Pano, did you really mean that you are crossing the woofer at 90hz LP, 6db/Oct? That's going to produce a lot of overlap with the compression driver. Even if it's measuring flat response, it doesn't seem like a good idea. Try a faster slope up higher and I think your midrange congestion might go away.
Hi Pano, Very cool stuff and I think it is very hard to tune.
More long wings is not a good idea because
You will have quarter wave resonance.
I noticed you can have store energy between the dipole dips : it ruins the midrange. But directivity might help to avoid a too perfect dipole behaviour of the speaker
. If you can trace the CSD, it can help.
Following your progress, very interesting and unusual.
Have fun !
More long wings is not a good idea because
You will have quarter wave resonance.
I noticed you can have store energy between the dipole dips : it ruins the midrange. But directivity might help to avoid a too perfect dipole behaviour of the speaker
. If you can trace the CSD, it can help.Following your progress, very interesting and unusual.
Have fun !
Yes. That is done to correct the rising response of the woofer on OB. John and I built a lot of OBs using a big inductor in series with the woofer. Usually 15-22mH. Thats a very low frequency 1st order LP. It works like a charm. Active is not quite as good as the big inductor, but it’s fast to implement.
In simulation an 80-90Hz LP on the woofer flattens its response and it then has an approximately 2nd order roll off at 420 Hz. Slope Q about 0.9. If the compression driver is high passed circa 800-850 it should sum correctly near 750Hz.
Of course real life is messier than that, so it will take some tweaking. The passive crossover we had in 2008 was better than my attempt yesterday, but I got close.
En effet, oui. It is proving tricky in the midrange. The difficulty makes me admire John B's crossover skills even more. If only I could remember the crossover we built way back then. 👁️ Changing the horn crossover point - I have run it up and down - swapping polarity and a few other standard tricks make only small audible differences, never major. And I don't know what that is telling me. Is it such a mess that a few changes don't really fix it, or is it a fault tolerant design?
True about the wings, but I would never go too deep. No more than 8"- 20cm. One could go deeper on a single side if that wing were tapered or triangular, which would help spread around the baffle peak.
One thing I liked about these when I first heard them (and now) is the scale. They are big, and they sound big. They can make singers seem slightly oversized, and wind or string ensembles seem a touch heavy, but they are never small or tiny or boxy. Rock really rocks. Symphonic recordings sound like you are sitting about 4 rows back in the balcony in the $$$ seats.
Very cool looking drivers. I really like that brown cones, smooth in front, rough at back.
For XO, have you tried 1st order at 2kHz for tweeter. That should smooth the 1-2kHz bump and give you similar crossing point.
Not yet, although looking at simulations yesterday that is one solution that stood out. I have programed that as a possible crossover and will try it as soon as I can.
Here is what I mean by the LP at 420 Hz explained in a graph.
Above we see in red the woofer response on OB taken at 2 meters from the baffle. The green line is the response of the woofer after the application of a 1st order low pass filter @80Hz and some PEQ to tame the peak around 1550 Hz. In blue is the transfer function of a 2nd order low pass at 420 Hz. The filter has a Q of 0.9, which is a steeper shoulder than Butterworth. The rolloff of the woofer now looks very much like the 2nd order Q=0,9 The baffle peak circa 450Hz has now become the shoulder of the LP rolloff.
Of course this is both driver and baffle dependent, and perhaps influenced by the room.
Here is that 2 second order filters would look like summed. The crossover point has a little ripple, but it's not horrible.
Note that the crossover happens at about 600Hz, while the LP is @420 Hz and the HP is 830 Hz. Spreading them apart makes them sum fairly flat.
I don't think that the compression driver is going to be molded into that pretty high pass shape, but it gives me an idea of what "could be."
Note that the crossover happens at about 600Hz, while the LP is @420 Hz and the HP is 830 Hz. Spreading them apart makes them sum fairly flat.
I don't think that the compression driver is going to be molded into that pretty high pass shape, but it gives me an idea of what "could be."
Looks like you need more EQing and a real crossover, not just a 6dB/oct. Is that with or without the 90hz LP in place? In either case, that much 400hz never sounded good to anyone.
I was wondering about the thin baffle you've got too. It's got to be resonating and adding something to the sound being such a large floppy thing only 1/4" thick.
I was wondering about the thin baffle you've got too. It's got to be resonating and adding something to the sound being such a large floppy thing only 1/4" thick.
No worries. Once low passed, the woofer has no more 400Hz than anything else in the pass band. Getting the horn to match is the difficult part.
Having built and heard many, many OB speakers with just a big inductor on the woofer, I know how well it can work. Most people don’t understand OB crossovers, which is why so many of them sound bad.
But it took me an entire year to understand OB crossovers - a year of making bad designs, because I was so used to “normal” crossovers. Now I’ve spent years on this forum explaining how OB bass works because most people aren’t familiar with it. In fact I have an entire thread about it, with well over a million views. The aggressive 1st order LP on the OB woofer works extremely well. It worked 15 years ago on these drivers.
The baffle is 1/2”, and I agree that may be part of what’s muddying the sound. Bracing might help some. I wonder if that can be measured?
Having built and heard many, many OB speakers with just a big inductor on the woofer, I know how well it can work. Most people don’t understand OB crossovers, which is why so many of them sound bad.
But it took me an entire year to understand OB crossovers - a year of making bad designs, because I was so used to “normal” crossovers. Now I’ve spent years on this forum explaining how OB bass works because most people aren’t familiar with it. In fact I have an entire thread about it, with well over a million views. The aggressive 1st order LP on the OB woofer works extremely well. It worked 15 years ago on these drivers.
The baffle is 1/2”, and I agree that may be part of what’s muddying the sound. Bracing might help some. I wonder if that can be measured?