A 1" compression driver that can play low / mate with a 15" ?

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

The crossovers are assembled and installed in the speakers. I was able to listen to each one in mono to check that everything is wired correctly and they are certainly a huge step up from the old setup. Soon I'm going to set them up in stereo and have a good listening session - probably from 30 yards away!

Thanks to the folks who chimed in and helped me make some decisions! If I get a chance to measure them before it gets too cold, I will post the results.

Regards,
Dan

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Well, the project is coming along nicely. Got my parts in and etched the boards. As you can see, I'm reusing the original connectors. I must have gotten something right because everything is lining up with no problem!

If only every project could go that smoothly...

[badman]

Re: A 1" compression driver that can play low / mate with a 15" ?

Good call dropping the overall HF if the additional complexity of another notch is a nonstarter. Not thrilled with the 20k peak but it'll be fine- many listeners may just call it "detailed" or "airy"

Looks really solid for the retrofit.

[Cuppa Joe]

Re: A 1" compression driver that can play low / mate with a 15" ?

You may now kiss the speaker....

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Well, I'd like to order the parts for this project soon so they can ship on Monday.

So if anybody knows any reason why these two transducers should not be wed in holy crossover, speak now or forever hold your peace. :D

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Thanks for having a look, guys.

I take it you're worried about the 5 ohm at 20k (and falling for a short while above)? A small series inductor on the tweeter might be helpful, both to suppress the 20k peak and to boost that HF impedance slightly, some amps don't like low impedance at HF because it can mess with feedback or modulation loops if the design isn't solid. (the inductor will require some re-tuning of the other components, most likely) I also might do some additional shaping to suppress that 6k bump- I suspect this design will sound pretty hot if it's left in.

I agree with you about the 6kHz hump being a potential hot spot. The problem I'm having now is that I've all but run out of PCB space for the crossover - I'm already having to populate the rear of the board with the inductors. Since this is a retrofit job, I'd like to stick with the original terminal cup and PCB size and just etch new 4"x6" boards for the speakers.

What I've done with this next revision is tweaked the Lpad resistor values a little to bring the whole HF response down about 1~1.5dB. I know this isn't much, but I'm trying to bring the peak more inline with the response at lower frequencies. I know psychoacoustic-wise, we are more sensitive to an abundance of one frequency than the absence of it. I think its better to have a slight hole at 4kHz.

As for the 20kHz peak, I'm not too concerned about it. I don't know how many people can hear that high, and the peak diminishes off axis. Again - PCB space is a limitation!

Nice job working with what you've got there.

Thanks!

Looks OK to me. 5 ohms at 20kHz isn't a concern, but you do want to be sure it doesn't drop much lower above that.

I've changed the topology for the tank circuit in a way that I am pleased with. I have replaced the parallel RC with the series L with a series arrangement with the shunt C. This achieves a deeper null and eliminates the need for a series R which allows it to fit on the board. More importantly, the impedance is much better controlled and is 8ohm at 20kHz and flattening. This looks to be one of those rare win-win engineering decisions!

Let me know what you think!

Thanks guys for the feedback.

[billfitzmaurice]

Re: A 1" compression driver that can play low / mate with a 15" ?

Looks OK to me. 5 ohms at 20kHz isn't a concern, but you do want to be sure it doesn't drop much lower above that.

[badman]

Re: A 1" compression driver that can play low / mate with a 15" ?

I take it you're worried about the 5 ohm at 20k (and falling for a short while above)? A small series inductor on the tweeter might be helpful, both to suppress the 20k peak and to boost that HF impedance slightly, some amps don't like low impedance at HF because it can mess with feedback or modulation loops if the design isn't solid. (the inductor will require some re-tuning of the other components, most likely) I also might do some additional shaping to suppress that 6k bump- I suspect this design will sound pretty hot if it's left in.

Nice job working with what you've got there.

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

BMS 4550 or 4552ND.

Bill, I have utilized the 4550. I would love to get some feedback on my design!

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Here is the circuit as it stands right now. HF is 3rd order electrical; I could basically meet the target curve with a 2nd order section, but the phase tracking at crossover is not as good. So the extra cap achieves some shaping and adds some beneficial phase shift. C5 bypasses the upper leg of the L-pad at high frequencies to lift the response and compensate for the CD horn.

LF section uses C4 to form a trap that attenuates the woofer breakup. R3 limits the depth, but is there because the impedance of the LF section plunges at HF without it.

So here is a question, what is a good compromise between notching out the breakup and low system impedance at HF? How important is HF impedance anyway? I know that there is much less musical energy up there. Am I barking up the wrong tree here?

Looking for input positive or negative here so I can move on to implementation with some confidence. Thanks!

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

And here is a glimpse of the response with a crossover that I am working on. With fancy inverted polarity trace!

Comments welcome!

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

And here is the HF contour plot.

Normalized to 0 deg!

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Here is the woofer and compression driver response at 0, 15, and 30 deg.

[Eigenmind]

Re: A 1" compression driver that can play low / mate with a 15" ?

Well I bought the BMS drivers and finally got around to testing them this weekend. I attached the on axis response of the drivers in the cabinet. Ignore the absolute SPL numbers. I will post some directivity plots shortly.

Note:

You can't just arbitrarily load the CD lower than the design. The crossover will be wrong, the horn won't load that low, and you'll wind up with worse results by "upgrading". Now, if you're willing to do a full-on upgrade, that's a different story, but it doesn't sound like you're prepared to make the sort of wholesale changes that a compression driver upgrade would necessitate.

I am designing a bespoke crossover for this system. I'm thinking about crossing over at 1kHz. I don't have flare rate data for the horn or anything, but the mouth is roughly 14.5" x 4" and I'm fairly confident that the cutoff frequency is somewhere around 800Hz-1kHz based on similar horns I have looked at.

[badman]

Re: A 1" compression driver that can play low / mate with a 15" ?

Note:

You can't just arbitrarily load the CD lower than the design. The crossover will be wrong, the horn won't load that low, and you'll wind up with worse results by "upgrading". Now, if you're willing to do a full-on upgrade, that's a different story, but it doesn't sound like you're prepared to make the sort of wholesale changes that a compression driver upgrade would necessitate.

GIVEN THAT.

JBL 2426 can be had used pretty inexpensively, is screw-on, and very capable of high output. Only problem is a (well damped) top octave, which can either lead to too little top or heavy EQ and harsh top octave sound. The BMS 4550 looks like a really nice option. Buy nice or buy twice, $150/unit isn't much for a unit you'll hopefully use for years.