Dave MacKay
Active Member
My Heresy speakers (now SHs) have K-52H squawkers. Are there known issues (like we're seeing with the K55M) with that driver?
Super Heresy 2.0 – dip at 4–5 kHz with SMAHL V2 / DE120?
- Thread starter dinkeyesjc
- Start date
Dave MacKay
Active Member
Are there any recommendations for replacements for the squawkers on Heresy (or Super Heresy) speakers using stock (K700/K701) horns, apart from other K55 variants?
I'm wondering if something exists for the squawkers like we have the DE120/SMAHL for the tweeters.
I'm wondering if something exists for the squawkers like we have the DE120/SMAHL for the tweeters.
mboxler
Well-Known Member
I have the K52H as well. It does well up to 6khz. Here's how it would look with the DE120 on a Type E crossover.
To maintain a good 6khz crossover, you can add two parts to the Type E tweeter filter. After the 2uf tweeter cap, place a 160uh inductor to ground, then add another 2uf series capacitor. This is similar to the AA, substituting a 160uh for the 245uh.
I'd tame that 10khz bump on the K52H as well, but it may sound okay with the tweeter mod alone.
Dave MacKay
Active Member
@mboxler The second plot certainly looks a lot better than the first one.
Just to be clear, in the first plot is the white line the K52H and the green line the DE120? And are the colours reversed (white = DE120, green = K52H) in the second plot?
If my records are correct, the tweeter portion of both the E and E2 networks are the same, so that the changes you suggest ought to apply equally well to either network.
It's hard to read the frequency axis on the plots -- do the changes alter the 6 kHz squawker/tweeter crossover?
I've modified an E2 schematic that I retrieved from the Klipsch forum to show the changes you suggested. I think the schematic has the polarity wrong for the tweeter. The values of the two new components are shown in red.
Please correct me if I've misunderstood the changes you designed.
What would you suggest to eliminate the 10 kHz bump in the midrange, if that proved necessary/desirable?
Last edited:
mboxler
Well-Known Member
Yes, I accidently reversed the colors. Sorry. The new schematic is correct, although I can't test the polarity without gutting my two-way Heresy's.
If you don't have those parts on hand, but have something close, I can simulate that to see if they are okay. As is the crossover is around 6khz. This is a simulation, so if you have a way of measuring that would be great. I'm always looking for verification!
I'll work on the squawker later.
Are you done yet?
mboxler
Well-Known Member
Picked up a nasty bug and yesterday was brutal.
Here's what I came up with.
And here is the before (green) and after (white) simulation
Dave MacKay
Active Member
@mboxler Mike, thanks for designing the filters. The squawker sure is a hard-to-tame beast.
I was expecting that I'd see a more consistent downward slope after the crossover frequency (6 kHz). There is a nice (and steeper than expected) drop-off but then -- even in the white plot --there's still a peak around 10 kHz. The peak has been lowered, but it's still fairly significant.
The additional filter on the midrange looks (to me) to be 1st order. Is it?
If the filter is 1st order, it should be dropping off at -6 dB/octave past the crossover. Because it's hard to read the vertical and horizontal scales on the graph, I can't tell if that's what it's doing. Is it?
mboxler
Well-Known Member
Now we are getting into how crossovers actually work. For simplicity, I'll use a K33E in a La Scala for an example.
The white plot is the SPL with no filter.
The green plot is a true first order low pass filter with a corner frequency of 400hz, meaning that the voltage is down 3dB at 400hz. Think active crossover.
If I used the active crossover, the SPL would be the red plot, and the SPL would be down 3dB at 400hz compared to the raw plot.
If I use a 2.5mh low pass inductor, the SPL would be the blue plot. Notice that the SPL difference is barely changes due to the rising impedance of the K33E.
My point is that a first order passive filter rarely equates to a 6dB slope.
Does this help at all?
Dave MacKay
Active Member
Thank you. That’s very informative.
I use an active crossover (MiniDSP Flex Eight) with my La Scalas. I particularly enjoy the ease with which it allows me to try out different configurations.
I am still using recapped E2 networks with my SHs.
