That 9kHz "glitch"

[Deang]

I can't find the thread where we discussed this thing. Looking at the plots of the PD-5VH, the only thing I see of concern is at 6kHz. Anything going on at 9kHz just seems too far down to cause any problems.

[BEC]

Dean,

That would be what I have been saying for a very long time. I can't find anything that I consider significant at 9 khz. Of course there was a recent post that said it was only there in about 20 percent of the K-55Vs at some time when someone looked at them. Perhaps that is why I haven't found one yet that has a peak that seems significant.

Bob

[Deang]

The word "bounceback" keeps coming to mind, but I can't remember how it's applied.

[sfogg]

Dean,

The one I measured was about 3dB down at 9kHz compared to 6kHz. That certainly isn't insignificant.

The other thing to keep in mind is amplitude alone doesn't tell the whole story. If the glitch is from a driver resonance/ringing its decay will be longer then it should be. IOW, it will be playing the 9kHz tone longer then it should. The 2d measurements give some info on how a driver/speaker reacts but not as much as 3d measurements.

If I still had a K55 I'd do 3d measurements on it but I no longer have any.

Shawn

[BEC]

Shawn,

I most often see that peak at around 12 to 15 db down. I did see a K-55V last week that had a peak at around 8 to 9 khz that was only down about 5 db but the whole response curve on that one was funny and the DCR was low by about 2 ohms. A diaphragm replacement put that one back to normal. I think that VC had a few shorted turns.

Bob

[TBrennan]

The fella that came up with the P-Trap never claimed all Atlas drivers had the peak or that all should have a P-Trap. SOME of the drivers he tested had the peak, SOME.

The fella thinks that those who have a particularly fierce speaker should try the trap. If you think your speaker sounds fine then leave it alone.

[markus111]

This is one thread with graphs of the 9khz glitch. Apparently only about 25% of the push-terminal type k55's had it. It was fixed with the solder terminal type.

http://forums.klipsch.com/idealbb/view.asp?topicID=40781&forumID=71&catID=19&search=1&searchstring=&sessionID={B1F86BB6-FABE-4DF6-9C61-EA6F71B4823E}

Mark

[Good2BHome]

----------------

On 4/16/2005 10:56:27 AM TBrennan wrote:

The fella that came up with the P-Trap never claimed all Atlas drivers had the peak or that all should have a P-Trap. SOME of the drivers he tested had the peak, SOME.

The fella thinks that those who have a particularly fierce speaker should try the trap. If you think your speaker sounds fine then leave it alone.

----------------

To compliment the above:

I had an extended conversation with the P-Trap inventor (over wine and cheese), within the last year. He said about 20-25% of the "push-terminal" Atlas K-55 drivers exhibited the problem. The problem is not present in the "solder-lug" K-55s. These began shipping after March 24, 1980 per "Dope from Hope" # 19-1.

Mike

[TBrennan]

"I had an extended conversation with the P-Trap inventor (over wine and cheese)"

Just don't go to Big Earl's with him, go to the Chinese joint.

[Deang]

The problem is that the soldered K-55-V with the dual phase plug was only used for a few years. Anyone with a Heritage speaker made before March 24th 1980 has the spring loaded version with the single phase plug. So, 1 in 4 have squawkers exibiting the dreaded "glitch". The cool thing about the P-trap is if you use it on a network that's feeding a K-55-V that doesn't have the problem, it doesn't really do anything, so why not add it for the sake of peace of mind?

Over the last month I've decided that the best solution is simply to improve the old networks. Let's face it, the Type A is a 40 year old design, and the Type AA isn't much newer. They were designed with a "least of what you can do is" mentality, with no modeling or optimization programs. It's ludicrous to be critical of aftermarket designs when Klipsch themselves abandoned these networks decades ago. Hats off to The Man, because I think they still sound very good. I really dig the sound of the simple Type A -- but I like the sound of it better with a bandpass on the squawker (which takes care of "the glitch). As far as the AA goes, I see nothing wrong with changing the tweeter filter to a true Chebyshev or Butterworth alignment if it brings improved performance -- which it should.

[Frzninvt]

That answer's my question, I was going to ask if there was any adverse affects of having it (P-trap) in there and the K-55's not having the glitch. My K-55's are really, really old though (43 years) where I have your Type B w/P's in them ('62 Cornwalls). The Khorn's did not need them I have the soldered terminal K-55V's in one pair and the K-55M's in the others.

I'd send them to Bob for inspection but they sound fine and have no abnormalities so I have just left them alone for now.

[Tom Mobley]

Dean,

>>"Over the last month I've decided that the best solution is simply to improve the old networks. Let's face it, the Type A is a 40 year old design, and the Type AA isn't much newer. They were designed with a "least of what you can do is" mentality, with no modeling or optimization programs. It's ludicrous to be critical of aftermarket designs when Klipsch themselves abandoned these networks decades ago. Hats off to The Man, because I think they still sound very good. I really dig the sound of the simple Type A -- but I like the sound of it better with a bandpass on the squawker (which takes care of "the glitch).<<<

I think some guy named Al already did this? Twice? Works for me.

[Deang]

Those aren't first order networks.

[Dflip]

I have asked this question before, but I will try again, does anyone have a picture or a schematic of what the P-trap would look like for a Cornwall B network, since I have the push-on terminals?

I would prefer the picture so I can see how it all fits together. It would need a 3.0 uF cap and a 1.3 mHy inductor which are soldered together and then placed between the squawker and the crossover?

Don

[Seadog]

The P-trap is the same regardless of the network. It is a 3.0 uF cap and a 0.10 mH inductor in parallel on the positive lead of the squawker.

[Deang]

Here ya go.

[Dflip]

Thank you Dean, your computer drawing skills are getting better. It made it much easier for my technically challenged self to see what goes where.

Don

[nola]

Hello all from the GTG in Hope, Friday night, June 2010.

Q1: How do we / can we test our K55s of the push pin type to see if we have the glitch? (I have 2 KHs, 2 LSs, 2 extra K55s, and about 8 HIs - so 14 potential driver issues). Is testing the squaker / driver combo with a test tone generator and microphone set up to generate a curve needed to check for a spike at about 9KHz? This would be a major pain for me given all of my stuff. Are there some K55s that are specifically known to have this glitch, while others simply don't (of the push pin family)? Most all of my gear would be c. 1976-9, so fo me this is a potential real issue that would be a time consuming pain and cost a bunch to resolve.

Q2: There several different sets of inductor and capacitor values given in various posts. So which set of values is more optimal, and why? Are there different values needed in differing networks (E Heresy I vs AA LaScala)? Any suggestion on which caps and inductors to buy. My post Katrina redos were done with Crites supplied Sonicaps.

I have stock tweeters, but for those with Beyma or Crites tweeter upgrades are P traps needed?

Q3: Should we have some general notice posting on which fix ups are to be considered depending on which speakers and networks are owned by forum users. For example: the K55 9Khz Q; the magnetic screws erroneously used to hold some inductors in place, replacing the gasket red rubber ring between the K55s and the metal horns, checking the foam and black gaskets, replacing old capacitors, etc.

Thanks all - at 1 am.

NOLA

[Seadog]

Hey Nola, it was great seeing you again in Hope last weekend.

I may have told you incorrect values on the P-trap components (after all, it was nearly 1 am Saturday morning and my memory is not so good anyway!).

The correct values are as stated above in this thread: 3.0uF cap, and 0.1mH inductor. That is exactly what was in the P-trap that I show you in my Heresy in the hotel room. Any other values will not be a P-trap (some other trap maybe, but not a P-trap). I am certain because "Mr. P" himself inspected my P-trap and verified that the values were correct. "Mr. P" was the fellow that identified the 9kHz problem and devised the P-trap for the fix (not sure why, but he prefers not to be known by his complete name).

Mr. P told me that not all - but a high percentage - of the push-on terminal style K-55V drivers have this problem to some degree. The P-trap will remove the 9Khz spike if it is there, but will not affect the sound if the spike is not there.

David