Khorn crossover capacitors + harshness problem

[idahodewitt]

Nevermind some of the questions in the last post. I did go to the ALK website and took a look. I'll check their stuff out a little further and repost if I have any remaining questions. I appreciate the lead.

Thanks, L.D.

[mark1101]

Judging by your last post............I am betting it is a smear issue. I can tell you that is a problem to an extent with all speakers using multiple drivers and that are not time aligned. Unless you use an active crossover with DSP and multiple amps, you won't be able to time align effectively. The trick is to eliminate the driver overlap using passives.........and that works very well.

Check out ALK engineering website. A "budget" way to solve the issue at hand (mid and upper frequency overlap) is to look at the AP12-350 networks. You will also need a pair of es5800s. This combo is the best bang for the buck..........It is a click below the full ESNs but because it has ES between the squawker and tweeter it is a world of improvement regarding smear. It kills all but between the woofer and squawker which is hardly a complaint of anyone.

If you want to spend more money for the better networks then go for the ES400/ES5800 combo. I had these on my Khorns, Jub clones, and run them now on my cinema stack (avatar speakers). I have spent LOTS on ALK network designs. I can't say enough about how good they are. You hardly ever see a used pair of ANY of his networks for sale. Truly a rare event. I say that because I saw a pair for sale on this forum today and you hardly ever see that.

I recommend reading his paper on the download page explaining the rationale for his ESN designs. The ESNs are not for everyone. They cost a lot and have a lot of components...............but they do the job of eliminating the overlap. The clarity as you increase the volume is incredible. Just keeps getting clearer and smoother as it gets louder. They work best at the medium to louder volumes.

http://alkeng.com/

[Deang]

Well, if you ask Al straight up what makes a bigger positive difference -- changing the horn or buying a pair of his Extreme Slope Networks -- he will tell you to buy the horns, and I agree.

The P-trap will also attenuate the mid-section a bit, which will help take the edge off when you crank it. I suggest you rebuild the networks, add the P-trap, and then re-evaluate.

[bluesboy]

I agree with Dean and Al on this one. The biggest improvment in sound quality that I could hear came with the switch from 401s and K-55s to Altec 511s and 902 drivers. I understand though that this is not an easy swap to make on stock K-Horns.

[mark1101]

I won't disagree because Al has told me the same thing......that his trachorn makes the biggest difference.

However, let's once again review the problem at hand..........and I am still not absolutely sure that it is smear, but I think it is because I always associate smear as a cause of poor performance in Heritage speakers when someone mentions "it gets harsh when I turn it loud". That is classic Klipsch smear. An out of time alignment system spewing the same music out of multiple drivers milliseconds apart at a high volume.

Al's ESNs definitely remove the smear issue. I am not sure WHAT the trachorn does. It can't fix time alignment issues or remove them. I know from everyone who has trachorns that they must sound darn good...........I just have no way of recommending them as a solution to the problem at hand.

ESN are guaranteed to fix the issue if it is a smear problem.

The biggest problem though is that we are running up a heck of a repair bill either way. I still say though that the AP12-350 / ES5800 is a bargain to fix this issue........and again I just have not heard trachorns nor do I understand technically how they would address the issue at hand.

[idahodewitt]

Now you tell me............ I had sold a pair of 511s not too long ago, but they had 806A drivers.

Now I have plenty of suggestions to consider and I'm not sure exactly what would be pleasing to my ears. I just know that the cornerhorns are not balanced as the volume increases and become unlistenable.

Basically, it comes down to the final dollar figure. I'll calculate the cost of building Al's network on my own, then that will help me come to a conclusion on which way to go.

I appreciate everyone's input. Thank you for the help.

L.D.

[mark1101]

I had Altec 511B horns and used them with K55s tro replace the stock Klipsch 401 horns. They had a nicer dispersion pattern and didn't beam like the 401 but I sold them. An incremental improvement. But they did absolutely nothing for smear. That is why I have no idea how a horn fixes a time alignment issue. Horn geometry and length may alter the time alignment but it will still be off the mark by some amount changing the problem........but not fixing it.

[Marvel]

That is why I have no idea how a horn fixes a time alignment issue. Horn geometry and length may alter the time alignment but it will still be off the mark by some amount changing the problem........but not fixing it.

A horn can't fix the time alignment problem. For a speaker with direct radiators, the acoustic centers of each driver is pretty much in the same verticle plane. Straighten out the horn in a Khorn bass bin and it would be far behind the mid and tweeter driver. Move those two back and they will start to line up.

I put an Eminence APT150 on top of my LS, but moved it back to where it was in line with the K55 sitting underneath it. The difference was astounding. Of course, it did nothing for the delay between the bass and mid. []

Bruce

Mark - I'm still making my drawings for the MWM bins - Colter sent me some other measurements and pics.

[HarryO]

Larry,

Just one upgrade at a time. Listen, then decide if you need any changes.

Harry

[idahodewitt]

Mark,

Have you tried or know anyone who has tried Al's 'improved ' crossover for the khorn? Just wondered if it was worth it's cost. Funds are limited at the moment, but I don't want a so-so/incremental improvement, so if I have to save and wait till I have enough, then that's what I'll do.

L.D.

[mark1101]

Check out Fenderbender. He has my old Khorns and ESNs. They were excellent at my house.....loved them a lot..........see what he has to say about them.

[Islander]

What procedure is used to get the driver timing aligned?

You can relocate the mid and high drivers to align physically with the woofer's acoustic center, but then you have the issue of sound reflecting off the expanse of cabinet in front of those drivers.

The other way to go is with bi-amping and active crossovers that allow you to dial in the necessary delay. That's the arrangement used with the Jubilee and JubScala and it makes an audible difference. It's pricey and complicated to set up, but then you just turn on your system and enjoy.

[idahodewitt]

Harry,

I bought Bob Crites CT125 tweeters & 500uh inductor to crossover the mid/tweet @ 4500hz, and a pair of 13uf & 4uf Sonicaps from Sonic Craft. I could not afford anything else at the moment. If that doesn't help, then I'm probably not going to be happy with the speakers. The only thing I'd still like to do that's currently out of financial reach at the moment is to install the Trachorn in place of the stock mid horn.

L.D.

[daleh]

Hello all.

I’m new to the forum and I picked up some ‘82 Khorns last summer. I’ve been pretty happy with the sound except for a little bump around 5 kHz and a room mode around 125 Hz. These have the AA networks in them and I re-caped them early on for and got a little more clarity in the highs. I put some makeshift bass traps in the room which is way too small at 10’6x16’ with the speakers on the short wall. Unfortunately the long wall is impossible due to layout. The bass bump is manageable though by moving the listening position out of the standing wave.

So the reason I’m posting here is the stock crossover had the tweeter rolling off under 4 kHz instead of 6 kHz. I determined this by measuring the on axis response with the squawker disconnected. So I plugged the parameters into a crossover calculator program and came up with about 2.7uF>130uH>8.1uF for the high pass section. My coil measured 270uH so I played around with a modeling program and adjusted the cap values to give me the 6 kHz crossover point I figured I needed. I swapped the new cap values into the circuit and the crossover point measured perfect. What I failed to notice was the tweeter level was now 10db to low, on the model and the speaker. There was no way to model the crossover a 6 kHz with that coil without lowering the overall output of the tweeter.

I had some cheap magnet wire from Radioshack laying around so I wound some 130uH coils and used the calculator values and everything is good now. So the obvious question is are my speakers unique in this problem or is it well known by all you Khorn veterans. Maybe the crossover was designed for 16 ohm tweeters. Mine measured 6.5 ohms DCR so I would guess they are 8 ohms nominal. Or, maybe I’m missing something and all that geeky tweaking was a waste of time.

So far they sound cleaner and better balanced in the highs to me.

[DrWho]

The electric crossover is point often times very different than the acoustical crossover point. Also, removing a driver will change the xover point too. At the very least you should throw in a terminating resistor.

More details on your test setup might help explain things a bit better. Were you measuring acoustic output with a microphone, or measuring at the positive lead on your tweeter, or both?

Btw, changing cap values shouldn't change the overall amplitude of a driver.

[PrestonTom]

Welcome to the Forum.

I had 1982 K-Horns also. That is a funny year since there were some changes in the drivers within the year, IOW, not all 1982s were the same.

It would be helpful to know which version the K-77 tweeter you have (eg. the "M") and also which mid range driver (the K-55). If it is the "V" (it probably is) were the original wire connections soldered on? Also, about the crossovers (probably the "AA") are they original or have they been modified. Does it look like the connections on the autoformer are original or have they ben altered

-Tom

[daleh]

The measurements were done with a microphone on axis at about 4’. The 55s are V’s with the soldered connections, not sure about the tweeters, maybe I’ll drag them out and look later. They have square magnets though. The DCR of the mids was 11.5 ohms and the tweets were 6.5 ohms. I didn’t measure ith actual impedances of the drivers because I was too lazy.[] The AA crossovers were bone stock when I got them and in good shape. I re-capped them when I got them with some type of metal film or foil caps I had. I matched them to factory specs with a Sencore LC101.

The reduced output of the tweeters was discovered while measuring with the microphone and also showed up in the modeling software. Interestingly the response curve of the tweeter looked really good except for the reduced output. I didn’t use a dummy load for the measurements and that could affect the measurements some. There must be some reason for the lower output with the bigger coil and higher crossover point but I’m not sure what it is. The response curve of the entire system was much flatter throughout the area of the 6 kHz crossover point after the modification.

I noticed the peak around the upper crossover point when I first set them up but I just left them like that until now. The peak was not really annoying but it sounds a lot smoother now. There was also a pretty severe impedance dip at 5 kHz that I saw with my scope and my little amp (SET45). I could only get 100db measured output at the crossover point before there was distortion on the scope. I swaped the output transformers from 3k:8 to 5k:8 since then (my last speakers were 16 ohm Altec VOT). I’ll have to measure again and see if this impedance dip has been reduced.

[Coytee]

I had sold a pair of 511s not too long ago, but they had 806A drivers.

I'm late to this chat but if it helps any, I seem to recall Al saying that the HORN swap was the important part... not the driver/horn. You could use the K55 on his trachorn and still gain the benefits of the swap.

[daleh]

Here’s a shot of the measurement I took. There is still a lot of interaction with the area just under the 6 kHz crossover point but no where near the stock configuration. I’m thinking of trying a 4^th order high pass to see if I can get it a little cleaner.

I don’t see a model# on the tweeter but it has a square brownish magnet.

The test rig is DBX measurement mic> M-Audio 610>TruRta.

[DrWho]

Hmmm, that's not rolling off very fast at all. I think going steeper would be a definite improvement. You can actually see the comb-filtering from the time-alignment (or rather lack thereof).

Btw, if you measure the impedance response, you can see at what frequency the cone-breakup happens with your squawker...and then try to avoid that with the steeper xover you implement. It's not showing up right now, but it might show up in a waterfall plot.

And not that truRTA is a bad tool, you might also check out REW:
http://www.hometheatershack.com/roomeq/
This program is free and also has an RTA with noise generation, but it also has a swept frequency measurement that calculates the impulse response....which then lets you gate the impulse so that you can measure the frequency response before the influence of your room (basically it allows a psuedo anechoic measurement). You lose frequency resolution when gating, so you lose low frequency extension - but it won't be a problem for the bandwidth you're measuring on this project. It will also provide you with ETCs, RT60's, and Waterfalls. Just thought I'd throw it out there as it seems like you know what you're doing with your measurement rig. [Y]