Help with hunting the 9KHz K-55-V artifact

[codhead]

Is anyone familiar with the exact frequency this artifact appears at?

I've been driving myself half crazy trying to reproduce it on my

Heresys and Cornwalls. Maybe I'm just just lucky enough not to have

drivers which suffer from this anomaly, but I'm not convinced.

Would I benefit from mic'ing my squawkers, and looking at the signal amplitude on a scope?

[JohnA]

Try this.

Atlas PD5-VH-VT.PDF

[Seadog]

You may be fortunate not to have the subset of K55-V drivers afflicted with the 9KHz flare problem. I was not convinced myself until I built P-traps and installed them in my CWs. In my case there was a clarification in detail. I never thought my drivers had the problem because it was not extreme, but the P-traps helped.

I have one extra P-trap (I built an extra one just in case), and if you want to borrow it to do the A/B experiment, send me a PM.

[Deang]

!

[sfogg]

This is a spring clip K55v...

http://forums.klipsch.com/forums/276844/PostAttachment.aspx

Shawn

[codhead]

John,

From the .PDF, it looks like there's a peak at 5250, and 8750. Wonder why no one makes an issue of the 5K peak?

[Deang]

The Atlas .PDF file shows the plane wave tube response. The response on the K-400 is different. Note the plot I posted and the one Shawn posted are nearly identical. The "artifact" is actually a spike in the response that takes the form of a "bounce back" into the tweeter's acoustic response. A bandpass or P-trap drives it down -- and it does clean up the top a bit.

[codhead]

Thanks for your responses on this. I'm thinking about upgrading my

Cornwall crossovers, but I'm a bit leery about adding additional parts

if I don't need 'em (need all the efficiency I can get with SET amps).

[JohnA]

The Atlas graph IS on a plane wave tube (long tube, maybe 4 meters) but

it shows the peaks. The response is different on the K-xxx horns,

but generally at the HF because the horn's vertical dispersion

collapses to boost HF response. The K-55-V needs that 5k peak to

reach 6k-ish. On a horn it doesn't look like much of a

peak.

[Corvette6769]

Okay, what am I missing here......if the crossover point is 6000 Hz, isn't everything in the 6kHz to 20kHz frequency range sent to the tweeter and as such how is there a 9kHz problem with the K-55-V midrange driver?

[chops]

Okay, what am I missing here......if the crossover point is 6000 Hz, isn't everything in the 6kHz to 20kHz frequency range sent to the tweeter and as such how is there a 9kHz problem with the K-55-V midrange driver?

No because it looks like it's a harmonic of some sort.

Dean, is this one of the reasons why you suggest dropping the CW's mid-tweet crossover point down to 4.5KHz when using Bob's CT125's? (which I now have)

Thanks,

Charles T

[colterphoto1]

Okay, what am I missing here......if the crossover point is 6000 Hz, isn't everything in the 6kHz to 20kHz frequency range sent to the tweeter and as such how is there a 9kHz problem with the K-55-V midrange driver?

From my understanding, here's how this becomes a problem.

If the crossover 'point' is 6000, depending on the slope of the crossover, the mid is still putting out signal at 9000. With a 12 db per octave slope, the signal going to the K55 would theoretically be down 12 db at 12000 hz. It's a log scale, so tough to guess, but with 9k being somewhere midway, it could potentially be at -6db or so, this spike coupled to the K77's strong output in this range could lead to on overall imbalance around this frequency.

I can't say that I've 'heard' this and none of my K55's are of the valued spring terminal type. I guess I'll look into doing the 'p' traps when I redo my networks. My AA/A's I got from BEC don't have this built in, perhaps I should have asked for this? (they replaced AL's in my LS)

Michael

[sfogg]

"......if the crossover point is 6000 Hz, isn't everything in the 6kHz to 20kHz frequency range sent to the tweeter and as such how is there a 9kHz problem with the K-55-V midrange driver?"

Not on the factory crossovers. They don't low pass the squawker... it receives everything from 400hz and up on the Type A or AA crossover.

That is why some crossovers like the ALK bandpass the squawker. They roll off the driver which tames that peak.

See this thread:

http://forums.klipsch.com/forums/thread/276852.aspx

Which has measurements of the K55V run full range vs. bandpassed.

Shawn

[colterphoto1]

ahh, very good point- I forgot about that!

[Deang]

"...is this one of the reasons why you suggest dropping the CW's mid-tweet crossover point down to 4.5KHz when using Bob's CT125's?..."

I don't think I ever suggested doing that with the Cornwall (or Heresy), in fact, I think I even recently posted that with those two speakers it might not be the best way to go. Ideally, one wants at least three octaves between crossover points on a first order network. At this point don't ask me why, it's been so long since I read it I can't remember the exact reasons for it. I have a bad habit of moving on to other things and then when I need to remember the why of something I only end being able to remember the what. At any rate, it's something to try, and if it sounds good then you don't have to worry about "the rule" so much:) -- and yes, the amount of rejection at 9kHz is higher when you move the crossover point down on the low pass section of the squawker. Truthfully though, if you go with a 6kHz squawker low pass/ tweeter high pass --.40mH litz air core in series with the squawker (from tap 3 to midrange positive on the terminal strip), and replace the 2uF tweeter cap with a 3.3uF cap -- you'll get improvement on the same scale as Bob's little mod, and I think if you compared the two ways of doing it you might find it impossible to decide which is "better".

[3dzapper]

I still think it easier to put a 3uF capacitor across a 0.1mH 18Ga inductor in the positive lead to the squaker. Might cost $5.

A P-trap can be seen in the lower Type B in the image:

Rick

[3dzapper]

The Atlas graph IS on a plane wave tube (long tube, maybe 4 meters) but it shows the peaks. The response is different on the K-xxx horns, but generally at the HF because the horn's vertical dispersion collapses to boost HF response. The K-55-V needs that 5k peak to reach 6k-ish. On a horn it doesn't look like much of a peak.

John, I was under the impression that the constriction in the inlet of the K-xxx horns was responsible for bringing up the response of the K-55 above 4500Hz. You are more knowledgable than I, have I been misguided?

Rick

[Deang]

I don't think it's behavior is isolated to the small throat horns -- look at the plot on Al's site of the Trachorn response compared to the 511B and K-401.

[WMcD]

Here is the Dope from Hope on the change in K-55-V phase plug. It was posted by some good person not too long ago.

My read is that that both versions have the artifact, it has just gone up a bit in frequency with the two-piece plug.

I've also read from people posting here that only some of the K-55-Vs have artifact. This DfH does not confirm that.

You can see the artifact is 10 dB down or so from the midband. That is perceived as about half a loud. So I don't doubt that it can be heard.

This DfH was not in the regular collection and I'd love to see a scan with better resolution, if anyone has one.

Gil

[WMcD]

I should have said, the two-piece has an artifact a bit below the one-piece and a new smaller one above.

Gil