PDF/SEAS/h1406_27tffnc_cg_datasheet.pdf
Made a mock-up baffle (open-backed, so disregard everything under 400Hz or so). and attached the drivers to it. I plan to surface mount the drivers and made a similar roundover to baffle edges as the finished enclosure will have. Made some measurements with UMIK outside, so there's a bit of ambient noise form neighbors and birds etc. Also measured the impedances with DATS. The mids are measured separately. The measurements are at 1m distance, at 0, 22 and 45 degrees, at tweeter height. The height from ground is 190cm on tweeter level. For anyone wanting to lend a hand, they are at this onedrive folder: https://1drv.ms/f/s!AoJmX0yPPNWy6QfpAWANbmA_d65c?e=gFeOVe
Designed for a bit in VituixCAD, and this is the best result I got so far. The phases are quite nicely aligned, the only problem I currently see is that the impedance dips quite low in the 10kHz region, around 2Ohm. The reference angle response is a bit hot but realistically they would be listened to around 10-20degree angle anyway. It's been a few years since my last speaker design/build, and I am rusty as hell. Life tends to get in way of hobbies sometimes 🙂
Would really appreciate if you have any tips and points on perfecting the crossover. Thanks in advance!
EDIT: Updated the onedrive along with VituixCAD sim for 2 mids, measured separately.
I'm designing a 3-way as a spare time project, in a WMTM alignment. The baffle is 30cm wide, 45cm high and the drivers I have at hand are:
Peerless 830868: https://www.samodelka.ru/pictures/peerless/peerless_830868.pdf
VIfa TC9FD: https://doc.soundimports.nl/pdf/brands/Peerless by Tymphany/TC9FD18-08/TC9FD18-08.pdf
Seas 27TFFNC/CG: https://www.costruireaudio.com/PDF/SEAS/h1406_27tffnc_cg_datasheet.pdf
Made a mock-up baffle (open-backed, so disregard everything under 400Hz or so). and attached the drivers to it. I plan to surface mount the drivers and made a similar roundover to baffle edges as the finished enclosure will have. Made some measurements with UMIK outside, so there's a bit of ambient noise form neighbors and birds etc. Also measured the impedances with DATS. The mids are measured separately. The measurements are at 1m distance, at 0, 22 and 45 degrees, at tweeter height. The height from ground is 190cm on tweeter level. For anyone wanting to lend a hand, they are at this onedrive folder: https://1drv.ms/f/s!AoJmX0yPPNWy6QfpAWANbmA_d65c?e=gFeOVe
Designed for a bit in VituixCAD, and this is the best result I got so far. The phases are quite nicely aligned, the only problem I currently see is that the impedance dips quite low in the 10kHz region, around 2Ohm. The reference angle response is a bit hot but realistically they would be listened to around 10-20degree angle anyway. It's been a few years since my last speaker design/build, and I am rusty as hell. Life tends to get in way of hobbies sometimes 🙂
Would really appreciate if you have any tips and points on perfecting the crossover. Thanks in advance!
EDIT: Updated the onedrive along with VituixCAD sim for 2 mids, measured separately.
Attachments
Last edited:
I suppose that's because you have a high frequency "short" via the 680 nF and 15 uF cap in the woofer network. Maybe add a resisor in series with the 680 nF cap?
I don't see 2 mids in parallel in the Vituixcad schematic, maybe i have missed something. However you do say the mids were connected in parallel for both measurements?
The two mids will have slightly different measured responses, so maybe you could import separate responses. Its a bit of a moot point, and may have very little if even discernible audible consequence.
I guess you had fun getting the phase to align.
The two mids will have slightly different measured responses, so maybe you could import separate responses. Its a bit of a moot point, and may have very little if even discernible audible consequence.
I guess you had fun getting the phase to align.
🙂
ok, then one more try:
the tweeter gets a 3 dB boost (double energy) around 10 kHz according to the electric response. That will reduce impedance to 50%.
You also need to consider baffle step, which is not included in your measurements because of open baffle. As consequence you will probably need to pad down midranges and tweeter which again may solve your low impedance problem.
Edit: as far as i know vituixcad allows to model/simulate bass response and baffle impact. That would allow to simulate lower end.
Last edited:
Sounds like I need to finish the box and do some remeasuring. I'm set on the front baffle size anyhow because I got them already, just need to slap some plywood together.
I came up with this XO. I didn't like the blended mid measurements so I traced the manufactures data. I did use your woofer measurements. Once you get better measurements I will try again. I made no attempt at modeling diffraction so there is no baffle step compensation.
Attachments
How do you protect the midranges from bass signals? I suppose the tc9s still need a cap in series..
Nice call, it seems they roll off naturally on the low end I forgot to put the high pass in. Here is an updated XO with the high pass on the mid. Obviously this is not a usable XO, just a model with the data given. I do this mostly for fun.
I updated the REW file with measurements and added freq. response files along with impedance files in the onedrive. Now both mids are measured separately. The distance from microphone is 1m, and the height of the tweeter from ground is 190cm.
Tried to model this with the measured responses. The impedance is better, but it has probably a too steep decline towards the highs.
Attachments
There is something off in your mid measurement in the 1-2KHz range as the huge dip is not present on the FR of the Vifa datasheet (dashed line).
Anyway, here's my take: in order to avoid all the reflections I merged the LF part of the woofer (@300Hz) for clarity. I used a 21L closed box for the LF simulation and applied a smoothing of 1/6 octave.
Anyway, here's my take: in order to avoid all the reflections I merged the LF part of the woofer (@300Hz) for clarity. I used a 21L closed box for the LF simulation and applied a smoothing of 1/6 octave.
