I'm estimating 90hz, but will probably cross over higher, like 120hz. I have yet to test them, and have some concerns around the excursion limits on the 6.5" woofers. I'll setup a few configurations and switch between them on the minidsp.
I originally had some other ideas around building these speakers. They are built using BWaslo's 60x90 synergy spreadsheet, a few inches larger in dimensions. I had the first prototype "unity horn" I was planning on having fit into the cabinet but decided to build a new horn to fit the enclosure precisely, and I decided what the heck, lets build a synergy and put some woofers in. This presented some difficulties around how to cross over the five sets of drivers. For now I plan on driving the 4x10" and 1x15" on separate amps both low passed from the minidsp driving the same frequencies. What would be ideal is to high pass the 10" woofers and low pass the 15" below the 4x10" but for now this will have to do.
Cutting mids square for close spacing
I decided to try GRS mids because the other buyout mids (Titan) were not performing great. In order to fit them on this horn they needed to be cut down so instead of trying to measure and mark where to cut I made a square jig out of a piece of plywood to help make the cuts. It makes this way easier. Use a good quality hacksaw blade and the cuts are fairly accurate. Cover the magnet and frame with masking tape to catch the filings.
I decided to try GRS mids because the other buyout mids (Titan) were not performing great. In order to fit them on this horn they needed to be cut down so instead of trying to measure and mark where to cut I made a square jig out of a piece of plywood to help make the cuts. It makes this way easier. Use a good quality hacksaw blade and the cuts are fairly accurate. Cover the magnet and frame with masking tape to catch the filings.
Attachments
I also replaced the buyout woofers with Dayton 6" pro woofers because some of the surrounds and dust caps came unglued. I noticed after mounting one the cone did not have enough clearance and would have smacked the horn so I routed out some spacers with the Jasper. Worked out perfectly to give just enough space, around 1/8 inch or 3/16 with the gasket.
Attachments
Test with frequency generator
I did a quick test of the drivers on the horn using an analog frequency generator on a breadboard electronics lab box, they seem to have smooth response within their passbands. Audibe was a null around 1950hz. I enlarged the mid ports to 3/4 inch (one port per mid). Hoping this works out without messing up the compression driver, but from a frequency sweep it sounded ok, no major problems. I'll be doing some analysis with REW once I get these put back together.
I did a quick test of the drivers on the horn using an analog frequency generator on a breadboard electronics lab box, they seem to have smooth response within their passbands. Audibe was a null around 1950hz. I enlarged the mid ports to 3/4 inch (one port per mid). Hoping this works out without messing up the compression driver, but from a frequency sweep it sounded ok, no major problems. I'll be doing some analysis with REW once I get these put back together.
DSP delay settings question
I'm trying to figure out the distance/time delay for the compression driver to match the mids and woofers on the new synergy horn (not the unity horn I originally posted about) and was wondering if anyone could help me understand better how to get delays set properly. My first attempt on these (despite the mid ports being too small, and mid driver efficiency issue) things were sounding good although I could tell something was slightly off somewhere.
I calculated out based on the null I guesstimated around 1950hz (assuming that's correct) I would want a .26ms delay on the compression driver to meet with the mids. Would this get me in the ballpark phase wise between the mids and compression driver?
Can the same approach be used on the woofers to determine approximate delay settings? I thought that one of the reasons Tom Danley placed the woofers in the horn was to offset them for better phase alignment by being closer to the mouth of the horn, but if they actually are delayed from bouncing back into the apex of the horn how does that work?
I also have the 4x10" sealed box woofers and 15" vented sub to figure out how to align to the Synergy horn.
I'm really interested in learning the MEH wizard functionality in Hornresp but I have not been able to understand how to put in all the variables.
In regards to measuring using REW I'm using a Dayton USB mic and was assuming that you can't or don't need to use a loopback reference to understand delay/phase between drivers?
Thanks for any help/suggestions from anyone on this. I'm really close to getting these wired up and running some measurement tests with REW, probably tomorrow afternoon!
-Nate
I'm trying to figure out the distance/time delay for the compression driver to match the mids and woofers on the new synergy horn (not the unity horn I originally posted about) and was wondering if anyone could help me understand better how to get delays set properly. My first attempt on these (despite the mid ports being too small, and mid driver efficiency issue) things were sounding good although I could tell something was slightly off somewhere.
I calculated out based on the null I guesstimated around 1950hz (assuming that's correct) I would want a .26ms delay on the compression driver to meet with the mids. Would this get me in the ballpark phase wise between the mids and compression driver?
Can the same approach be used on the woofers to determine approximate delay settings? I thought that one of the reasons Tom Danley placed the woofers in the horn was to offset them for better phase alignment by being closer to the mouth of the horn, but if they actually are delayed from bouncing back into the apex of the horn how does that work?
I also have the 4x10" sealed box woofers and 15" vented sub to figure out how to align to the Synergy horn.
I'm really interested in learning the MEH wizard functionality in Hornresp but I have not been able to understand how to put in all the variables.
In regards to measuring using REW I'm using a Dayton USB mic and was assuming that you can't or don't need to use a loopback reference to understand delay/phase between drivers?
Thanks for any help/suggestions from anyone on this. I'm really close to getting these wired up and running some measurement tests with REW, probably tomorrow afternoon!
-Nate
Have you tried using the Input Wizard tool to generate a representative example design, to serve as the starting point for specifying your MEH system?
Attachments
Hi David,
The part I struggle with is inputing the horn geometry and also how to represent the sealed back mids which I don't have any parameters for. So far my design approach has been port placement based on crossover frequency using the perimeter of the horn at the port entry as well as placing the mid port as close as possible to the compression driver.
I reassemled them last night and did some preliminary testing. One thing that surprised me was inverting the compression driver was needed to avoid setting what seemed to be too large a delay to get a smooth transition around the mid/high crossover region. I did switch from a Dayton compression driver to a PRV along with different mid drivers and woofers, so everything was modified in the horn except the woofer ports. Thinking the diaphragm location must be different between these two compression drivers. Made quite a difference but still has issues with smooth response but I was able to flatten things out with a few PEQ's. I plan on spending time later today when the wife and kids leave to do some more testing and actually listen to some music for once. The thought that I could have gotten things more optimal and not fully understanding how to interpret phase and impulse response bugs me a little but regardless the result has been one of the best sounding speakers I have built or heard.
Regards,
Nate
Last edited:
How I (sort of) time aligned my Unity
First, my "Synergy" is a Yorkville U15 Unity that I run active. I also made the huge stock mid ports smaller. So it is only a two way ("tweeter" and mids) and one x-over to worry about. I changed stock x-over frequency to 1K Hz. My cancellation notch is about 1.9 KHz. This would equate to a travel time of about 0.26 mSec. Please note my only measurement app is REW and my own ears
Using the technique: invert one driver (e.g. tweeter) and try to find the greatest null with 1K pink noise and/or a sine at 1 KHz. It is important to do this with the x-over in the circuit. Using this, I got a much smaller delay, about 0.03 mSec.* I did a few sweeps around the x-over freq (500 Hz to 2 KHz) and get amazingly flat phase (and freq.) response through the range. I don't claim high skill: this could be dumb luck, faulty readings or both In any case it sounds wonderful. The highs are more important (human hearing), but I suppose you could dial them in and do similar with lower x-overs.
If it matters, my x-over is JRiver, mids are 6 dB/oct Low pass and tweets are 24 dB/oct high pass at 1 KHz.
*If such a low delay is accurate, I can easily understand how those skilled in the art (damn it! I read too many patents.) can align the two ranges with a simple ? x-over and crafty placement of ports.
First, my "Synergy" is a Yorkville U15 Unity that I run active. I also made the huge stock mid ports smaller. So it is only a two way ("tweeter" and mids) and one x-over to worry about. I changed stock x-over frequency to 1K Hz. My cancellation notch is about 1.9 KHz. This would equate to a travel time of about 0.26 mSec. Please note my only measurement app is REW and my own ears
Using the technique: invert one driver (e.g. tweeter) and try to find the greatest null with 1K pink noise and/or a sine at 1 KHz. It is important to do this with the x-over in the circuit. Using this, I got a much smaller delay, about 0.03 mSec.* I did a few sweeps around the x-over freq (500 Hz to 2 KHz) and get amazingly flat phase (and freq.) response through the range. I don't claim high skill: this could be dumb luck, faulty readings or both In any case it sounds wonderful. The highs are more important (human hearing), but I suppose you could dial them in and do similar with lower x-overs.If it matters, my x-over is JRiver, mids are 6 dB/oct Low pass and tweets are 24 dB/oct high pass at 1 KHz.
*If such a low delay is accurate, I can easily understand how those skilled in the art (damn it! I read too many patents.) can align the two ranges with a simple ? x-over and crafty placement of ports.
Refinished horns with grills added.
Kids and cats, too many accidents and mischief in the house to leave woofers exposed. I replaced a 10" woofer due to the dust cap getting crushed by a cat's rear legs when they jumped inside the horn enclosure while assembling (I always buy spares, and PE sent me a replacement one due to distortion/noise issues).
Kids and cats, too many accidents and mischief in the house to leave woofers exposed. I replaced a 10" woofer due to the dust cap getting crushed by a cat's rear legs when they jumped inside the horn enclosure while assembling (I always buy spares, and PE sent me a replacement one due to distortion/noise issues).
Attachments
You do need some way of getting a reference (or holding one between measurements), but a loopback with USB mic doesn't really work well because to do it right, you'd need to loop back using an inaccessible stereo 'record' channel inside the microphone.
Your best approach for getting out relative phase between drivers (which is what you really want) using a USB mic is probably Jeff Bagby's method, neatly summarized by Erik at Is there a clear way to measure speaker time delay, phase and standing waves?
On mine, the tweeter always needed more delay to get toward linear phase (I used an all-pass network on Cosyne to do that, which wasn't easy). The crossover to the mid and woofer adds delay to them, beyond just their positions, the tweeter's didn't (at least in its passband). Which is problematic, because if you physically move the tweeter/compression driver's throat back, it moves the midrange's reflection notch lower in frequency so that the tweeter then has to work to lower frequency.
That is interesting, you mean the 'analog' crossover circuit causes the delays. Most of the time i need to delay mids and woofs to CD horn. Though i must say i am strictly digital crossovering and the CD's in my systems are like way back into to the cabinet. So I assumed the same for this build.
Interesting indeed, definitely have more work to do to get these to minimum phase. What helped immensely was getting the desktop with the digital out working with REW for testing instead of running off a laptop connected to the analog in, which was getting me very frustrated from anomalous measurements.
It took some time to get the frequencies tamed down, especially in the high and low mids, and setting delays based on nulls using the inverted phase approach that Soldermizer mentioned. Also set the lowest possible order crossover slopes. Solved issues getting smooth transition from the mid woofers in the horn to the sealed 10" woofers and ported subs.
How to handle a five way system with a four way crossover? By using two of the channels mono sum for the sub and low, and the six other channels are essentially three way stereo above 70-100hz (some overlay between the low mids and low drivers).
The difference in splitting the woofers this way instead of driving both low outputs in stereo 4 way is startling. Much more controlled/tight bass by band passing the 4x10's and low passing the 15's. Also was able to bridge the Crown amp for the sub channel this way (600 watts, drivers could handle more but not needed).
I've been using/trying Tidal music lossless client, listening to Holst the Planets Suite, Philharmonica Orchestra (released last month) and was very impressed, best recording yet I've heard of this piece. Otherwise pretty much anything I listened to was excellent if the recording was good, very revealing.
It took some time to get the frequencies tamed down, especially in the high and low mids, and setting delays based on nulls using the inverted phase approach that Soldermizer mentioned. Also set the lowest possible order crossover slopes. Solved issues getting smooth transition from the mid woofers in the horn to the sealed 10" woofers and ported subs.
How to handle a five way system with a four way crossover? By using two of the channels mono sum for the sub and low, and the six other channels are essentially three way stereo above 70-100hz (some overlay between the low mids and low drivers).
The difference in splitting the woofers this way instead of driving both low outputs in stereo 4 way is startling. Much more controlled/tight bass by band passing the 4x10's and low passing the 15's. Also was able to bridge the Crown amp for the sub channel this way (600 watts, drivers could handle more but not needed).
I've been using/trying Tidal music lossless client, listening to Holst the Planets Suite, Philharmonica Orchestra (released last month) and was very impressed, best recording yet I've heard of this piece. Otherwise pretty much anything I listened to was excellent if the recording was good, very revealing.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.