No, I think we have a misunderstanding: What I meant before, and what I mean again, is that the offset ( = horizontal distance only) is 6,4cm = 2,52in with my speakers. This is front edge to front edge of the drivers.
Center to center (vertical distance only) is 18,5cm = 7,3in with my speakers.
Center to center (vertical distance only) is 18,5cm = 7,3in with my speakers.
Last edited:
Ok, thanks for clarifying - axial offset distance. The 2.52in is very close to what I think the simulation called for - about 3in I think. But best to do it manually. If you have measurement microphone, it will tell you the perfect distance. However, it is close enough as that distance is only valid for one listening point on a non-coaxial speaker. My simple solution was to put it on the bottom to increase the offset relative to earheight that is at level of woofer. Your way is the best way to do it. You might want to line the hard ledge between the tweeter and woofer box with felt to reduce the reflection there that might cause a cancellation dip in the full range response. That's about 4-5cm distance (driver centerline to hard wall ledge) or a 3400Hz to 4300Hz 1/2-wave cancellation dip.
Detailed offset calculation
Dear XRK,
I cannot follow you all the way, I think things are actually different what you perhaps think. Furthermore I doubt my own certainty, I doubt if I really got the right result with my listening experiment, perhaps I will redo it, let me explain.
If you put both drivers on a plane baffle, the voice coil of the Dayton (woofer!) is about 5cm = 2in behind the voice coil of the Scan Speak (horizontal distance). I have attached a more the less true to scale drawing: You say, that you put the speaker upside down, and that the woofer is on ear height. At a listening distance of 1,5m = 59in = 4,9ft, the woofer's (!) acoustical center will still be 1,6in behind.
In my case, with my speakers, the ear is at the height just in between the centers of the two drivers, and the horizontal offset of the voice coils is about 1,4cm (my 6,4cm minus the 5cm the woofers voice coil is even further back due to the woofer's size). 1,4cm = 0,55in. So the Scan Speak (acoustical center) is 0,55in behind. This means, to my ears that 0,55in delay of the Scan Speak is right. Compared to your setup you need 2,15in (0,55 + 1,6) more delay to the Scan Speak.
But to sum it up, after doing these calculations, I doubt my own results of the listening experiment. Because I think I should reach somewhere close to the 3in, isn't it? That would mean an edge to edge offset of 12,6cm = 5in. I have only tried up to 8cm. I think I will redo the listening experiment, perhaps it sounds even better when I go even more backwards with the Scan Speak. I don't think it to be, because it already did sound on the spot, really on the spot, but who knows... Or your 3in simulation delay is wrong. Can you see any other possibilties of errors?
Florian
Dear XRK,
I cannot follow you all the way, I think things are actually different what you perhaps think. Furthermore I doubt my own certainty, I doubt if I really got the right result with my listening experiment, perhaps I will redo it, let me explain.
Yes I remember you saying, that the delay should be 3in. - I think this calls for a distance offset between the acoustical centers of the drivers, which will produce that delay of sound, right? The acoustical center is somewhere close to the voice coil of a driver.
If you put both drivers on a plane baffle, the voice coil of the Dayton (woofer!) is about 5cm = 2in behind the voice coil of the Scan Speak (horizontal distance). I have attached a more the less true to scale drawing: You say, that you put the speaker upside down, and that the woofer is on ear height. At a listening distance of 1,5m = 59in = 4,9ft, the woofer's (!) acoustical center will still be 1,6in behind.
In my case, with my speakers, the ear is at the height just in between the centers of the two drivers, and the horizontal offset of the voice coils is about 1,4cm (my 6,4cm minus the 5cm the woofers voice coil is even further back due to the woofer's size). 1,4cm = 0,55in. So the Scan Speak (acoustical center) is 0,55in behind. This means, to my ears that 0,55in delay of the Scan Speak is right. Compared to your setup you need 2,15in (0,55 + 1,6) more delay to the Scan Speak.
But to sum it up, after doing these calculations, I doubt my own results of the listening experiment. Because I think I should reach somewhere close to the 3in, isn't it? That would mean an edge to edge offset of 12,6cm = 5in. I have only tried up to 8cm. I think I will redo the listening experiment, perhaps it sounds even better when I go even more backwards with the Scan Speak. I don't think it to be, because it already did sound on the spot, really on the spot, but who knows... Or your 3in simulation delay is wrong. Can you see any other possibilties of errors?
Florian
Attachments
The 3in required offset already includes the voice coil offset. That is, the FRD files for the simulation were measured with both drivers on a flat baffle with the mic on the tweeter axis (hence woofer voice coil will physically be located behind the tweeter by about 2in). With this, it was found in the XSim program or PCD that the tweeter needed to be moved backwards by about 3in (or net extra 1inch or so) in order to get the right alignment. So essentially moving them so that the voice coils are about the same axial distances away.
What do you mean with axial distance?
I think I have understood what you wrote. The tweeter's voice coil has to be about 1in behind the woofer's, talking about pure distance. Right? In my setup it turned out to be 0,55in.
Still: If you now regard my setup as reference then, if my setup seems to be right then, this still means, that you need 2,15in more delay to the Scan Speak in your setup, as I have calculated before.
I think I have understood what you wrote. The tweeter's voice coil has to be about 1in behind the woofer's, talking about pure distance. Right? In my setup it turned out to be 0,55in.
Still: If you now regard my setup as reference then, if my setup seems to be right then, this still means, that you need 2,15in more delay to the Scan Speak in your setup, as I have calculated before.
Well, it’s related to diffraction - it’s 1/2-wave reflection destructive interference cancellation that can cause dips. Diffraction is related and caused by smaller discontinuities and edges of the baffle. The hard edge here is like a wall - so I don’t consider it diffraction like baffle edge diffraction that can be reduced with a round-over.
I mean a thin 2-3mm of felt not thick foam there. But if you tried it and it doesn’t sound good to you, no worries. It’s what your ears like that counts.
In my case the delay is correct as seen from the step response and flat phase as measured. A right triangle with tweeter sharp peak at the leading edge and in same polarity as woofer shows that the delay is correct.
I mean a thin 2-3mm of felt not thick foam there. But if you tried it and it doesn’t sound good to you, no worries. It’s what your ears like that counts.
In my case the delay is correct as seen from the step response and flat phase as measured. A right triangle with tweeter sharp peak at the leading edge and in same polarity as woofer shows that the delay is correct.
Last edited:
Sorry, this was a fast shot, sure there is reflection.
Perhaps measurements could be even better? Your setup with your speaker upside down, both drivers on a plane baffle is far different than my setup, you don't have the delay I have got. I will do a measurement of my speaker the next days, we will see, what it looks like!
Okay, thank you, I will try a thinner felt-like material!
Florian
Perhaps measurements could be even better? Your setup with your speaker upside down, both drivers on a plane baffle is far different than my setup, you don't have the delay I have got. I will do a measurement of my speaker the next days, we will see, what it looks like!
Okay, thank you, I will try a thinner felt-like material!
Florian
I know that things are afoot in the Vunce household. Anxiously waiting for updates on the TL progress.
You and me both. I recently whipped up some boxes for the full range driver to sit on top, somewhat in the style of the B&W 801's.
I'm using a Fostex FF85wk on top of the rs-225s, in an open backed baffle.
Shared album - Gable Barber-Smith - Google Photos
Disregard the neo3's in that photo, I ended up not using them. The new version I've been working on here and there, is the same basic baffle, but cut flush with the top of the main box, and will have a ~8"x14" box on top, that has no back currently. I'm going to test stuffing it progressively with denim insulation and magic erasers, measure and listen until I find something I like (hopefully).
Shared album - Gable Barber-Smith - Google Photos
These were the first boxes I made, and ended up scrapping them, the new boxes are very similar, just with the dimensions mentioned above.
I use the same basic crossover you initially developed a few years ago, with a few very minor tweaks to suit my room and ears, and for the sake of experimentation of course
The bass is so wonderfully tight and controlled, but certainly doesn't dig deep. I don't need ~35hz in my room, there is a good bit of room gain and too much down low starts to bloom more than I like.
The TL is interesting, as a bit more lf response could be quite nice. The rs225 is a heck of a driver, when I initially measured them in the test box I built, I cranked them up with some very low frequency signals and they seemed to handle them quite well indeed. In their price range, they are an immense value.
I'm interested to see how the tl progresses.
No they don't. Thankfully the speakers are to be used in a fairly small room, so I don't need much power at all. I may bi-amp at some point, I have a few low power tube amplifiers here that I can play with as well.
Progress Report: First Sound
Yesterday, I lined the TL enclosures in key areas with 1.5” mattress topper “egg crate” like foam. In addition, the upper chamber behind the drivers and down the back port was then filled with fiberglass stuffing. The cone for the Scanspeak 10F driver was also semi-filled with fiberglass. I had decided to keep the XO components outside the enclosure for easy access to tweak, I am curious how different capacitors and inductors will impact the SQ. At that point, I was ready to close the side up and power up. But my dilemma was I didn’t want to glue the side panel on yet, so I broke out the ”go-to” 3m blue painters tape and taped it as tight as I could.
Hooked up the XO and a small IcePower amp, fired up and there was music! Folks, this speaker has excellent potential. Right out of the box the Scanspeak 10F sounded very clear and enjoyable. When I got up close to the TL slot I could hear defined bass but it wasn’t impactful. Then I squeezed the sides of the enclosure tight together. Wow! It was like a subwoofer was added to the room. I asked my wife to give a listen, she had her back facing the speaker so as not to see what I was doing, every time I squeezed the enclosure she would remark how the bass increased.
So, that experiment led me to tonight’s work. The taped on side is no good, the transmission line not working properly because it’s not airtight. I also removed the fiberglass insulation and replace it with a lighter polyfill. Then I applied the adhesive and attached the side panel. Then put approximately 70lbs. of weights to get the adhesive to seal the perimeter and all internal TL pieces.
Now, its got to sit for 24hrs...... no immediate gratification tonight
Forgot to add this: These foam speakers are a bit top heavy, I added 5lbs of lead shot in the sealed square chamber at the base to help counteract that.
Yesterday, I lined the TL enclosures in key areas with 1.5” mattress topper “egg crate” like foam. In addition, the upper chamber behind the drivers and down the back port was then filled with fiberglass stuffing. The cone for the Scanspeak 10F driver was also semi-filled with fiberglass. I had decided to keep the XO components outside the enclosure for easy access to tweak, I am curious how different capacitors and inductors will impact the SQ. At that point, I was ready to close the side up and power up. But my dilemma was I didn’t want to glue the side panel on yet, so I broke out the ”go-to” 3m blue painters tape and taped it as tight as I could.
Hooked up the XO and a small IcePower amp, fired up and there was music! Folks, this speaker has excellent potential. Right out of the box the Scanspeak 10F sounded very clear and enjoyable. When I got up close to the TL slot I could hear defined bass but it wasn’t impactful. Then I squeezed the sides of the enclosure tight together. Wow! It was like a subwoofer was added to the room. I asked my wife to give a listen, she had her back facing the speaker so as not to see what I was doing, every time I squeezed the enclosure she would remark how the bass increased.
So, that experiment led me to tonight’s work. The taped on side is no good, the transmission line not working properly because it’s not airtight. I also removed the fiberglass insulation and replace it with a lighter polyfill. Then I applied the adhesive and attached the side panel. Then put approximately 70lbs. of weights to get the adhesive to seal the perimeter and all internal TL pieces.
Now, its got to sit for 24hrs...... no immediate gratification tonight
Forgot to add this: These foam speakers are a bit top heavy, I added 5lbs of lead shot in the sealed square chamber at the base to help counteract that.
Attachments
While we wait for Vunce's enclosure to dry...
I wanted to bypass the miniDSP HD I had and try a passive XO with my SBAcoustics drivers.
They are dual racetrack 5x8" with the wonderful SB65 2.5", on OB.
I couldn't find anything that excited me, and in the end, I came back to this thread and decided to keep it simple, even though it didn't look amazing in the sims.
Still... it turned out ok.
It's not transient perfect, but it's not horrible either.
With some bass boost at 60Hz, they are good enough for casual listening.
For more, or movie watching, I cancel the boost and cross them at 100Hz to dual subs.
So, sometimes, simple is indeed enough.
Thanks X for this thread.
I wanted to bypass the miniDSP HD I had and try a passive XO with my SBAcoustics drivers.
They are dual racetrack 5x8" with the wonderful SB65 2.5", on OB.
I couldn't find anything that excited me, and in the end, I came back to this thread and decided to keep it simple, even though it didn't look amazing in the sims.
Still... it turned out ok.
It's not transient perfect, but it's not horrible either.
With some bass boost at 60Hz, they are good enough for casual listening.
For more, or movie watching, I cancel the boost and cross them at 100Hz to dual subs.
So, sometimes, simple is indeed enough.
Thanks X for this thread.
Attachments
Hi Vunce,
These speakers look great, and you are really taking your time and working carefully. They will reward you with great sound. Very nice work on the foam damping pads and bracing. Don’t forget to seal the hole for the wires - maybe plumbers’ putty so it doesn’t harden up?
I was thinking you could have put one of those 12lb dumbbells inside at the bottom for a stabilizer weight.
So you just happen to have 10lbs of lead shot on hand? Who has 10lbs of lead shot lying around? Amazing preparedness or you use them for melting your own 17th century flintlock musket rounds?
The wait is killing me.
Super work!
These speakers look great, and you are really taking your time and working carefully. They will reward you with great sound. Very nice work on the foam damping pads and bracing. Don’t forget to seal the hole for the wires - maybe plumbers’ putty so it doesn’t harden up?
I was thinking you could have put one of those 12lb dumbbells inside at the bottom for a stabilizer weight.
So you just happen to have 10lbs of lead shot on hand? Who has 10lbs of lead shot lying around? Amazing preparedness or you use them for melting your own 17th century flintlock musket rounds?
The wait is killing me.
Super work!