Hmm, I printed it off, and I measured the ports to be 3 inches x 1.4 inches. I think on yours it scaled to something like 3 x 2.3...
This looks very promising X! Make them out of wood and put them on top of your slot loaded woofs. will make a hell of a speaker.
My mistake, the PDF grid is correct - if you count the squares it is 2.4 in wide. I wrote down the wrong first digit as 1 instead of 2. Sorry about that.
I need to update the drawing and ask an admin to swap the attachment.
You're welcome! I'll help in any way that I can...hopefully the more I can contribute will make this project more accessible to more people who want to build the Trynergy! 🙂
I noticed that you'd said the Elmer's foam core board was harder to cut. That's what I'd been using from the outset, so my perception of it was possibly a little skewed 🙂
I noticed that you'd said the Elmer's foam core board was harder to cut. That's what I'd been using from the outset, so my perception of it was possibly a little skewed 🙂
I switched back to the lightweight cheap $1 FC from the dollar store for the woofer enclosures and they cut like butter with a new utility knife. The paper is thinner and crunchier - cracks and splits easily. It's also about 1/4th the weight and price. 🙂
Trynergy in Stereo
I finally finished the second Trynergy and am listening to them in stereo. Have not had much of a chance to test out yet as it is late here, but first couple of tracks sound as I expected, fantastically magical! Great imaging and clarity.
I will report more with measurements and sound clips later...
I finally finished the second Trynergy and am listening to them in stereo. Have not had much of a chance to test out yet as it is late here, but first couple of tracks sound as I expected, fantastically magical! Great imaging and clarity.
I will report more with measurements and sound clips later...
Attachments
Trynergy sounclips in Stereo
Here are some clips of the new stereo setup. The EQ is still very rough and time alignment has not been optimized yet.
00 - dave brubeck
01 - dire straits
02 - sting
03 - fiona apple
04 - anne bisson
05 - live at small's zaid quartet nasser
Here are some clips of the new stereo setup. The EQ is still very rough and time alignment has not been optimized yet.
00 - dave brubeck
01 - dire straits
02 - sting
03 - fiona apple
04 - anne bisson
05 - live at small's zaid quartet nasser
I only have the small PC Speaker in the workshop here ... Are those recordings from a mic? Which one are you using?
Hi Sabbelbacke allow me to comment.
Having followed this thread and the sound clips, these recordings the lowest audio performance from the horns when listened at my setup.
Now why, because the miniDSP in a stereo setup run out of channels we now only listen the horn themselves 2-way and their steep slope fading in at 80-100Hz i guess. Same as xrk971 wrote time was late and EQ fiddling and delay alignment between mid woofers and fullranger not optimized.
Go backward this thread and find seen from my view much better sound clips where miniDSP run 3-way with dedicated sub and we get a full range sound clip though in mono.
As recorder think xrk971 use a dedicated small handheld stereo device, imagine after rest and some calibration fiddling we get better sound clips performance 🙂.
Having followed this thread and the sound clips, these recordings the lowest audio performance from the horns when listened at my setup.
Now why, because the miniDSP in a stereo setup run out of channels we now only listen the horn themselves 2-way and their steep slope fading in at 80-100Hz i guess. Same as xrk971 wrote time was late and EQ fiddling and delay alignment between mid woofers and fullranger not optimized.
Go backward this thread and find seen from my view much better sound clips where miniDSP run 3-way with dedicated sub and we get a full range sound clip though in mono.
As recorder think xrk971 use a dedicated small handheld stereo device, imagine after rest and some calibration fiddling we get better sound clips performance 🙂.
Last edited:
Surely with the equivalent cone area of a 12" per side, it shouldn't be difficult to get bass down to, say, 60Hz at moderate levels.
Chris
Chris
I am using a Zoom H4 recorder. The band pass nature of the bass injection ports and the relatively weak motor in the woofers will not let much bass happen below 90Hz without a lot cone excursion. I was hoping that the recordings would be able to give a sense if the stereo imaging and dynamics of stuff above 90Hz. You have to be imaginative and ignore the lack of bass supplied by the sub in previous recordings. As Byrtt mentioned, without the sub I may be better off in mono for sound quality. However, the stereo gives soundstage for many recordings. More tweaking of the EQ and time alignment generally yields something better.
New XO, EQ & Time Alignment
I think I was able to make some good progress today with the XO, EQ, and Time-Alignment. It appears that flipping the phase of the 3FE22 and increasing delay between the two to 1.50ms makes all the difference. I also shifted the XO point to 375Hz. The group delay looks much better now with a small 2ms anomaly at 500Hz but otherwise sub 2m from 100Hz on up. The minimum phase is pretty flat as well. Harmonic distortion looks pretty good but starts to climb due to cabinet resonance below 150Hz. The consistency between unit 1 and 2 is also very good (using same EQ settings). Overall, I think I am much happier with how this is turning out. It is approaching a true phase coherent point source from 100Hz on up. No time for careful listening yet, but the sound is better (less phasiness and soundstage is stable). I still need to do some work to see if the phase anomaly at 500Hz can be eliminated.
Here is the frequency response for both Trynergies showing the new XO (ignore the low freq's for the magenta curve - the HVAC kicked on):
Here is the minimum phase (min-phase) for horn No. 1 (there is a single wrap at 500Hz - not shown for clarity):
Here is the min-phase for horn No. 2:
Here is the harmonic distortion (HD) for horn No. 1:
Here is the HD for horn No. 2 (note the curve below 100 Hz is different than No. 1 - I forgot to add polyfill stuffing to the rear chamber of the woofers, so the Q is higher - also evidenced by higher HD it appears that the stiffer Elmer's FC board did not make the horn have any better HD performance - my recommendation is to use the cheap $1 stuff for easier building):
Here is the impulse response (IR) and step response (SR) for horn No. 1:
Here is the IR and SR for horn No. 2:
Here is the group delay (GD) for horn No. 1:
Here is the GD for horn No. 2:
I think I was able to make some good progress today with the XO, EQ, and Time-Alignment. It appears that flipping the phase of the 3FE22 and increasing delay between the two to 1.50ms makes all the difference. I also shifted the XO point to 375Hz. The group delay looks much better now with a small 2ms anomaly at 500Hz but otherwise sub 2m from 100Hz on up. The minimum phase is pretty flat as well. Harmonic distortion looks pretty good but starts to climb due to cabinet resonance below 150Hz. The consistency between unit 1 and 2 is also very good (using same EQ settings). Overall, I think I am much happier with how this is turning out. It is approaching a true phase coherent point source from 100Hz on up. No time for careful listening yet, but the sound is better (less phasiness and soundstage is stable). I still need to do some work to see if the phase anomaly at 500Hz can be eliminated.
Here is the frequency response for both Trynergies showing the new XO (ignore the low freq's for the magenta curve - the HVAC kicked on):
Here is the minimum phase (min-phase) for horn No. 1 (there is a single wrap at 500Hz - not shown for clarity):
Here is the min-phase for horn No. 2:
Here is the harmonic distortion (HD) for horn No. 1:
Here is the HD for horn No. 2 (note the curve below 100 Hz is different than No. 1 - I forgot to add polyfill stuffing to the rear chamber of the woofers, so the Q is higher - also evidenced by higher HD it appears that the stiffer Elmer's FC board did not make the horn have any better HD performance - my recommendation is to use the cheap $1 stuff for easier building):
Here is the impulse response (IR) and step response (SR) for horn No. 1:
Here is the IR and SR for horn No. 2:
Here is the group delay (GD) for horn No. 1:
Here is the GD for horn No. 2:
Attachments
Last edited:
Minphase plot does not tell much to me, unwrapped regular phase is more informative. The total phase shift is what matters, and the possible minor/major "bumps" in the unwrapped plot. Change (shorten) the time window to get clean phase plot + close range measurement is good.
You have tuned the midwoofer to slight "forward lean" compared to fullranger (= negative EGD, can also be seen in step response). I also tried this kind of arrangement by delaying the comp driver from "optimal" and reversing it's phase. This can work, but I'm not sure if it's phase accurate setting by-definiton. I found better solution that preserves the square waves better, has lower EGD and paints deeper sound stage among other things.
Have you tried square waves with the new setup? Very informative kind of wave shape... 🙂
edit. By the way, at least partial documentation is nice; measurement distance, used time windows for different plots etc...
You have tuned the midwoofer to slight "forward lean" compared to fullranger (= negative EGD, can also be seen in step response). I also tried this kind of arrangement by delaying the comp driver from "optimal" and reversing it's phase. This can work, but I'm not sure if it's phase accurate setting by-definiton. I found better solution that preserves the square waves better, has lower EGD and paints deeper sound stage among other things.
Have you tried square waves with the new setup? Very informative kind of wave shape... 🙂
edit. By the way, at least partial documentation is nice; measurement distance, used time windows for different plots etc...
Last edited:
Legis,
These were close range measurements at 12.5cm from plane of horn mouth exit right in centerline. No time windowing applied. What do you mean by forward lean of mid woofers? I should say that I have not done time alignment via test tone at XO freq and shift delay in steps. Regarding reverse phase of HF, I think Danley does that on his horns and it is often done with multiways in general. I will post the regular phase soon. Like I said, just a single wrap at 500Hz. One of the best results I have achieved so far. I am going to hold off on square waves until the indicators say that it will look good. As Bwaslo said, I need phase at 0 or 180 not 90 or 270, needs to be linear and flat, freq response needs to be flat, etc. So I am still not so good on many of these that I don't think a square wave test is going to tell me much. Thanks for the feedback though.
These were close range measurements at 12.5cm from plane of horn mouth exit right in centerline. No time windowing applied. What do you mean by forward lean of mid woofers? I should say that I have not done time alignment via test tone at XO freq and shift delay in steps. Regarding reverse phase of HF, I think Danley does that on his horns and it is often done with multiways in general. I will post the regular phase soon. Like I said, just a single wrap at 500Hz. One of the best results I have achieved so far. I am going to hold off on square waves until the indicators say that it will look good. As Bwaslo said, I need phase at 0 or 180 not 90 or 270, needs to be linear and flat, freq response needs to be flat, etc. So I am still not so good on many of these that I don't think a square wave test is going to tell me much. Thanks for the feedback though.
Last edited:
I mean the step of the mids starts earlier than the step of the fullranger, hence the mids' EGD plot goes to negative. The band of the fullrangers come after the mids. Also seen in the step response (the downward slope between -600u ond 0ms).
Phase wraps do not mean anything per se, they just show when the absolute phase goes over +/- 180deg or whatever the transition degree is. unwrapped phase is more informative imo, the total phase shift is easier to see. With higher order filters the phase shifts/distorts more.
Seeing the phase correctly might require one to use shorter time window than the REW-default 500ms, even if measured very close.
Last edited:
I'm not sure IF one can achieve "flat" phase with IIR filtered multiway speaker (with FIR filters it can be different story).
Here's picture of Danley SH50 (from the spec sheet pdf), which shows ~540deg phase shift between 50hz and 10khz. That hints towards 2nd order filters in 3 way design (= total 6 poles, 90deg phase shift per pole).
An externally hosted image should be here but it was not working when we last tested it.
The phase plot that bwaslo showed (in post #215) has intrigued me, I suspect it can be "minimum phase" (like it's named in REW) kind of phase plot but I could be wrong? I would like to see the "regular phase" plot measured with REW by bwaslo.
Last edited:
Good hard work xrk971 continue and you suddenly find the magic tricks and Trynergy's potentiale, thanks sharing.
Legis think a bit inspiration to what showed at bwaslo post #215, is build into his free XO program XSim, maybe worth a try.
Picture-1 is the program and build in example file which make square waves ~650-2000Hz, change anything at left in schematic then in real time change plots on right.
Picture-2 is double-click tweeter S1 from schmatic, then by a dot in "derived" you get picture-3 which is a frd-file real world measurement of driver assigned to S1, also a measured impedance file can be assigned the driver.
Picture-1 is the program and build in example file which make square waves ~650-2000Hz, change anything at left in schematic then in real time change plots on right.
Picture-2 is double-click tweeter S1 from schmatic, then by a dot in "derived" you get picture-3 which is a frd-file real world measurement of driver assigned to S1, also a measured impedance file can be assigned the driver.
