Deang Posted July 20, 2005 Share Posted July 20, 2005 I decided to try a bandpass in the squawker section of the AK-3. I think the improvement is notable, and since it's such an inexpensive and simple mod, AK-3 users may want to give this a try to see what they think. I've been wanting to try this ever since I read through PWK's patent where he included a network utilizing a steep slope bandpass for the K-55-V -- where the filter aggressively cuts the output at 5.5kHz. Also, in that particular network he sized his values so he could take advantage of a solid state amplifier's ability to increase output at lower impedances. The result was what he referred to as a "peaking circuit" that lifted the K-55-V's falling response at the crossover point. Since I'm primarily a tube amp user, and am currently using K-55-M's --there wasn't much point in building that particular filter. However, based on the info in the patent, and all the positive listening experience I've had with the bandpass on both John Albright's first order design and the ALK -- I decided to put one into my AK-3's. Now, a big difference here is that the former networks are constant impedance designs, and the AK-3 isn't -- so I based the inductor value on 30 ohms -- which was really a "best guess". The mod necessitates the removal of the 5mH inductor in parallel with the squawker, and then the simple addition of a .9mH air core in series with the 13uF primary capacitor. The inductor is placed between the cap and the input tap of the autotransformer. I come up with a FqL of 403Hz and a FqH of 5246.7Hz. It sounds very good, but I would like to see what the transition looks like to the AK-3's steeper sloped, elliptical tweeter filter. Al? Quote Link to comment Share on other sites More sharing options...
Deang Posted July 20, 2005 Author Share Posted July 20, 2005 In the bottom right hand corner you can see my hack job to get the inductor in. Actually, it's just sitting there. I didn't have anything fancy laying around when I decided to do it so I used an old Dahlquist DQ-10 1.27mH air core and peeled it down to the needed .9mH value. Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 20, 2005 Share Posted July 20, 2005 Guys, In case anybody is interested, I put this mod in the computer to test it. It works! I think it's a good mod. Al K. Quote Link to comment Share on other sites More sharing options...
Seadog Posted July 20, 2005 Share Posted July 20, 2005 Dang it, now I have another project to add to my list! Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 20, 2005 Share Posted July 20, 2005 Guys, Here's the computer plots before and after Dean's mod. No more 400 Hz to light trying to come out of the squawker! Al K. Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 20, 2005 Share Posted July 20, 2005 Dean, You asked for it! Here's the ES-Khorn (ES400T + ES5800) ploted using the same scale factors and the Dean mod plots (2 dB / div and 5 Ohms / Div). Al K. Quote Link to comment Share on other sites More sharing options...
mikebse2a3 Posted July 20, 2005 Share Posted July 20, 2005 Dean said: The mod necessitates the removal of the 5mH inductor in parallel with the squawker, and then the simple addition of a .9mH air core in series with the 13uF primary capacitor. The inductor is placed between the cap and the input tap of the autotransformer. I come up with a FqL of 403Hz and a FqH of 5246.7Hz. It sounds very good, but I would like to see what the transition looks like to the AK-3's steeper sloped, elliptical tweeter filter. Al? ---------------------------------------- Hey Dean and AL Its been a while since I looked at this but I thought the 5mh inductor in parallel with the squawker is used to deal with the crossover between the woofer and squawker in the AK-3? Since the impedance of the 5mh inductor goes up with frequency I thought it was pretty much out of the way at the upper end of the squawker but was being used to balance the response of the squawkers lower frequencies with the woofer's upper frequencies. If this is the case then could the 5mh inductor be left in the circuit to keep the transition between woofer/squawker the same as was orginally designed? mike Quote Link to comment Share on other sites More sharing options...
Deang Posted July 20, 2005 Author Share Posted July 20, 2005 Al, -- since you and Bob have me so gun-shy about trusting my ears, can you show me what the transition looks like -- what does the AK-3 tweeter filter look like with the new squawker curve (I'm worried that I'm cutting the squawker too soon). Mike, -- interesting. Al can either model it, or I can stick the inductor back in for a listen. I have a feeling it's a one thing or the other proposition. Quote Link to comment Share on other sites More sharing options...
mikebse2a3 Posted July 21, 2005 Share Posted July 21, 2005 Dean I did a quick calculation of the impedance of the 5mh inductor. 400Hz = 12.564 ohm 1000Hz = 31.41 ohm 5000Hz = 157.05 ohm Looks to me that the 5mh inductor is mostly affecting frequencies below 1000Hz and I would think removing it from across the squawker will give you a different sound based on a slightly elevated response from 400Hz to around 1000Hz and whatever interaction(constructive/destructive) is created in the woofer to squawker crossover region. When I was changing my AK-2 network to an AK-3 I noticed that Klipsch was increasing the squawker level by abot 2db with the T4 versus the T5 autoformer and also raising/tailoring the frequency levels in the 400Hz crossover region by (reducing the capacitance in parallel with the woofer from a 140mfd to 100mfd) and (the 4mh inductor to a 5mh inductor in parallel with the squawker). So I believe some of the differance in the sound your hearing(good/bad) is also affected by removing the 5mh inductor so I would really be curious what you notice if you reinstall the 5mh inductor and just do the .9mh coil in series with the 13mfd for your bandpass for the squawker. mike Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 21, 2005 Share Posted July 21, 2005 Guys, There is really no substitute for taking acoustic measurements with a calibrated mike right in front of the speaker to know what is going on at the woofer / squawker crossover. The phase between the two has a lot to do with it and the computer says that the changed only oves the phase a small abount. Amplitude changes through the netwrok are in play too. You simply need to do acoustic tests to know. Al K. Quote Link to comment Share on other sites More sharing options...
Frzninvt Posted July 21, 2005 Share Posted July 21, 2005 Dean, please send me the parts to do this and the two 13uf caps as well, tell me what I owe you and I will PayPal it to you. Thanks! Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 21, 2005 Share Posted July 21, 2005 Guys, Here's a computer comparison of the woofer / squawker acoustic summation of the AK-3 network with and withOUT Dean's mod. The simulation is VERY "ify" because it requires that you give it the acoustic path length difference between the squawker and woofer. I can only guess at that number. I assumed 30 Inches or 2.2 mSec difference. The program sums the computed amplitude and phase data from two channels of the crossover and adds the phase shift caused by the path length difference. The response changes drastically with the path length difference. The program is not a polished and finished program, it just a crude plot. The scale factors are at the top. The polts are similar enough to assume the mod will be ok at the woofer / squawker crossover. The squawker / tweeter crossover is not worth considering because of the wavelengths at the 6000 Hz. It will depend on where you are sitting in the room! The only way to know how the mod sums up acoustically is to actually measure it! Al K. Quote Link to comment Share on other sites More sharing options...
BEC Posted July 21, 2005 Share Posted July 21, 2005 ---------------- On 7/21/2005 5:26:44 AM Al Klappenberger wrote: Guys, There is really no substitute for taking acoustic measurements with a calibrated mike right in front of the speaker to know what is going on at the woofer / squawker crossover. The phase between the two has a lot to do with it and the computer says that the changed only oves the phase a small abount. Amplitude changes through the netwrok are in play too. You simply need to do acoustic tests to know. Al K. ---------------- I would agree with Al in that this needs careful acoustic testing. I would add that when you change the relationship between the drivers, you really need an anechoic chanber and the proper equipment to see how the drivers "blend" producing the audio spectrum. Bob Crites Quote Link to comment Share on other sites More sharing options...
Deang Posted July 21, 2005 Author Share Posted July 21, 2005 I would agree with Al in that this needs careful acoustic testing. I would add that when you change the relationship between the drivers, you really need an anechoic chamber and the proper equipment to see how the drivers blend producing the audio spectrum. I agree too, and it's the next step. However, without an anechoic chamber, "careful" acoustic testing is impossible. Mike mentions the possibility of a "slightly elevated response from 400Hz to around 1000Hz", and I think his numbers show that what he is saying might be true. The best I can do may not be good enough (RTA in my room), but I'll try it and see what it says. I'll do the measurements with and without the 5mH inductor in place. I don't think the squawker "blends" very well using any of the autoformered Klipsch networks. It sounds good, but you can tell it's a little rough. Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted July 21, 2005 Share Posted July 21, 2005 Guys, There is an alternitive to an anechoic chamber for testing speakers. If you generate a fast loud spike, it can be analyzed by an FFT analyzer. You simply shut off the mike after the time it takes for the spike to get to the mike. Any reflected energy from the walls is rejected. In truth though, I did this once using my SD375 analyzer and a pulse generator and found I got exactly the same results as using continuous white noise! I never bothered with the spike procedure again! It's pain in the butt! Al K. 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.