doityourself Posted March 29, 2009 Share Posted March 29, 2009 My homebuilt khorn plans from speakerlab has a bass contour control and am wondering if it's more than an L-pad wired in with the woofer. I'm running the 100db Audax midrange but know that the bass has a pronounced bump in the 180hz range. I've tamed mine somewhat with a couple of coils wired in series for perhaps a total mh of 9 to help roll off some of the midbass. I've thought about designing a notch filter centered on 180hz. Perhaps just a 100watt Lpad and cutting the woofer output by 4db might sufffice. Any thoughts? Quote Link to comment Share on other sites More sharing options...
Guest " " Posted March 29, 2009 Share Posted March 29, 2009 sounds messy....big time deviation from the stock material I have. Quote Link to comment Share on other sites More sharing options...
doityourself Posted March 31, 2009 Author Share Posted March 31, 2009 Well, all of this talk of the tractrix horn made me want to build one for my audax midrange. Got one built tonight from Edgar's paper on using a horn with cone speakers. It's the 300hz horn (9x18inches-6inches deep). Will run some freq. response tests later tonight. My thinking is that if the midrange was stronger I might be more willing to put with the bass unit's strong upper bass hence the horn. Should increase db output by 5-7db. We'll see. Quote Link to comment Share on other sites More sharing options...
doityourself Posted March 31, 2009 Author Share Posted March 31, 2009 I just ran the test. My meter was about a meter in front of speaker. Turned tweeter all the way off-woofer still running. The horn starts to come alive at 630 hz (up 5db) and is generally up about 3db over the rest of the midrange (using 1/3rd white noise) but at 4khz actually looses 3db! My wife just held up the horn to the speaker. In Edgar's paper he said best results were with a 1 inch space between the speaker and throat. I'm thinking I need the speaker sealed tight to help out the very low end of horn fs. He was getting a peak at 500hz he was trying to tame. I was looking for an average 3db gain and I think this got it. Edgar came up with aprox. 5db gain (at least he rated the horn with an earlier version of my audax as a 105 db efficiency horn). I just need to add the speaker board and stiffen up the sides of the horn. The sound reminds me of an altec horn system (thinking movie theatre or an model 19 Altec). Not sure I'll like the "horn" sound but I'll build another horn and tweak the bass unit crossover with just one 5.4mh coil, increase the tweeter output and do some listening. Quote Link to comment Share on other sites More sharing options...
Marvel Posted April 1, 2009 Share Posted April 1, 2009 The horn will need to be mounted to the baffle with the Audax. You won't have any valid numbers just holding it up to the cabinet. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 2, 2009 Author Share Posted April 2, 2009 I put a 10x10inch baffle on it last night and sealed against the speaker it didn't change the readings much. You can tell when you get with 2 inches of the speaker the sound changes. In Edgar's paper he actually preferred a 1inch gap between and older model audax and baffle to kill a 500hz spike. Can't decide how to finish the project. Quote Link to comment Share on other sites More sharing options...
colterphoto1 Posted April 2, 2009 Share Posted April 2, 2009 "I just need to add the speaker board" Like Bruce said, this will change everything. Now your woofer will be acting as a compression driver with all the waveform entering the horn. I'd look for even more output on axis when you complete the mod. EDIT- I typed this before reading your last post- I'm surprised at your results. Did you stiffen up the sides of the horn? Michael Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 2, 2009 Author Share Posted April 2, 2009 Here's a more finished product. I couldn't see myself cutting up the old boxes yet so I removed the T35 and put it on top of the speaker and added a baffle to horn and painted it flat black (need to do some more painting). Made the otherside one and listened and like it! I may finish these off in a box of their own or leave it alone and remount the tweeter back in the box horizontal to clear the horn roof. Not a perfect job by any means. I'm not going to reinforce the sides as the 1/8inch masonite is in stress and plenty "stiff". Don't think I detect any "ringing". Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 2, 2009 Author Share Posted April 2, 2009 Here's a full length picture. Bottom unit is the later Speakerlab model that has been scaled down on the outside but internals same dimensions as Khorn. Quote Link to comment Share on other sites More sharing options...
JL Sargent Posted April 2, 2009 Share Posted April 2, 2009 You mention ringing. I thought only the metal horns ringed. Can other horn materials ring too? Interesting project. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 5, 2009 Author Share Posted April 5, 2009 I mentioned ringing but perhaps resonating is a more proper term. I haven't owned any midrange horns before to know of the affect. Here's a picture of the more finished product. Have to finish the outer corner trim, stain, and finish. Ran a freq. response with Audax sealed against horn and it is very flat except between 3-4khz but comes alive again at 6.3 (typical Audax response-perhaps a lower crossover to kill energy at 6.3 is needed). No horn coloration or maybe I got used to it already. The horn really helped the lower range of midrange (500-800hz). Balances nicely with the khorn upper midbass bump. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 5, 2009 Author Share Posted April 5, 2009 Here's a freq. resposne graph. It does show the dip at around 4khz. My "unhorned" audax showed the same dip. Nothing to write home about but with 1/3 octave noise with my meter it shows within 2db from 500-2khz. Perhaps running the T35 down to 3500 and using a 12db filter on the audax would get rid of the dip. Quote Link to comment Share on other sites More sharing options...
Marvel Posted April 5, 2009 Share Posted April 5, 2009 I have some Galaxy Audio 5 inch drivers (used in the Hotspot monitors) and have contemplated using one for a mid driver for my LaScalas. They ae rated 200-18k. What is the eff. of your Audax? Looking good, btw. Bruce Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 5, 2009 Author Share Posted April 5, 2009 The efficiency is rated at 100db/watt but some charts show only 95 db eff. between 400-1khz and than a rising response. There has been some conflicting graphs among the years. The midrang Edgar Horn article suggested using a JBL LE-5 driver with a 2inch throated horn (just longer than mine) for a 105 efficient horn. Not sure how hard those drivers are to get a hold of. The Audax is about $83 at Madisound. I saw a pair of Audax midranges (earlier model than mine but same model as Edgar used) that didn't sell for $35 on ebay a couple of weeks ago. Edgar mentioned that some cone midranges don't make good horn drivers but doesn't mention what models he tried. Just make a simple horn like mine on the first photo and experiment. Start off with a 2 inch throat and chop it off if needed. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 15, 2009 Author Share Posted April 15, 2009 I ran the T-35 down to 3500hz with a coulple of more capacitors to fill in the gap nicely in the 4khz range. Here's the finished product. Sounds nice. Quote Link to comment Share on other sites More sharing options...
DrWho Posted April 15, 2009 Share Posted April 15, 2009 You say the dip shows up with the driver without a horn too? Could you post that measurement? You're measuring just the driver by itself right? No other speakers in the mix? And no crossover? What you're showing is very odd behavior for a driver. Looks very much like some form of phase cancellation or cabinet vibration. Do you have the ability to play solid test tones and sweep them around? I would try sweeping tones at somewhat elevated levels (not too loud) to see if you can identify any mechanical noises. Do both drivers measure the same? Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 16, 2009 Author Share Posted April 16, 2009 Here's a chart I posted a yr ago or so. Not the same style of chart but you can see the dip around 4khz. I need to produce an updated chart with lower tweeter crossover freq. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 16, 2009 Author Share Posted April 16, 2009 Here's a graph with the revised 3500 hz tweeter crossover. Still have a dip beginning at 2.5khz but new crossover did fill in dip somewhat. Dip is barely noticeable on 1/3rd octave white noise test. I can't say I hear anything lacking. The T35 has a noticeable rise from 3.5 to 8khz or so flattens and than drops. Some testing with only midrange horn running the response comes back at 5khz and extends to 6.3khz but is 4db shy of what it is at 1khz. I'm running a coil or roll off top end that is calibrated for an 8 ohm speaker at 5khz but this driver has about 15 ohm impedance at 5khz thus the coil isn't doing much. Quote Link to comment Share on other sites More sharing options...
doityourself Posted April 16, 2009 Author Share Posted April 16, 2009 Here's a graph by reversing the tweeter polarity and also reverseing the wiring to the midrange-tweeter. It solved the "suckout" at 3.5khz and caused a greater "suckout" at 300hz. I'm relieved now that the upper suckout is fixed. I didn't think my new "horn" rolled off that bad in the upper range. I'll see if it actually sounds different. Quote Link to comment Share on other sites More sharing options...
DrWho Posted April 17, 2009 Share Posted April 17, 2009 Here's a graph by reversing the tweeter polarity and also reverseing the wiring to the midrange-tweeter. It solved the "suckout" at 3.5khz and caused a greater "suckout" at 300hz. I'm relieved now that the upper suckout is fixed. I didn't think my new "horn" rolled off that bad in the upper range. I'll see if it actually sounds different. Just reverse the tweeter and leave the mid polarity normal and your 300Hz will come back. Btw, when you disconnect speakers in a passive xover to measure independant drivers, you should attach a resistor in place of the speaker to emulate the speaker load to ensure the passive xover works properly. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.