Haha, thanks! I've just realised that the jig will act as a nice template for marking and cutting the port holes too.
There is a little bit of tension in the plastic, and I haven't tried, but maybe once it's set in shape a heat gun could be used to release that tension. I'm not sure it would matter though, as the construction glue isn't going anywhere
I might have missed it so I apologise if I have, but what are the dimensions/volume of the injection ports?
To dampen the sides I could layer on some Gorilla glue, or make a brace that runs half-way up the side wall, from throat to mouth, connecting it with vertical braces between the top and bottom panels.
The 3fe25s arrive today...the woofers are on back-order
There is a little bit of tension in the plastic, and I haven't tried, but maybe once it's set in shape a heat gun could be used to release that tension. I'm not sure it would matter though, as the construction glue isn't going anywhere
I might have missed it so I apologise if I have, but what are the dimensions/volume of the injection ports?
To dampen the sides I could layer on some Gorilla glue, or make a brace that runs half-way up the side wall, from throat to mouth, connecting it with vertical braces between the top and bottom panels.
The 3fe25s arrive today...the woofers are on back-order
Last edited:
You can use a 2.0 in dia round hole or an equivalent area crescent shape like I did. It is not super critical, within 10% area variation is OK, as the sims don't show a huge dependence on this. Also location is not too critical if within 70% to 80% of total axial distance from throat.
I would leave it alone as the tension helps it hold its curved shape and increases stiffness - which is good. You will need to throw some caulking on it and add another layer of plastic, foam core, cardboard, whatever to add mass and damping. Then add bracing to the CLD layers.
Last edited:
I just got a notice that the buyout 6.5 inch polycone woofers are back in stock. Get them while you can... at $4.88 ea it is a terrific value. You need 8 for a pair of Trynergies.
An externally hosted image should be here but it was not working when we last tested it.
http://www.parts-express.com/6-1-2-poly-cone-midbass-woofer-4-ohm--299-609
Last edited:
The woofers have just been shipped, so PE were pretty quick to process them!
What is the overall thickness of the panels you're mounting the woofers to? I see you have a number of layers, so I want to mimic that to get the port volume as close to yours as possible
As for the damping, I'll have a look to see what makes the most sense. Thanks for the suggestions!
What is the overall thickness of the panels you're mounting the woofers to? I see you have a number of layers, so I want to mimic that to get the port volume as close to yours as possible
As for the damping, I'll have a look to see what makes the most sense. Thanks for the suggestions!
Ok. Well, given that I have the 3/8 plywood lying around, I can glue an extra layer on for the speaker zone, acting as a brace and also a constrained layer for damping (and given that it has more mass than the Foam Core Board, may be enough by itself). That'll give me a 3/4" thickness to work with, which seems appropriate. Had I thought about it before, I would have used 1/2" board as a compromise...though I don't think I have anything of that thickness large enough at the moment....
Last edited:
Fwiw, the "minimum phase impulse respose" or step or square wave response isn't really meaningful, it is based only on the dB curve, not the actual phase curve.
If you want to see actual impulse, step, or square wave responses you can download Xsim freeware to calculate all that from the dB and degree response curves. Just draw a schematic with just a driver connected to the amp and assign your curves to that driver. Then "Add Graphs" will get you the other three graphs (prepare for disappointment on the square wave, though, that takes more than a flat 'degrees curve' to make something resembling a square wave).
If you want to see actual impulse, step, or square wave responses you can download Xsim freeware to calculate all that from the dB and degree response curves. Just draw a schematic with just a driver connected to the amp and assign your curves to that driver. Then "Add Graphs" will get you the other three graphs (prepare for disappointment on the square wave, though, that takes more than a flat 'degrees curve' to make something resembling a square wave).
Bwaslo,
Thanks for the tip. Your build was the inspiration for me to start making a Synergy one way or another. I also like your other multiway designs with the SEOS. I suspect the square wave response is not going to look very good. This is probably where a dedicated 1in CD and qnty 4 sealed back mid drivers wil excel. If I were to try to measure the square wave response directly with a square wave generator and recorder - could that work and what frequency square wave would one use? Something in the bandwidth of the woofer I suppose? It seems I will have to dial in time alignment precisely before attempting this. Fortunately, music is not composed of square waves.
Xrk971-
Sure, if you have a mic, generator, and a scope, you can look at square wave response that way. Frequency range of the gen should be from about 50Hz (or your min freq) up to maybe 4kHz - because the mic won't likely go above 20k and you need to see at least the 5th harmonic to get a squarish waveform. A waveform accurate speaker will of course have to do it over a range of frequencies, not just at a "lucky" one! It's actually pretty hard to do, so don't getbhung up on it, particularly since there isn't any evidence it is even audible as far as I know.
Sure, if you have a mic, generator, and a scope, you can look at square wave response that way. Frequency range of the gen should be from about 50Hz (or your min freq) up to maybe 4kHz - because the mic won't likely go above 20k and you need to see at least the 5th harmonic to get a squarish waveform. A waveform accurate speaker will of course have to do it over a range of frequencies, not just at a "lucky" one! It's actually pretty hard to do, so don't getbhung up on it, particularly since there isn't any evidence it is even audible as far as I know.
Adjusting and optimizing time alignment
I finally had some time to get the time alignment dialed in between the HF full range and the mid woofers. I have my XO set up as 48dB/oct BW at 350Hz so I set a tone generator to 350Hz and flipped the phase of the full range driver. I took a ruler and measured the distance from the throat to the injection ports. Used that distance to set the initial guess of where delay is. It is approximately 0.9 ms so I set the relative delay between the HF and mid at that value. I set the mid absolute delay to be the max value allowed by miniDSP 2x4 at 7.5 ms since the sub is located far away in the corner of the room. The HF delay was then offset with less delay at 6.6ms to start. Then I turned on the RTA utility in REW and observed the 350Hz peak and slowly adjusted the delay on the HF in 0.01ms steps while observing if the RTA peak goes up or down. Adjust it in one direction or the other until you see the peak starting to fall. It will pass through a minimum and then begin to rise again. At that point the time alignment is perfect and you want to flip the phase back to normal and the amplitude should increase dramatically. To find the minimum when close it was easier for me to use my ear to listen for the minimum point. The 0.01ms steps are tiny but they make a difference - there is a noticeable change in 10 microseconds of time alignment. After completing this, write down that delay as you will use it over and over whenever reconfiguring the Trynergy setups in DSP. The initial guess was close but the small adjustments made a huge difference.
In listening to music played through it, the time coherency of dynamic tracks with percussion, bongo drums, tambourines, etc all have a tremendous amount of impact and realism. It was like someone flipped a switch and things popped to life. The Trynergy sounded great before and now sounds even better. Take the extra time to do this critical adjustment you will be very happy you did.
I finally had some time to get the time alignment dialed in between the HF full range and the mid woofers. I have my XO set up as 48dB/oct BW at 350Hz so I set a tone generator to 350Hz and flipped the phase of the full range driver. I took a ruler and measured the distance from the throat to the injection ports. Used that distance to set the initial guess of where delay is. It is approximately 0.9 ms so I set the relative delay between the HF and mid at that value. I set the mid absolute delay to be the max value allowed by miniDSP 2x4 at 7.5 ms since the sub is located far away in the corner of the room. The HF delay was then offset with less delay at 6.6ms to start. Then I turned on the RTA utility in REW and observed the 350Hz peak and slowly adjusted the delay on the HF in 0.01ms steps while observing if the RTA peak goes up or down. Adjust it in one direction or the other until you see the peak starting to fall. It will pass through a minimum and then begin to rise again. At that point the time alignment is perfect and you want to flip the phase back to normal and the amplitude should increase dramatically. To find the minimum when close it was easier for me to use my ear to listen for the minimum point. The 0.01ms steps are tiny but they make a difference - there is a noticeable change in 10 microseconds of time alignment. After completing this, write down that delay as you will use it over and over whenever reconfiguring the Trynergy setups in DSP. The initial guess was close but the small adjustments made a huge difference.
In listening to music played through it, the time coherency of dynamic tracks with percussion, bongo drums, tambourines, etc all have a tremendous amount of impact and realism. It was like someone flipped a switch and things popped to life. The Trynergy sounded great before and now sounds even better. Take the extra time to do this critical adjustment you will be very happy you did.
Great thanks this guide, to keep that precision as you just calibrated is this something to consider and add.
That if there is a noticeable change in 10uS step, then on the note where one wrote down the delay setting, good idea to add description cable setup length/type amp and device setup, this because changing something in these routes amps/cables could change +/- delay in that route.
The route chains is now a part of speaker, so changing something in the route then if one want max performance/precision could lead to new calibration.
Length of interconnect cables will not change delay at 10uS level as delays in wires are about 1nsec per foot. However, changing things like input coupling caps on the amps or the slope of the XO makes a big change and the delay will have to be re optimized. Also changing the amp will affect the phase/delay as well.
I apologise, I've not been able to work on them since Wednesday evening for one reason or another. The woofers are set to arrive on Tuesday, so then I can work out their placement, and the shape of the port holes
When you mentioned that the location of the ports should be between 70 and 80% of the total horn length (from the throat), would 75% be applicable, or is it best to aim closer to either of those specific numbers?
Thinking about the overall shape of the speakers, I'm also going to try and come up with chambers for the woofers that would fit a box shape that encloses the entire system, probably with access holes in them to reach the woofers when necessary. I suppose what I could do is make the box, and then strategically add braces that effectively seal a chamber of the right size. I could do the same for the mid-high driver too.
When you mentioned that the location of the ports should be between 70 and 80% of the total horn length (from the throat), would 75% be applicable, or is it best to aim closer to either of those specific numbers?
Thinking about the overall shape of the speakers, I'm also going to try and come up with chambers for the woofers that would fit a box shape that encloses the entire system, probably with access holes in them to reach the woofers when necessary. I suppose what I could do is make the box, and then strategically add braces that effectively seal a chamber of the right size. I could do the same for the mid-high driver too.
75% is good - if you had to pick one end vs the other pick 80% as the closer in you go the more it affects the tractrix HF horn profile where it counts. If putting woofers in an enclosure you can now use that as a bass reflex and tune it with volume and ports to reach 45Hz easily. The dynamics of the bass may not be as good as sealed though.