DIY 2-way Visaton - 8" WS20E + 1" G25FFL + WG WG148R

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi, I am litlle bit playing with measuring of this 2-way Visaton combination - 8" woofer WS20E + 1" G25 FFL tweeter with Waveguide WG148R fitted on it. Now both drivers are placed in test enclousure (36L volume), both drivers are countersunk into front baffle and all front edges have radius.
So I was trying to measure both drivers independently in enclousure by using HolmImpulse. (off-axis horizontal measurement between 0-75 deg after 15deg.). Then I've tried to simulate/assemble first version of crossover (see 2nd pic - simulated using by Xover 2.0.4 - LoudspeakerSoft - Download - with quite nice deuce with Holm measurement made on real crossover). I was inspired by TQWT project. Not sure about it, if is it possible to use it in my project.
Anyway, please look at my HolmImpulse data on 4th picure, if they are correct ? I had previously some problems with pre-ringing during measuring, but now it seems ok.....
Thank you for any comments, hints.

Note: sorry for my english, I'm trying my best :)
 

Attachments

  • WS20_test_box1.jpg
    WS20_test_box1.jpg
    209.2 KB · Views: 645
  • test_xover_vs_holm_resize.png
    test_xover_vs_holm_resize.png
    226.1 KB · Views: 622
  • test_xover_details_resize.png
    test_xover_details_resize.png
    387 KB · Views: 591
  • holm_mereni_detail.png
    holm_mereni_detail.png
    46.3 KB · Views: 583
Just another Moderator
Joined 2003
Paid Member
Hi Pet007 I'm wondering about your time zero on the holm measurements. Also it looks like something (could be the mic preamp, or could be the sound card) is inverting the signal, not a biggie but probably good to set the option in holm (I think on the first tab) to invert the signal.

With your Time zero off, I think the phase data may also be off which would affect your simulation.

If you want to check whether your measurements are good, then measure each driver individually as you have done, and then without moving anything measure with the real crossover in circuit. Then simulate the actual crossover and compare the result of the sim to the actual measurement.

If they are close then you have got measurements you can rely on for simulating different versions of the crossover.

With the Time Zero, I always set it (click options for the measurement, and choose the auto time zero (I usually use highest positive peak), then do it a second time) I then go to the second tab and tick the lock time zero button (use last detected). I generally set the time zero for the tweeter.

I take the measurements on the tweeter axis and do not move anything for doing the woofer measurement. If you do this your measurements have all offset information built in, so you do not need to specify ant offsets in the simulation program.

You can of course do it the other way and measure on axis with each driver and then put the offsets into the sim program. I just find it easier (and I suspect more accurate) to do it as I have related.

edit: I've attached an example of my tweeter and woofer measurement, note how the peak on the tweeter lines up pretty much with 0cm. I've also turned on the phase, see how it is fairly constant slope across the measurement.... If time zero is not right the phase will look very different.

Tony.
 

Attachments

  • mtm_measurements.png
    mtm_measurements.png
    32.4 KB · Views: 148
Last edited:
Hi, thank you for your reply !
About zero time in Holm... I had my microphone aprox. 7 cm under the tweeter axis (on the botom edge of waveguide, see attached photo). It means there must be some offset for tweeter too, but for phase measuring between drivers it's IMHO no problem. I did lock zero time after first (second) tweeter measurement, then switch to another slot and measure woofer's repsonse.
Ad first negative peak. First peak should be all the time positive ? It means the speaker is generating positive signal (voltage) at first moment, after input signal comes to speaker terminals ? => I have to measure positive voltage peak ?
 

Attachments

  • IMG_20140221_061540_resize.jpg
    IMG_20140221_061540_resize.jpg
    364.9 KB · Views: 155
Just another Moderator
Joined 2003
Paid Member
Hi Pet007 sorry I missed your reply before. Yes for crossover modeling I think you will be fine (for relative phase between drivers) with the mic 7cm below the tweeter. However you probably should measure at what will be the normal listening axis with respect to the tweeter, as vertical lobe will change off axis, and you probably want to use on axis (vertically) measurements for that.

It doesn't really matter whether your first peak is negative or positive, but it generally should be the largest one. I like to adjust the sound card to result in the measured impulse being positive, neagtive just means 180 degrees of phase shift the other way. But your graph that you posted shows your first peak at around 10cm which definitely isn't right.

I've attached a measurement of my MTM's (same drivers as the first measurements but with the crossover) one with the impulse offset 10cm. You can see how it makes the phase information wrap multiple times. You really don't want that. Much better to have phase only rolling over once or twice over the measurement range :) Note that in this one the first impulse is actually negative. measurement taken outside at 1M, and 1/8th octave smoothed.

Tony.
 

Attachments

  • mtm_no_offset.png
    mtm_no_offset.png
    54.3 KB · Views: 139
  • mtm_offset.png
    mtm_offset.png
    57.6 KB · Views: 115
Thanks, perfect. I built yesterday another version of crossover with baffle step compensation (not sure if it's connected well - based on theory) and then I was listening about two hours and must say, it was good for me. The bass range was much better then before even on low volumes (I listen world music, ambient, dubstep, jazz etc.). On attached picture is crossover simulation/measurement in Holm (I can see only one phase change on crossover point).
 

Attachments

  • T80_T100BFC_zapojeno_porovnani.png
    T80_T100BFC_zapojeno_porovnani.png
    112.5 KB · Views: 226
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.