By the way, while Herr Dr. Waslo is tuned in: Thanks so much for sharing your spreadsheet. I didn't follow it strictly, but it was very helpful as I followed a related path. A thousand thanks for generously sharing the product of your labors!
Few
Few
I thought the slots would be invisible but no such luck. They’re right against the corners and in the 90 degree face of the horn. Perhaps I broke some other rule? It’s a problem only EXACTLY on-axis so maybe some just ignore it. I guess I’ll add it to the very long list of learning opportunities this pursuit provides.
Few
Few
Yep, so much to learn about these things (synergy's/unity's).
And another yep about some folks maybe ignoring EXACTLY on-axis. Me for one 🙂
Like i mentioned earlier, even with no ports whatsoever in the horn, when the mic is dead center both H and V, i've measured some weird stuff coming from the CD.
Fortunately, it only takes a little mic movement to make the anomalies mostly disappear.
I've come to use 10-15 deg off-axis horizontally, with vertical anywhere close to on-axis, as the reference axis to tune to.
Usually this gives the nicest overall set of polars horizontally. I haven't done near as much work with vertical polars.
One thing that keeps surprising me about port location is, I can't measure justification for putting them in the corners.
A round hole in the center is working as well as anything I've tried in the corners...actually a little better.
Perhaps this is due to the fact my builds use only one cone driver size (12, 10, or 8") to get up to the CD xover (480Hz to 650Hz).
Which we all recognize doesn't require any small ports in the throat very close to the CD, for mids to be able to reach higher in frequency.
So port centers are generally in the neighborhood of 5" from CD, maybe far enough away from CD throat not matter as much...dunno...just know what i measure.
Last edited:
Few,
My SynTripP design did not seem to have any on-axis high frequency dip such as yours, though I'd be surprised if any of my measurements could be verified to within 1 degree accuracy ;^). The four SynTripP mid entry holes had similar placement, though had different diameters on either end, and were slightly rounded, while yours appear symmetrical and hard edged.
What distance from the horn throat and mouth was the microphone position for the on axis measurement made in the OP that shows the dip?
Did you find the on-axis 4.8kHz dip at other distances also?
The mid range output seems to have a peak at almost the same frequency, are the exit ports about 2.75" length end to end?
Art
Same here. Not enough samples to conclude that it was any kind of rule or anything, but having ports in the corners always seemed to give somewhat worse measured results.
I'm thinking now that maybe using the little half-pipe covers (or something similar) with different lengths and positions on each port might be the way to go -- to avoid the deep suckouts and break up the symmetry a little without making the radiation pattern wonky. A neat thing about the covers that FEW came up with is that they would seem to allow for easier experimentation, as you could move them around with some temporary stickup to find how they best work out. A hell of a lot easier than moving ports around or messing with their shapes.
+10,
Thank you again and again Bill,
I would never have been able to undertake a synergy without your speadsheet.
Thank you again and again Bill,
I would never have been able to undertake a synergy without your speadsheet.
Thanks, all, for your interest and thought-provoking questions. When working solo, it's very easy to get stuck in an unproductive mindset. I appreciate the reality-checks.
I use a diy turntable marked in 5 degree increments so I think estimating 1 degree horizontally isn't too crazy. The vertical mic position is repeatable to 1 mm or so. 1 degree at 1 m corresponds to a displacement of ~17 mm so that's not too hard to track. Bottom line: I don't think I exaggerated the angular resolution by much.
The notch doesn't disappear 1 degree off axis, but it is reduced noticeably and reproducibly with displacements on that scale. All the measurements I shared so far were made at the same 1 m distance I mentioned in there somewhere (1 m from the mouth of the horn). I haven't measured systematically at many distances, but the notch did show up at distances other than 1 m.
The woofer slots are about 67 mm long I believe. That sticks in my memory based on a 4.8 kHz wavelength calculation I did when coming up with reasons for the notch.
The woofer slots are fairly sharp-edged because they're deeply chamfered on the woofer side (smooth on the interior of the horn). It's an interesting question whether rounded or sharp edges yield better results. I was mostly trying to reduce the resonance in the slot-woofer system; I wanted to avoid a thick neck of uniform thickness that would maximize the Q of the Helmholtz resonance I was plagued by previously. I don't know if that made the tweeter disturbance more pronounced or not.
Bill, I agree playing with the half-tubes is very convenient. I just used double-stick tape to tack the tubes in place for testing. Since the notch disappeared with my first (symmetric) arrangement, and I was leaving town in a few hours, I didn't play much with the many possibilities. I'd be very interested to hear from anyone who tries a similar approach on their own speakers.
I'm not sure what to make of the mixed reports regarding slot position. I've heard from many sources that putting them near the edges is not the silver bullet that is sometimes stated. I see the logic of that argument, but there's no disputing experimental observations!
Few
I use a diy turntable marked in 5 degree increments so I think estimating 1 degree horizontally isn't too crazy. The vertical mic position is repeatable to 1 mm or so. 1 degree at 1 m corresponds to a displacement of ~17 mm so that's not too hard to track. Bottom line: I don't think I exaggerated the angular resolution by much.
The notch doesn't disappear 1 degree off axis, but it is reduced noticeably and reproducibly with displacements on that scale. All the measurements I shared so far were made at the same 1 m distance I mentioned in there somewhere (1 m from the mouth of the horn). I haven't measured systematically at many distances, but the notch did show up at distances other than 1 m.
The woofer slots are about 67 mm long I believe. That sticks in my memory based on a 4.8 kHz wavelength calculation I did when coming up with reasons for the notch.
The woofer slots are fairly sharp-edged because they're deeply chamfered on the woofer side (smooth on the interior of the horn). It's an interesting question whether rounded or sharp edges yield better results. I was mostly trying to reduce the resonance in the slot-woofer system; I wanted to avoid a thick neck of uniform thickness that would maximize the Q of the Helmholtz resonance I was plagued by previously. I don't know if that made the tweeter disturbance more pronounced or not.
Bill, I agree playing with the half-tubes is very convenient. I just used double-stick tape to tack the tubes in place for testing. Since the notch disappeared with my first (symmetric) arrangement, and I was leaving town in a few hours, I didn't play much with the many possibilities. I'd be very interested to hear from anyone who tries a similar approach on their own speakers.
I'm not sure what to make of the mixed reports regarding slot position. I've heard from many sources that putting them near the edges is not the silver bullet that is sometimes stated. I see the logic of that argument, but there's no disputing experimental observations!
Few
When cutting plastics always turn the blade backwards to avoid those issues, and keeps the plastic from splitting...😉