Crossover problem

[wintermute]

Major improvement!! 🙂

 

[Siggma]

This is the first one I did for this speaker, before it - all were parallel. The thing is, no matter how well the simulated phase tracking is - when measured - it is never such. Usually in the range of 5-8db dip at xover freq, while the predicting software shows -40db for example. Tried asymmetrical crossovers, symmetrical LR2 - no deep dip at crossover frequency.

I was wondering, should I look elsewhere to obtain good phase tracking or the response I get is normal for inroom measurements with the defined setup. Unfortunately, speakers are very heavy to bring them out and measure there.

Did I read that you're using the Vifa XT25TG tweet? You didn't say in your original post but I thought I saw it referenced. That's a newer design. Nearly all the off-axis sound is reflected off the center nipple so it might align with a symmetric slope. If the crossover design assumption is made with a normal flat cone and this is a reflected, phase aligned design maybe that's the issue?

 

[Mario Pankov]

It turned out to be a software issue which is now fixed. Don`t have the time to revise the crossover and add the RLC circuit for the XT25, bought the parts already but it will need some adjustment.
Problem was I was getting unreliable measurements due to a sound calibration issue.

 

[Siggma]

Hello and happy new year to all!

After I read your post, I double checked that I have the sound card calibration, it was checked. However, when I did a loopback measure, response was ragged as speakers inroom measurement. Unfortunately, it appears to be a Holm problem and I fear the inconsistency and weirdness of measurements comes from it. Anytime I restart Holm, calibration is lost, even if applied in the calibration panel.
When I did try to recalibrate, it does it and during the information analysis crashes. The only way around I found was to let it crash and then go to the measurements where it has DAC-ADC loopback on the blue panel, recheck that it is calibrated and then go ahead. Picture looks normal now with smooth response and more realistic phase plot. I use the 64 bit version on Win 7 64, not sure if this would be the problem.

Might be too late but it sounds like a folder permission issue saving HOLM settings. Try running HOLM as Administrator (right click and select "Run as Administrator") and see if the issue disappears. Windows 7 does some gyrations to folder names to link them together and sometimes programs can't save to their own installation directory. You might also try reinstalling it using "Windows XP" as the software version target.

Hope this helps.