MEH speakers look pretty neat! Many of them seem like more output (+ expense) than I think is necessary for my space, so I'm considering a 2 way with a 2" and an 8". I have a sub to take care of the really deep stuff. I do sort of enjoy the conceptual "Wow! / Meh" juxtoposition.
I've kinda wanted to do a 2-way with a 2" FR as a low tweet for a while, but they're so low efficiency. I started thinking about wave guides for the tweeter, but then thought maybe it should be a MEH. The idea being to increase sensitivity on the 2" down to 500 Hz or so, and cross to the 8" between 500 and 1k, and hopefully keep the horn size domestically compatible. I have active / DSP available, but would like to be able to make a passive XO for it.
These are some buyout drivers I got a while back. These drivers are NLA
2" - Tymphany TC6FC02-04
https://www.parts-express.com/Peerl...ll-Range-Line-Array-Drive-264-1350?quantity=1
8" - DynaLab V22-DR-0003
https://www.parts-express.com/DynaLab-V22-DR-0003-8-Poly-Woofer-4-Ohm-299-4004?quantity=1
So I started to monkey with HornResp. I don't really know much about HornResp, so we'll see if the brain trust thinks I've done it right...
I started out trying to enter T/S directly, but that seems a little wonky and the T/S wizard seems better.
It looks like the MEH wizard simulates response like this... (2")
(and 8")
So, questions.
Hmmm... anything else? I'll come back as I think of more questions / make progress...
Thx
I've kinda wanted to do a 2-way with a 2" FR as a low tweet for a while, but they're so low efficiency. I started thinking about wave guides for the tweeter, but then thought maybe it should be a MEH. The idea being to increase sensitivity on the 2" down to 500 Hz or so, and cross to the 8" between 500 and 1k, and hopefully keep the horn size domestically compatible. I have active / DSP available, but would like to be able to make a passive XO for it.
These are some buyout drivers I got a while back. These drivers are NLA
2" - Tymphany TC6FC02-04
https://www.parts-express.com/Peerl...ll-Range-Line-Array-Drive-264-1350?quantity=1
8" - DynaLab V22-DR-0003
https://www.parts-express.com/DynaLab-V22-DR-0003-8-Poly-Woofer-4-Ohm-299-4004?quantity=1
So I started to monkey with HornResp. I don't really know much about HornResp, so we'll see if the brain trust thinks I've done it right...
I started out trying to enter T/S directly, but that seems a little wonky and the T/S wizard seems better.
It looks like the MEH wizard simulates response like this... (2")
(and 8")
So, questions.
- Have I done a reasonable job setting this up? I might wish for more HF extension from the 8", but I think it has to be some 11 cm down the horn given it's diam. I think these dimensions makes 27 deg center-to-wall, or 54 deg wall-to-wall for the horn, and 40 / 80 deg for the end flare to smooth the horn exit (square horn shape). Am I missing any best practices for choosing horn dims?
- I see the power predictions for output pictured above. I didn't see any directionality prediction, is that in HornResp? Horns are all about the directivity so I was expecting something like that. On the other hand, I only specified sectional Sq Cm not horizontal and vertical dims so maybe HornResp couldn't hazard a guess. I had to make a spreadsheet to estimate horn the angles.
- That woofer null I presume comes from the wave entering the horn part of the way down, and some of it going back up the horn and then coming back down and interfering. That makes sense to me. But also the woofer front chamber and horn entrance is going to be a bandpass, right? Does HornResp have any facility to predict that, or help choose the entrance holes for the 8"? I was considering going to a different program to simulate a bandpass to investigate entrance sizes / velocity.
- HF response on the 2" falls off. 2" spec is 85 dB @ 2.8 V and basically flat to HF on-axis. Other simulators frequently show a lot of HF roll-off that doesn't appear on-axis. I'm sort of assuming that is what's going on here (total output vs on-axis?). Does anyone have experience with how real that might be?
Hmmm... anything else? I'll come back as I think of more questions / make progress...
Thx
Consider that you may be making it work below it's optimum range, as uncompromised 1kHz waveguides are not small.
It is somewhere..
Yes, the VTC (volume of throat chamber) will bandpass the woofer.
Your drivers are four ohm nominal, the 2" sensitivity spec is 85.9 dB/1meter at 2.83 volts (two watts) around 82.9 at one watt (2 volt).
The conical horn won't increase the 2" sensitivity much above 1kHz or so.
That said, upper sensitivity rises to ~87.5dB on axis above 1kHz.
Check that it does not exceed it's 1.55mm Xmax at SPL levels you'd like to achieve at your listening distance (-6dB for each doubling of distance, 88 dB @1m, 82dB @2m, 76@4m).
I'll keep looking...!
Oh, VTC, ok I see it. And ATC... and AP1 / AP2 and the length, like a horn segment. Oh man, I had just figured to have a hole (cylindrical), but put expansion (reduction?) on it? I wonder if that does anything useful, I'll have to play with it.
But I see the bandpass upper frequency on the woofer as a hole in response for the 2". That makes sense, a resonance going both ways. It looks kind of alarming but I wonder if it's really something to stress about.
I figure the radiation from the 2" will eventually narrow enough to not get much off the horn. So I've been wondering where that transition would be. The sim predicts output above nominal (85ish dB, sim is also at 2.83 V) until about 4k. But then it predicts falling below that level, which I'm not sure I trust. The FR from the data sheet suggest 60 deg off-axis is holding on until about that same 5ish kHz (give r take). If the driver is wide dispersion up that high (it's only a 2") I figured it would still be feeling the horn as a load / getting sensitivity. Am I off-base?
Thanks for your input!
Using a small full range driver, depending on the horn's design and wall angles, probably looking between a -6 to -20 drop before a throat dimension notch, then a Hf shelf.
Using a 3.5" Tymphany TC9FD-18-08 the response difference between it on a cardboard 45x30 degree horn and front loaded on a 1liter bottle looked something like this:
There is a lot of throat detail work required to get the most out of small "full range" drivers:
https://www.diyaudio.com/community/...point-source-horn.285030/page-63#post-4644522
A quick cardboard or foam core mock up can give a rough idea of what your full range driver will do.
Art
Attachments
Oh, that's a big thread, but it looks right on the money. They got good HF results, but you think that's b/c of their more advanced horn geometry, have I got that right? I'll see how much of that thread I can wade through.
Conical horns have polar response drawbacks compared to other horn profiles.
The throat transition area will create a problem for any 2" cone driver.
A conical family waveguide should normally be adapted at each end as to have no significant issues, regardless of the driver.
Danley initially went with a squared cross section (as opposed to round). This helped with arrayability and probably made construction easier, but it also gave a place for the ports to hide in the corners as he explained. Such a view tends to indicate the compromises you can get away with. Not that a near perfect waveguide isn't something that's worth striving for in general..
Danley initially went with a squared cross section (as opposed to round). This helped with arrayability and probably made construction easier, but it also gave a place for the ports to hide in the corners as he explained. Such a view tends to indicate the compromises you can get away with. Not that a near perfect waveguide isn't something that's worth striving for in general..
Cause it seems like fun... The spatial coincidence thing seems like a cool aspect. The directivity seems interesting also, tho I'm trying to keep the horn mouth to 1 sq ft (give r take), so maybe it doesn't keep control as low as larger examples. Mostly as an opportunity to learn. I've already learned and refined my thoughts a fair bit, so thanks to all for the comments!
it looks like @bushmeister managed to get pretty good results after doing some throat work. I've got a 3D printer, and have already done up CAD for a circle to square adapter to fit my 2.5" driver to a 6 cm square. But I see bushmeister brought his adapter in over the surround and used some wool, so I'll have to keep at the thread to see if I can find more detail into that work. I've got other questions about how they did what they did as well, I'll prob pop in there to ask after I've read some more. I've seen a number of neat @xrk971 projects by now, and that seems like one of them. But their results look positive, so are you suggesting that I would need equiv throat transition tricks to get as good on-axis, but my polars would never be as good based on a conical horn?
I can print something smallish at the throat end. At the mouth most examples I see have a segment that opens a bit faster to smooth the exit, and I figured I'd do that. But most of the horn will be in wood as my print volume is 200 mm cube (or a bit less, likely).
Yes.
I've compared a number of conical horns polar response to other horn designs, they just are not as good as other "constant directivity" designs can be.
Conical horns are not bad, but since you have 3D printing ability, no need to limit your prints to flat surfaces, especially in the critical throat area.
I think combinations of 3D print, plywood and "bendy ply" could give the best performance/$$/ease of construction ratio for DIY MEH builds.
Art
Last edited:
Hmmm, that's a substantial widening of the decision-space. I'm on-board, but I'm going to need to learn some more horn stuff before I can chart a course.
... especially since the Directivity tools only seem to activate for single segment horns.
How will I investigate improved polars with an initial exponential section 🤔
How will I investigate improved polars with an initial exponential section 🤔
You could try to estimate the wavefront change on just the throat and exponential sections to use as an entry for the conical in hornresp. Perhaps this separates the task for printing if you simply aim to get spherical or flat out of the first part.
And in lies your fundamental problem…..you’re now constrained to constant directivity down to maybe 800hz so given the multi driver MEH vs say an equal size 1.4” compression driver on a horn, there’s nothing gained for the effort but in fact, likely much better performance given the passband of the single compression driver option. Think of it this way…..a Mack truck with an 18 foot trailer.
