Phase data at 1m w/ EQ
Here is the response as measured at 1m away (on axis) and acoustic signal phase. Surprisingly, the 1.7kHz boost did not throw the phase out of whack - appears to still vary linearly at the region I was most worried about. I would like it not to have any sudden phase wraps from 500Hz to 5kHz where all the point-source imaging signal cues are held. This horn actually did pretty well except for the one wrap at 600Hz. I don't know what is happening there but don't recall applying any EQ there.
Here is the response as measured at 1m away (on axis) and acoustic signal phase. Surprisingly, the 1.7kHz boost did not throw the phase out of whack - appears to still vary linearly at the region I was most worried about. I would like it not to have any sudden phase wraps from 500Hz to 5kHz where all the point-source imaging signal cues are held. This horn actually did pretty well except for the one wrap at 600Hz. I don't know what is happening there but don't recall applying any EQ there.
Attachments
There is a check box to unwrap the phase in the settings panel in the upper right area near the deg corner of the graph. IMO looks great between 700-6K. Great overall really
You mean like this? But then you lose info of where the wrap happens?
Or one can show it as minimum phase and excess phase, where the region bounded by 700Hz to 6kHz is clearly seen:
Attachments
Wanted to express your very good work with the thumbs up icon, the words "and a repeat" was meant for the thumbs up icon to be expressed twice, sorry the confusion.
Thanks for the phase plots post #161/163, trying to get wiser on bandwidth/frequency response/phase turn in minimum phase system. Bandwith/frequency response typical is our first goal and very audioable though as less as possible phaseturn seems good to reproduce waves. Horn seems to add some phaseturn though i only can find phase plot without horn in sealed/Nautaloss by looking here at you Vifa driver http://www.diyaudio.com/forums/full-range/247598-nautaloss-ref-monitor-11.html#post3766222 where turn seems 180º DC-HF.
PS guess the XO of 250Hz -24dB/oct is turned off in the plots.
Byrtt,
The 250Hz XO was indeed turned on in the above plots. That is probably cause of phase wrap near 300Hz.
Sorry I misunderstood your "repeat" - thanks for the kind words. I guess we say "two thumbs up" to mean same thing.
The more I look at the data and my listening impressions of this horn I am more certain that it is a worthwhile project for someone looking to try a fullrange horn as a project and do not have a wood shop to build one. The foam core with CLD is a very viable alternative for a home audio horn. Its light weight makes it very easy to place in room or move around. I had a lot of fun listening to it and now that it is put away, I am starting to miss its sound. The impact of the horn sound is unmistakable. You can be much farther away and it feels like you are in front of it. Vocals also have an enhanced intelligibility. I might have to build the stereo pair now
The 250Hz XO was indeed turned on in the above plots. That is probably cause of phase wrap near 300Hz.
Sorry I misunderstood your "repeat" - thanks for the kind words. I guess we say "two thumbs up" to mean same thing.
The more I look at the data and my listening impressions of this horn I am more certain that it is a worthwhile project for someone looking to try a fullrange horn as a project and do not have a wood shop to build one. The foam core with CLD is a very viable alternative for a home audio horn. Its light weight makes it very easy to place in room or move around. I had a lot of fun listening to it and now that it is put away, I am starting to miss its sound. The impact of the horn sound is unmistakable. You can be much farther away and it feels like you are in front of it. Vocals also have an enhanced intelligibility. I might have to build the stereo pair now
Last edited:
The data is only lost/altered permanently if you extract and then overwrite the original record with the minimum phase data.
The phase is constantly falling so when it is unwrapped it no longer fits the scale of the graph, you can change and offset the scale to get it back in the display window.
I was thinking about the 3D print idea some today. What if you were only to print the corner fillets and throat transition and at the mouth have the fillets transform into corners with pipe couplings for PVC pipe, leaving you to insert the panels... ?
xrk971,
Your results are very good indeed! And there is much less HF ringing after EQ.
I have thought of making a small FAST with a 3-inch fullrange driver, but perhaps, as you've shown, it's worth it to horn-load the fullranger to gain some sensitivity. And as I planned heavy EQ anyway, horn resonances are a non-issue.
Your results are very good indeed! And there is much less HF ringing after EQ.
I have thought of making a small FAST with a 3-inch fullrange driver, but perhaps, as you've shown, it's worth it to horn-load the fullranger to gain some sensitivity. And as I planned heavy EQ anyway, horn resonances are a non-issue.
Last edited:
If you want a compact FAST look here, I used a 3.5in TC9FD and works very well. BMS got good results with 4FE35 too. The tractrix is not what anyone would call compact.
http://www.diyaudio.com/forums/full-range/249984-cheap-fast-ob-literally.html
Also, if you had such high sensitivity on the fullrange, you will need a very sensitive woofer or sub woofer like a TH to match. But if you have space and matching bass support, the horn is indeed a good thing to have on top. Note that it works well here because the 5MR450NDY is a special driver that is really a mid range high fs driver with a very stiff cone, and powerful motor. I am not sure how well a regular 3in fullrange will do, although a simulation may price useful as a start but will not be able to include effects of cone flex under high pressure loading that the driver will see as a compression driver.
Last edited:
xrk971,
Thanks, this slot-loaded open baffle looks really nice too. Perhaps this is a good first project for me, as cutting an enclosure out of cardboard makes this speaker much more doable.
Perhaps a tractrix horn will be useful if I ever decide to make a speaker with larger SPL capabilities, reusing the old drivers. With something like 80 db/W Aura NS3 this could be really useful. How much of increase in sensitivity were you able to get in the end? I've found an earlier post where you said you were able to reach 108 db at 1 Wt/1m, which is a solid increase of 13 db if the driver's sensitivity is 95 db/Wt/m. But is it before or after EQ?
Thanks, this slot-loaded open baffle looks really nice too. Perhaps this is a good first project for me, as cutting an enclosure out of cardboard makes this speaker much more doable.
Perhaps a tractrix horn will be useful if I ever decide to make a speaker with larger SPL capabilities, reusing the old drivers. With something like 80 db/W Aura NS3 this could be really useful. How much of increase in sensitivity were you able to get in the end? I've found an earlier post where you said you were able to reach 108 db at 1 Wt/1m, which is a solid increase of 13 db if the driver's sensitivity is 95 db/Wt/m. But is it before or after EQ?
That was with EQ, although not much was applied at 1khz. You can expect anywhere up to about 10dB with a front horn typically. The OB a lot loaded speaker is a much easier first project but needs DSP and biamping which sounds like you have. I would not build a tractrix around an old low sensitivity driver. They need efficient powerful motors to work well.
xrk971,
Wow, that's a very big increase in SPL. Now I know how to make outdoor parties on the cheap
I am not opposed to bi-amping, and I think wiring together a PLLXO with some second-hand AV receiver won't be difficult in comparison to making an enclosure. Making it out of cardboard is much easier in my condo though, as opposed to cutting up plywood. I only got one question regarding your slot-loaded OB: what is the reason for the relatively big center-to-center spacing between the sections? I noticed that the fullrange driver could have been moved closer to the shelf above the LF-drivers. Did you choose to space drivers further apart because this shelf introduces resonances when close to fullranger?
Wow, that's a very big increase in SPL. Now I know how to make outdoor parties on the cheap
I am not opposed to bi-amping, and I think wiring together a PLLXO with some second-hand AV receiver won't be difficult in comparison to making an enclosure. Making it out of cardboard is much easier in my condo though, as opposed to cutting up plywood. I only got one question regarding your slot-loaded OB: what is the reason for the relatively big center-to-center spacing between the sections? I noticed that the fullrange driver could have been moved closer to the shelf above the LF-drivers. Did you choose to space drivers further apart because this shelf introduces resonances when close to fullranger?
Lol.
no, purely driven by practical needs like getting fullrange up higher to be closer to ear level. Tilting back helps. The woofer sections are low to keep lower center of mass for stability. The center to center distance that I ended up with, purely by chance, happens to match the 1/2-wave distance corresponding to the 340Hz XO (20 inches) for excellent integration of the woofer and fullrange as a single non localized source. It was not all luck if you read the thread you will see I modeled most of it in akabak first and even had a CAD model before making it out of cardboard. If I ever build a wooden speaker this will probably be high on my list. You may want to check out the design by Ztransform who came up with almost the same box independently - his is very well executed in CNC mdf. Look here for more info and inspiration. You will see that he too uses the TC9FD on top - an excellent driver for OB. http://www.diyaudio.com/forums/multi-way/252593-novel-open-baffle-construction-techniques.html
There is a check box to unwrap the phase in the settings panel in the upper right area near the deg corner of the graph. IMO looks great between 700-6K. Great overall really
Thanks for the tip and positive feedback on the tractrix experiment.
0.70x scale tractrix with 3 inch fullrange driver
I am exploring the possibility of a smaller less expensive version of the tractrix horn but still with a similar killer performance. Not sure if possible until I build a prototype but if the sims hold up for this none like they did for the 5MR450NDY, it could be really interesting. The 0.7x scale provides a very manageable size of 20 in wide x 12 in tall x 15 in deep. This means that a single sheet of foam core can be used for the top and bottom panels, thus cutting foam core requirements in half. The horn should be able to provide about 400 Hz on the low end and upwards of 18kHz on the high.
Here is the predicted SPL at max thermal load of 40watts or 12v, cone displacement here is 0.9mm so xmax is still far off so HD should be very low:
Sensitivity at 2.83v is about 108dB at 1m.
The best part is this driver only costs about $20. Anyone interested?
I am exploring the possibility of a smaller less expensive version of the tractrix horn but still with a similar killer performance. Not sure if possible until I build a prototype but if the sims hold up for this none like they did for the 5MR450NDY, it could be really interesting. The 0.7x scale provides a very manageable size of 20 in wide x 12 in tall x 15 in deep. This means that a single sheet of foam core can be used for the top and bottom panels, thus cutting foam core requirements in half. The horn should be able to provide about 400 Hz on the low end and upwards of 18kHz on the high.
Here is the predicted SPL at max thermal load of 40watts or 12v, cone displacement here is 0.9mm so xmax is still far off so HD should be very low:
Sensitivity at 2.83v is about 108dB at 1m.
The best part is this driver only costs about $20. Anyone interested?
Attachments
Last edited:
Here is another candidate that only costs $14 ea. This is the Vifa TC7FD04 2.5 inch line array driver. It actually has 91dB sensitivity at 2.83v as it is 4ohms. It can't take too much thermal power, but you won't need much power as it is super sensitive so a couple of watts should be more than enough.
The previous example was for a Faital Pro 3FE25, which can take more power.
The previous example was for a Faital Pro 3FE25, which can take more power.
Attachments
Think your knowledge and feeling for best result should rule what driver and impedance used, info my personal preference is higher voltage swing 8 ohm and up, think 4 ohm get too dampened but very good for cars and their low voltage high current.
Know it's early in process but for 2 way home system combine this 10" available in Europe at cost 63€, your teaser 20" x 12" horn, XO 800-1000Hz area, and DSP'ed system.
Know it's early in process but for 2 way home system combine this 10" available in Europe at cost 63€, your teaser 20" x 12" horn, XO 800-1000Hz area, and DSP'ed system.
Attachments
There is an 8ohm version of the 3FE25, it probably works fine too. Slightly higher Qts - but I am seeing that in the sims, a higher Qts is not necessarily bad for a horn. Probably want lower Qms though to damp out oscillations. That Monacor 8in looks like a nice driver. Are you thinking of putting it to an injection port a la Synergy or separate sealed box?
Was thinking off sealed because of simplicity, but now you ask remind me i haven't thought of the wave distance between the two drivers center related to XO frequency, need to check.
Another question have rumbled you know always popped up when laying in bed but forgotten in daytime. A feature in your Nautaloss the good black hole that handle the backwave, would Tactrix horn benefit Nautaloss and black hole instead of ordinary sealed or is LF in horn too high a frequency to benefit this better backwave dampening.
Another question have rumbled you know always popped up when laying in bed but forgotten in daytime. A feature in your Nautaloss the good black hole that handle the backwave, would Tactrix horn benefit Nautaloss and black hole instead of ordinary sealed or is LF in horn too high a frequency to benefit this better backwave dampening.