These look very interesting
I really like the look of these. I've been looking for something like this to mount drivers on my Jeeps' A pillar. Planning to use a DSP also. Do you know who does stuff like this professionally? With something like this I might even be able to use 4" mids, as long as they had small neodymium magnets. Maybe a one piece to include tweeter and mid in one pod? How much to get something like this produced?
Inventor isn't that hard to learn. Just need to learn the basics and play with it. I find it very useful for all my hobbies. I also used it for my car audio setup. Here's a waveguide design for a Vifa tweeter:
I made that pod to use the Vifa in my car and minimize diffraction. Worked very well. Actually, the great sound I got in my car is the reason I'm building new home speakers!
I printed the above model trough an online printing service.
DSP in my Pioneer radio takes care of the crossover and time alignment. The speakers disappear completely leaving you with a great stage that extends beyond the actual car boundaries. It sounds bigger than it is. I hope to achieve that in my living room as well.
But I had Inventor running at home due to my job. It is way to expensive for a home user. If I ever start a business of my own I won't hesitate to buy a license.
I really like the look of these. I've been looking for something like this to mount drivers on my Jeeps' A pillar. Planning to use a DSP also. Do you know who does stuff like this professionally? With something like this I might even be able to use 4" mids, as long as they had small neodymium magnets. Maybe a one piece to include tweeter and mid in one pod? How much to get something like this produced?
Its really easy to replace drivers, since they are front mounted. I have seven extras of these extinct 6 ohm drivers
Thanks. The roundover doesn't give any waveguide effect at all
I designed it for wave guide purpose, but measurements showed its completely ineffective. However they looks awesome & adds strength to the baffle
I am using a 500W computer SMPS for power supply of the car amp.
rockk19,
Congrats - now I want one too!
wesayso,
Looks like more folks are taking the step towards their own two towers!
You mentioned "underpowered amp" - what amp are/were you using? The sensitivity of your two towers must be pretty high?
Sensitivity is high enough at about 96 dB/watt but the boost in low frequencies takes away from that.
My amplifier is a Pioneer A757 Mark II. Build just before they forgot how to build decent amplifiers. Bought it new many moons ago and had it serviced last winter. Its about 100 watt at 8 ohm. It weighs 19 Kg.
But the arrays can take about 240 watt each. So in time i'll look for something different. Maybe Hypex NCore? I'm hearing good things about those.
Ask me again in 3 weeks... I'm sure we can work something out
.
I found a nice option for low cost $20 200w using IRS2092 amp with $20 53v dual rail 450w SMPS. There were problems reported by the SMPS users but we figured out that it was because all outputs (there are auxiliary output taps at 12v and 7v) need minimum load of about 100mA to be stable.
It sounds very nice and clean. Low distortion.
http://www.diyaudio.com/forums/class-d/275915-200w-irs2092-amp-20-a.html
It sounds very nice and clean. Low distortion.
http://www.diyaudio.com/forums/class-d/275915-200w-irs2092-amp-20-a.html
Inspired by line array concept, I've just completed my own version.
25x per channel of 3.5 inch black kevlar coned Tymphany TPY03.
This is my initial test setup & measurements. I am using a car amplifier
All the measurements are taken at 3 metres distance form the speakers, 15 deg off-axis.
Included in the attachments are an unequalized frequency response of the system, a full range frequency response/ distortion graph of the system equalized with miniDSP, a 90Hz & above frequency response/ distortion graph ( showing 90Hz at 107dB @ 3m with 0.31% 3rd order HD & 1.48% THD
Great job! Congratulations.
Yep, NCores should be good choice - have read that they are very neutral sounding and powerful. I have Hypex UcD, and find them very good Class-D amps.
Btw, the NCores will end up being much smaller and lighter than your current amp.
Last edited:
System sensitivity
I measured the sensitivity of my arrays & its a little less than what I was expecting.
The impedance of my arrays are 6 Ohms.
So, I took readings at 2.45V rms both for Pink Noise and 1KHz.
The results were same.
For miniDSP equalized system, I got 81 dB at 3 metres distance from the speakers.
For the unequalized system, I got 84.5 dB at 3m.
Wesayso,
Have you done pink noise sensitivity tests for your equalized arrays?
I wonder how you got the 96dB figure.
I measured the sensitivity of my arrays & its a little less than what I was expecting.
The impedance of my arrays are 6 Ohms.
So, I took readings at 2.45V rms both for Pink Noise and 1KHz.
The results were same.
For miniDSP equalized system, I got 81 dB at 3 metres distance from the speakers.
For the unequalized system, I got 84.5 dB at 3m.
Wesayso,
Have you done pink noise sensitivity tests for your equalized arrays?
I wonder how you got the 96dB figure.
I just looked up the reference and it was a comparison at listening position of the array compared to a conventional point source system. Kinda cheating in a way but done to make it comparable. If you'd measure the array at 1 M it wouldn't really be down as much as a point source.
The 3 meter distance would mean ~9.5 dB down compared to 1 m (with a point source).
According to this calculator here: Efficiency and sensitivity conversion - loudspeaker percent and dB per watt and meter loudspeaker efficiency versus sensitivity vs speaker sensitivity 1 watt = 2,83 volt box chart - sengpielaudio Sengpiel Berlin
(scroll down to Sound pressure level and amplifier power)
Your array would measure 94 dB / 1 watt at 1 M to arrive at 84.5 dB at the listening position (if it were a point source). Not that far off I'd say. In MiniDSP you are the judge of what frequency you boost or cut. So the EQ-ed value depends entirely on your own judgement of applying the EQ, right?
The 3 meter distance would mean ~9.5 dB down compared to 1 m (with a point source).
According to this calculator here: Efficiency and sensitivity conversion - loudspeaker percent and dB per watt and meter loudspeaker efficiency versus sensitivity vs speaker sensitivity 1 watt = 2,83 volt box chart - sengpielaudio Sengpiel Berlin
(scroll down to Sound pressure level and amplifier power)
Your array would measure 94 dB / 1 watt at 1 M to arrive at 84.5 dB at the listening position (if it were a point source). Not that far off I'd say. In MiniDSP you are the judge of what frequency you boost or cut. So the EQ-ed value depends entirely on your own judgement of applying the EQ, right?
It's been a while since the last post here...
I managed to get one conjugation network complete before going on holiday. Did some quick tests on it but nothing definitive yet. I hope to get the second one finished soon. I won't comment on how it sounds yet, to early to tell.
Meanwhile I got a bit side tracked working on the mid/side processing which has been giving me good results in various ways. I've said more than once that I liked what JRiver's "Surround Field" processing did in my setup.
However since I upgraded to JRiver version 20 I kinda lost my way. I could swear things sounded different to me.
There had always been some element I could not quite grab that threw me off. Not so much in the sweet spot, but when moving out of the general listening area. In the sweet spot I've been wowed many times with chills running down my spine. But moving around in the room something was off, kinda bugging me. I never successfully managed to get similar results with using Voxengo's MSED. It was similar in a way but missing something on its own. For a long time I had been using JRiver's "Surround Field" in combination with a boost in MSED of the mid channel to get that equal loud as the sides again. Until I stumbled over Pano's http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html thread.
A recent post by member CBW there mentioned LoCo.
Basically it also used mid/side processing with an equaliser moving the frequencies at ~400 to the sides and ~4000 to the mid signal.
It references this paper: FrequenzabhHoerereignisrichtung
A good read on mid/side processing is linked in one of those papers as well: http://www.audiosignal.co.uk/Resources/Stereo_shuffling_A4.pdf
I had stumbled on a paper about the "S" curve before, don't remember where. It might even have been the same paper I linked above. I decided to try it in my setup, first in JRiver 19 and now in v20. So far I like what I hear.
Here's how I set it up:
The first MSED step is encoding the stereo signal to mid/side, the second one decodes the signal back to stereo.
In the equaliser I run 4 PEQ's
First at 4000 Hz, Q= 0.3 and boost is 1.2 dB on mid (left)
Second at 400 Hz, Q= 0.5 and boost is - 1.5 dB on mid (left)
Third at 4000 Hz, Q= 0.3 and boost is -1.6 dB on side (right)
Fourth at 400 Hz, Q= 0.5 and boost is 1.5 dB on side (right)
I threw in a High Shelf on the mid (right) channel at 13300, Q=0.71 and gain is 2 to get just a bit more focus in the phantom centre.
These are al preliminary numbers, subject to change of coarse. But so far it's very pleasing to listen to. Sounds great in the sweet spot. But I enjoy it also when walking away from the listening area. Notice the slightly less boost on the mid at 4000, I did that after some listening and also played around with the center frequency.
My bizarre dislike (pretty strong word here, it wasn't that bad) outside the listening area seems to be gone. When siting in the door opening of my adjoining kitchen to the garden it seems there is a band playing in the living room. Feels stronger/more real than before. Something I was looking for but had not accomplished yet. Before this change it did sound more hollow somehow. Less focussed on the centre parts, more room ambience from within the recording. Great in the listening sweet spot, less so outside of that area.
The old way of processing did sound "wider" at times but like I said. Off site listening was more compromised. Too soon to tell though if this setup is a keeper. Only time can tell. The first impression is good though, hence I wanted to share it on here. I made the change a couple of days ago.
I always wondered about the sound outside of the listening area. I remember Gmad talking about his setup where the limited 4 cycle processing gave him better sound everywhere. I guess the "Surround Field" option in JRiver was boosting the sides a bit too much. I could easily measure the difference. Next time with the mic out I'll show it. I've been trough many plugin demo's to find an alternative. This above solution might give the most pleasing results, as it's fully configurable.
I already spend way to much time listening to several tracks and enjoying myself way to much. I should be busy doing something else.
I managed to get one conjugation network complete before going on holiday. Did some quick tests on it but nothing definitive yet. I hope to get the second one finished soon. I won't comment on how it sounds yet, to early to tell.
Meanwhile I got a bit side tracked working on the mid/side processing which has been giving me good results in various ways. I've said more than once that I liked what JRiver's "Surround Field" processing did in my setup.
However since I upgraded to JRiver version 20 I kinda lost my way. I could swear things sounded different to me.
There had always been some element I could not quite grab that threw me off. Not so much in the sweet spot, but when moving out of the general listening area. In the sweet spot I've been wowed many times with chills running down my spine. But moving around in the room something was off, kinda bugging me. I never successfully managed to get similar results with using Voxengo's MSED. It was similar in a way but missing something on its own. For a long time I had been using JRiver's "Surround Field" in combination with a boost in MSED of the mid channel to get that equal loud as the sides again. Until I stumbled over Pano's http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html thread.
A recent post by member CBW there mentioned LoCo.
Basically it also used mid/side processing with an equaliser moving the frequencies at ~400 to the sides and ~4000 to the mid signal.
It references this paper: FrequenzabhHoerereignisrichtung
A good read on mid/side processing is linked in one of those papers as well: http://www.audiosignal.co.uk/Resources/Stereo_shuffling_A4.pdf
I had stumbled on a paper about the "S" curve before, don't remember where. It might even have been the same paper I linked above. I decided to try it in my setup, first in JRiver 19 and now in v20. So far I like what I hear.
Here's how I set it up:
The first MSED step is encoding the stereo signal to mid/side, the second one decodes the signal back to stereo.
In the equaliser I run 4 PEQ's
First at 4000 Hz, Q= 0.3 and boost is 1.2 dB on mid (left)
Second at 400 Hz, Q= 0.5 and boost is - 1.5 dB on mid (left)
Third at 4000 Hz, Q= 0.3 and boost is -1.6 dB on side (right)
Fourth at 400 Hz, Q= 0.5 and boost is 1.5 dB on side (right)
I threw in a High Shelf on the mid (right) channel at 13300, Q=0.71 and gain is 2 to get just a bit more focus in the phantom centre.
These are al preliminary numbers, subject to change of coarse. But so far it's very pleasing to listen to. Sounds great in the sweet spot. But I enjoy it also when walking away from the listening area. Notice the slightly less boost on the mid at 4000, I did that after some listening and also played around with the center frequency.
My bizarre dislike (pretty strong word here, it wasn't that bad) outside the listening area seems to be gone. When siting in the door opening of my adjoining kitchen to the garden it seems there is a band playing in the living room. Feels stronger/more real than before. Something I was looking for but had not accomplished yet. Before this change it did sound more hollow somehow. Less focussed on the centre parts, more room ambience from within the recording. Great in the listening sweet spot, less so outside of that area.
The old way of processing did sound "wider" at times but like I said. Off site listening was more compromised. Too soon to tell though if this setup is a keeper. Only time can tell. The first impression is good though, hence I wanted to share it on here. I made the change a couple of days ago.
I always wondered about the sound outside of the listening area. I remember Gmad talking about his setup where the limited 4 cycle processing gave him better sound everywhere. I guess the "Surround Field" option in JRiver was boosting the sides a bit too much. I could easily measure the difference. Next time with the mic out I'll show it. I've been trough many plugin demo's to find an alternative. This above solution might give the most pleasing results, as it's fully configurable.
I already spend way to much time listening to several tracks and enjoying myself way to much. I should be busy doing something else
.
Last edited:
The total array design looks like this:
Final version as build looks like this:
(not too far off I presume)
How do you stop the towers toppling over? Have you anchored them to something? When you turn them up in volume, do they sway? I am being serious if naive, I'm new to this sort of thing so please forgive me if my questions seem obtuse.
They are not fixed to anything, just standing on the feet you see.
When playing full swing they do not move at all. It's amazing how quiet the enclosure is. No movement and even with the ear against the sides you only hear the sound coming out of the front. The total enclosure weight is about 1000x the moving mass (cone etc.).
When playing full swing they do not move at all. It's amazing how quiet the enclosure is. No movement and even with the ear against the sides you only hear the sound coming out of the front. The total enclosure weight is about 1000x the moving mass (cone etc.).
Last edited:
Do they do everything you hoped they would?
Do they sound the best speakers you've ever heard?
Well done for being so innovative and putting in such effort, I don't think I'll ever get to this level of expertise.
Thanks Bob
That is THE question (lol). The thing they do best is being a very cohesive source. That makes a lot of different types of music sound very good! I'm not done experimenting yet, to see what else they can do or achieve. Still learning here.
The size of the speaker does help in giving very convincing sound for big orchestra's etc. very overwhelming
. In short, in more than one way they are better than I ever heard before. But in some things I've heard better. The upper part, high frequencies isn't always as crystal clear as I would want it to be. Being very honest here. And I'm still voicing, getting closer and closer to very pleasing voices. But it seems to take me some time. Every recording is different and that makes it hard to find that one right pleasant curve. Part of my trouble is a rule my girlfriend set: no more damping panels! In my mind I could still benefit from some damping or diffusion here and there (lol).
For movies they are simply awesome! No doubt about it.
If I remember correctly, Zaph showed that Vifa TC9 driver to be very flat at the high end, but the Vifa spec sheet showed it to have a bit of a peak in the FR at about 10kHZ. I have that same driver in my computer speakers that I recently built (bi-amp'd, 8 inch woofer X=500HZ), and I think I may be hearing what could be a peak at 10kHZ. The treble in my main system sounds a lot smoother and "lush" (Seas Millenium 1 inch domes). You might try putting a slight dip in the response at 10kHZ.
Another thing I might try is to electronically bandwidth limit the towers to 7kHZ, and use a single (per side, 2/3 the way up to the top) 1.5 inch Fountek Ribbon tweeter (Zaph says it's one of the best) to do the high end. Or possibly a Dayton (affordable) "Air-Motion" tweeter (which I think is di-pole). I think the major advantage of floor to ceiling line arrays may be in the frequencies below about 7kHZ, based on my understandings. I could be wrong, but it's something to try.
For room treatment, I've found that corners are the worse enemy. I nailed 2 inch diameter cotton rope (from a fabric store) into most 2 and 3 surface corners of my living room (not at the floor), and the clap hands test showed a major improvement in the decay response. Bad ringing was virtually eliminated. No large panels or bass traps of any kind. Much better WAF. I better add that cotton rope is very flammable (I didn't know), so soft non-flammable foam rubber would be a wiser choice.
One of the problems with building highly accurate speakers is that the shortcomings of the recording process can be exasperated (made worse or at least more noticeable). I have never regretted designing and building my 4 section Baxandall tone control circuit. Easy to use and makes bad recordings listenable. There are just too many reasons why flat is not likely optimal, and this varies from recording to recording. Especially when you're a musician who might be digging up very old recordings of dubious quality.
Again, congratulations for creating such a beautiful work of art.
Thanks Bob
That is THE question (lol). The thing they do best is being a very cohesive source. That makes a lot of different types of music sound very good! I'm not done experimenting yet, to see what else they can do or achieve. Still learning here.
The size of the speaker does help in giving very convincing sound for big orchestra's etc. very overwhelming
And I'm still voicing, getting closer and closer to very pleasing voices. But it seems to take me some time. Every recording is different and that makes it hard to find that one right pleasant curve. Part of my trouble is a rule my girlfriend set: no more damping panels! In my mind I could still benefit from some damping or diffusion here and there (lol).
For movies they are simply awesome! No doubt about it.
What about adding a tweeter in a tiny cab of it's own rather like B&W to get that clarity you seek?
Canvases make aesthetically pleasing damping panels and your boss[ha ha ha] could pick her own pictures?