| zaph type waveguide measurements, request for help - Click HERE for Original Thread |
| mazurek |
I have produced the following measurements, but in general have failed in my waveguide experiment, and keep racking up problems.
Refer to http://www.zaphaudio.com/hornconversion.html for background.
I tried a new mounting method whereby I sink a screw head in epoxy in the wells formed by the waveguide ribs. I added epoxy over three days, and let it dry thoroughly. As soon as I mounted the tweeter to the waveguide, the epoxy completely separated. Anyone know a better material to bond to abs for me to sink screw heads in? After that failure, I just mounted the tweeter using screws compressing it onto an old baffle.
I got some haphazard measurements of before and after. My goal for this experiment was to find the difference in magnitude in order to develop a notch filter so I can update my crossover when I finally get working waveguides. The measurements are done using TrueRTA in my room, raw measurements are shown, as well as those smoothed using a 30 average forward and backwards running average. This should be adequate to get some starting crossover values to work from.
Zaph was right, the second rib of the waveguide is pretty much spot on for the Seas 27TDFC.
I'd like to ask if anyone would be willing to modify some of the mcm horns for me. I asked the superviser of the machine shop in my engineering department at school if she or one of her students could quickly help me when they had the chance. She flat out told me to go away. I'd rather not wait until the school year to get a friend to do it under the table for me, and don't want to have anyone face her wrath.
I did the first modification by putting the waveguide onto a drill chuck and turning it with a saw, an extremely dangerous diy lathe. This would be quick work for anyone with an actual lathe. Besides the danger and lack of precision of my stupid method, the abs needs to be machined faster than I can do, or it deforms.
Sorry about the cropped picture, diyaudio limits the resolution, and I didn't want to rerun the scripts and format it again. |
|
|
| mazurek |
| It's such a shame the tool shop is idle most of the time. The fellow in my lab that builds all our setups does ridiculous things by hand because the supervisor hates him. |
|
|
| noodle_snacks |
| A router, straight bit and an appropriate jig is all you need to modify the MCM horns. Maybe you should try approaching a relative or a local woodworking club or something. You need a box that is open on two sizes, big enough to slide the horn in, with a third opening on the top, using a straight bit take off small amounts (testing with the tweeter itself) until you have it the right size, its not difficult as long as you are careful. Let me know if this doesn't make sense. |
|
|
| mazurek |
I actually talked to a buddy here with a router, and he was uncomfortable about the brass insert. I know brass is easy to machine, but any special advice for the router? Do you just go around the brass? I feel real uncomfortable telling other people just to go for it.
About your box instructions, I'm getting the idea you just make a box to put the face of the horn back flush with the face of the router baseplate? |
|
|
| noodle_snacks |
| Cut the brass section off with a hacksaw or similar first. I'd take 3mm passes with the router until i was within 3mm, then adjust in very small steps from that until it is just right (if you are clever you can do the final pass on both at the same time and save a fair bit of fiddling). The box would basically let you place the horn face down with the router on top (sort of like the badly drawn attached picture, you should be able to see the horn underneath with the platform for the router) |
|
|
| mazurek |
| Thanks, I need a better hacksaw. |
|
|
| bluebeard |
I dunno. Quick judgement based on the fact that you're interested in something that I also happen to be.. and that I've had very similar run-ins with supervisors of depts in schools who really should NOT be in the position they inhabit for such a lack of interest in student's wholesome curiosity..
..but I think you ought first to write a letter to some higher authority in your school and see if you might add some weight to any other discontent she may have generated up till now. Ref this forum.
Maybe she's hugely talented, but she might be better off laboring away in a cubicle on some widget that really needs her attention instead of disappointing the ambitions of her students.
I could be wrong. Maybe you're just a pest. But I doubt it.
Personally, I've been toying with the idea of using one of the MCM waveguides to cast a mold so that I might make a concrete baffle. Nylon inserts could be cast in place to match the tweeter mounting holes. |
|
|
| ralphs99 |
Hi Mazurek,
My method for cutting down the MCM horns was to cut with a hacksaw and then grind to the desired throat opening with a belt sander. It worked perfectly and only took 10 minutes to do! A belt sander will cut through metal or plastic quite happy.
Personally, I've abandoned the idea of using the MCM or similiar horns as a wavegiude for a dome tweeter as it doesn't really work very well.
Cheers, Ralph |
|
|
| mazurek |
bluebeard, I agree that the school has a problem. The mechanical engineering machine shop does not even let students in whatsoever. The industrial engineering machine shop at least offers a class for students, which I took. I don't know why the people that run them are so negative towards student projects, the same superviser that shot me down yesterday was incredibly helpful and positive when I took her class.
raphs99, please explain to me what you have moved on to. I thought the MCM horn as a waveguide as a good idea to deal with diffraction issues in a larger baffled speaker I'm making. The woofer dictates the baffle to be at least 11.5" wide. Do you use compression tweeters. I understand there are many out there that are of excellent quality, but am very unfamiliar with them. |
|
|
| sreten |
| quote: | Originally posted by ralphs99
Personally, I've abandoned the idea of using the MCM or similiar horns
as a wavegiude for a dome tweeter as it doesn't really work very well.
Cheers, Ralph |
Hi, care to elaborate ? :)/sreten. |
|
|
| mazurek |
Yeah, at least on paper they are a solution to me for:
time alignment
diffraction
power response
Other solutions to these issues are:
low tweeter mid crossovers
diffraction equalization in crossover
sloped baffles
It seems as though the waveguide is a simpler solution. I'm still reading your site though for tips. |
|
|
| ralphs99 |
Hi Mazurek,
Don't get me wrong, waveguides have a lot going for them! As you say, time alignment, diffraction and power response are key advantages. But they're not easy to implement.
The MCM horn is after all an exponential horn. It's designed to be coupled to a compression driver. My testing with JBL 2407 and 2426 compression drivers shows that it works extremely well in this capacity- an almost perfect power response.
A compression driver tends to output a flat wavefront that couples to the throat of the horn and gradually expands via the horn conture. I'm not a horn expert, so I'm sure that there are better explanations of the propagation methods of horns.
So just sticking an amputated exponential horn on the front of a dome tweeter and expecting to work perfectly is a bit optimistic. With all due respect to Zaph, he does excellent work and never made any claims for the MCM horn that weren't met. Time alignment- Yes, Diffraction control - Yes, Even power response vs frequency - No!
For want of a a better description, the wavefront propagating down the horn doesn't seem to stick to the horn walls properly and the horn loses directional control as frequency increases. The result is we're back to having a falling power response vs frequency.
The power response I have measured with the MCM horn coupling to a dome tweeter is rather poor. It is arguably a bit better (more consistent) than a bare dome tweeter, but far from ideal. I'm guessing that the problem is due to the spherical wavefront produced by a dome shaped diapragm not matching the horn that expects to see a flat wavefront.
A better alternative is an oblate spheroid waveguide. This appears to be the waveguide shape you often see in comercial products such a Genelec, JBL, Mackie and others. This shape produces a much more even power response.
Additionally there can be also be a penalty in the on-axis repsonse for any kind of waveguide coupled to a dome tweeter. The coupling between diapragm and throat is very critical. Standing waves are easily setup across the throat giving rise to a notch in the on-axis amplitude response.
If you look carefully at Zaphs amplitude response plot, you will see a 3dB dip at 10kHz and some funny behavoiur close to this frequency. I observed similar behaviour in my experiments both with the MCM wavegiude and also with an oblate spheroid waveguide. Interestingly, the effect is only seen on axis (or close to it). As such it is non-minimum phase behaviour and it cannot be EQ'ed out, as then the power response is upset. Putting a physical barrier across the throat gets rid of the problem (or at least moves it to an ultra-sonic frequency), but creates another problem due to off-axis cancellation. Ultimately a compromise is necessary between the two effects.
I'm planning on completing and writing up my experiments soon complete with graphs etc. Hopefully this weekend!
Cheers, Ralph |
|
|
| mazurek |
Thank you for the detailed response. So it seems that waveguides are worth pursuing, just its more complicated than putting a horn on a dome.
I eagerly await the results of your experiment. I won't ask any more questions until you post your results, so you can spend all your typing time on your results. |
|
|
| mazurek |
| I was looking again, it says mackie uses a logarithmic waveguide, I also looked up oblate spheroid. Is an oblate spheroid just a sphere cut out of a solid? I suppose on second look, they don't use a logarithmic shape so I guess that must refer to dispersion. Did you buy your other waveguides or manufacture. Depending on your results, maybe I will try making some by cutting out several holes at multiple levels and filling in between with some wood filler. |
|
|
| ralphs99 |
Hi Mazurek,
It's up! Fresh and ready to go. I hope you find some answers and some questions.
http://www.aeronet.com.au/waveguide.htm
I've also started a new thread to deal with any discussion that arises from the new page.
Cheers, Ralph |
|
|
| mazurek |
Thank you very much for posting these results. I'm awaiting a response from my friend that worked in the CNC lab last year, I believe he may still work there. If he does, I may get one shot at making a custom waveguide.
I just disassembled a tweeter, I did not realize the voice coil is attached to the face plate, this limits experimentation options. It looks like the only dome tweeter with a suitable faceplate is the Seas Millenium. Otherwise I would need to use the Vifa XT-25, or super expensive scanspeak ring radiator.
Will have to mull this one over, as I will most likely get one shot. Also, I'm thinking of using a 5" mid crossed to a 5" waveguide, the tweeter would extend below the mouth frequency of the waveguide, but would have similar directionality to the mid a the crossover frequency. |
|
|
|
|
