Looks like the woofer is lagging. But what does the phase look like from the tweeter only. Is there a slowly rising phase or is it flat. Without a crossover I mean.
It looks like you'd need some textbook examples from BYRTT to guide you.
Great pictures of bushmeister's progress and xrk971's new devices last knight also measurements was some fantastic
When woofers gets a sealed back had some thinking it maybe be wise to look in REW measurements and impedance sweep after resonance that fourth order symmetrical band pass add in sims, because if it gets there in real world think it be worth find frq point and Q and throw a PEQ in to repair. See picture 1 what i mean and see picture 2 how much error it add compared a clean 80-1300Hz BW2 pass band, did also try add a electric low pass BW2 at 400Hz to worst scenario in picture 2 but it don't get much better without cleaning that resonance.
When woofers gets a sealed back had some thinking it maybe be wise to look in REW measurements and impedance sweep after resonance that fourth order symmetrical band pass add in sims, because if it gets there in real world think it be worth find frq point and Q and throw a PEQ in to repair. See picture 1 what i mean and see picture 2 how much error it add compared a clean 80-1300Hz BW2 pass band, did also try add a electric low pass BW2 at 400Hz to worst scenario in picture 2 but it don't get much better without cleaning that resonance.
Attachments
Last edited:
Great pictures of bushmeister's progress and xrk971's new devices last knight also measurements was some fantastic
When woofers gets a sealed back had some thinking it maybe be wise to look in REW measurements and impedance sweep after resonance that fourth order symmetrical band pass add in sims, because if it gets there in real world think it be worth find frq point and Q and throw a PEQ in to repair. See picture 1 what i mean and see picture 2 how much error it add compared a clean 80-1300Hz BW2 pass band, did also try add a electric low pass BW2 at 400Hz to worst scenario in picture 2 but it don't get much better without cleaning that resonance.
Thanks - Making the enclosures is a long way off yet - I have got three more woofers to fit first.....
But as soon as I have - I will be enlisting your expertise!
While it's worth a listen, there's something still a bit off.
I expect the STEP to fly above the zero line for about ~3 ms to indicate good behaviour. But lot's of causes could make it deviate right now. Can you plot the combined impulse as well?
Ideally the phase would look like this plot X showed us:
So I think the woofer is still lagging a bit. But keep in mind the above plot is using ideal responses.
Your tweeter might already have a slowly climbing phase response in the horn. The textbook x-over is based on pure flat phase from the drivers before crossover.
I expect the STEP to fly above the zero line for about ~3 ms to indicate good behaviour. But lot's of causes could make it deviate right now. Can you plot the combined impulse as well?
Ideally the phase would look like this plot X showed us:
So I think the woofer is still lagging a bit. But keep in mind the above plot is using ideal responses.
Your tweeter might already have a slowly climbing phase response in the horn. The textbook x-over is based on pure flat phase from the drivers before crossover.
Last edited:
Definitely worth a listen though. See if certain drum hits jump out more than with the Linkwitz x-over. To get what we want we need to hit that perfect slope as best we can.
Keep an eye on the FR and phase for the individual drivers. The woofer is the hard part as the entry holes give it a band pass with a resulting change of phase. That's where knowing what to aim for will be handy.
Keep an eye on the FR and phase for the individual drivers. The woofer is the hard part as the entry holes give it a band pass with a resulting change of phase. That's where knowing what to aim for will be handy.
Here HARSCH at 500Hz with system bandwidth set as BW2 60Hz - BW2 22kHz.
Bushmeister think your SR is decaying to fast and going into negative area and wonder if drivers is by accident in reverse polarity or it is because there need to be one more symmetrical woofer for wave front to fire real forward.
Bushmeister think your SR is decaying to fast and going into negative area and wonder if drivers is by accident in reverse polarity or it is because there need to be one more symmetrical woofer for wave front to fire real forward.
Attachments
Doesn't tuning the XO seem a bit premature considering there's 180 reverse phase data in these measurements given the testing without an enclosure and rear wave damping?.....not to me mention the future addition of three more woofers. Lol
Proof of concept is there though......this can be done in a compact package!
Proof of concept is there though......this can be done in a compact package!
Last edited:
LOL! You're probably right mayhem!
I just can't help myself! As you say - it is more just a proof of concept as it seems like such a nice point source design with excellent controlled directivity, I just want to see if I can get a nice step and phase too!
Enclosures will take weeks/months to build as it is so cold and wet here and I will need to do it outside. Plus my missus isn't too keen on the 'alien wrapped in blankets' in the study, so I am making the most of this opportunity!
The SR Byrtt shows here is what yours should look like when correct. I think there is additional delay and the woofer is lagging behind quite a bit. If you look at SR of each driver individually you will see the large woofer "hump" needs to move to the left by almost 4ms. So you are
off by multiple cycles.
The most accurate way I know to find the acoustic offset is to measure it by interferometry from the mic position at say 0.5m away. Measure each driver response individually and connected in parallel. All running full range. The interferogram you get with both drivers is then the target response in a XO sim program like PCD or Xsim. Load all 3 minimum phase FRD files into PCD or Xsim, the adjust the physical driver offset in the sim until the simulated combined response matches the measured combined target. It's amazing when it does because it will get you to sub mm accuracy of the offset. A bit more work but worth it. The other advantage is that the raw FRD files are then massaged with EQ and XO's in PCD to match the desired Harsch target curves of BW24 and Bessel12.
Then transfer the PEQ sims from PCD to miniDSP.
But all of this needs to happen after the other woofer goes in and the horn gets a baffle. So glad you are testing out the waters with the Harsch XO! LOL
Last edited:
Now we are talking! You should be able to hear that change.
If you can find it, try and listen to Harry Connick Jr with The Last Payday.
Listen to that pool shot. It should sound scary real with a good impulse.
I'd also look at what a PEQ at this 1/6 octave smoothing scale would do for phase and frequency targeted at that little 450 Hz bump.
If you can find it, try and listen to Harry Connick Jr with The Last Payday.
Listen to that pool shot. It should sound scary real with a good impulse.
I'd also look at what a PEQ at this 1/6 octave smoothing scale would do for phase and frequency targeted at that little 450 Hz bump.
Horizontal plot is same as distortion plot, and no there are no other distortion point except when put higher volume but then mic amp distorting.
the hybrid amp has also most even harmonics. I have put on 0 dB as reference so it is right measurement.
regards
Before I put my mic away - does anyone want any more measurements?
I have done polars, distortion, impulse, step, and phase.
I think, given this is a single woofer, running wrapped in blankets, that we have an excellent proof of concept here.
Good CD down to about 500hz horizontally, about 800hz vertically, low distortion, high sensitivity, flat phase, nice step, point source behavior without any vertical lobing etc...
Anything else anyone wants me to look at before I disappear into a listening session?
I have done polars, distortion, impulse, step, and phase.
I think, given this is a single woofer, running wrapped in blankets, that we have an excellent proof of concept here.
Good CD down to about 500hz horizontally, about 800hz vertically, low distortion, high sensitivity, flat phase, nice step, point source behavior without any vertical lobing etc...
Anything else anyone wants me to look at before I disappear into a listening session?