This really isn't the place for discussion of current drive and such.
However I can tell you that the large inductance between the woofer and the amp is part of what makes John's OB designs work and sound so good. All his OB designs and builds use a large inductor on the woofer. It's an integral part of the design.
As to why that might be, it's a good topic for discussion in another thread. Just remember that it IS an important part of this design. But let's not pull this too far off topic. Thanks.
However I can tell you that the large inductance between the woofer and the amp is part of what makes John's OB designs work and sound so good. All his OB designs and builds use a large inductor on the woofer. It's an integral part of the design.
As to why that might be, it's a good topic for discussion in another thread. Just remember that it IS an important part of this design. But let's not pull this too far off topic. Thanks.
Agreed, I am not trying the derail the treat. I personally am very fond of current drive. But I use it for mids/tweeters only. Have three versions of F1J with various outputs and about dozen of lm3886 amps with current drive feedback. Not on woofers though. I made lots of simple active crossovers myself, so I am exploring that route. But I am open to exploration, still have some woofers laying around.
I had this thing as well, a more loud midrange than I liked. On my part, I thought my slight narrowing of the baffle made the Vifa louder and thus put in damping material around (not behind) the elements. To me, that smoothened the sound a bit so the mid was quite ok. I used 20mm waffle foam for this.
Same. And I tend to like a good, strong SS amp for the woofer section that is placed close to the woofers as possible. But in the case of the Manzanita design, the big inductor does so much to make the design work and sound correct, that my normal procedures go out the window.
Okay, why wait... Two stripes of felt placed into the cut out of the TC9 (0.5×1 cm), and the effect was immediately here - but I'm not sure if it is the right direction. The "sharpness" is gone, but at the same time there was a loss of some room, stage, liveliness and the treble was also compromitted. For sure, it was a different sound image and I missed something.
So I removed the felt
So I removed the felt
I was curious and took some measurements on my Manzanita:
Frequency response 6 dB smoothing:
Frequency response 24 dB smoothing:
Frequency response 20° off axis 6 dB smoothing:
Impulse response:
Distortion:
I was searching for the reason of the "sharpness" or "screaming", but the diagrams are okay.
However, with the sweep sine signal, in the upper mids there was an audible uncomfortable, almost throbbing area, like an echo. But I don´t see it in the diagrams, do you?
Interesting, for sure
Frequency response 6 dB smoothing:
Frequency response 24 dB smoothing:
Frequency response 20° off axis 6 dB smoothing:
Impulse response:
Distortion:
I was searching for the reason of the "sharpness" or "screaming", but the diagrams are okay.
However, with the sweep sine signal, in the upper mids there was an audible uncomfortable, almost throbbing area, like an echo. But I don´t see it in the diagrams, do you?
Interesting, for sure
Thanks for the plots. Is that taken at the listening position with one speaker playing? The impulse would indicate one speaker.
What I do see is a rise above 3 kHz that might sound bright, but usually harshness comes lower. I would rather see a slowly falling response above ~630 Hz.
Can you export the impulse response? Not gated if you can. Does Omnimic do waterfall plots or other time graphs? REW has very good ones, but AFAIK the measurement has to be done inside REW.
What you might try is to increase the value of R2, the resistor that is in parallel with the small inductor on the Vifa. See on your plots if you can get a falling response in the top end, and listen to what that does.
I suspect it's a resonant peak, but I'm not seeing it right away. Bending down the Vifa response might point you in the right direction.
What I do see is a rise above 3 kHz that might sound bright, but usually harshness comes lower. I would rather see a slowly falling response above ~630 Hz.
Can you export the impulse response? Not gated if you can. Does Omnimic do waterfall plots or other time graphs? REW has very good ones, but AFAIK the measurement has to be done inside REW.
What you might try is to increase the value of R2, the resistor that is in parallel with the small inductor on the Vifa. See on your plots if you can get a falling response in the top end, and listen to what that does.
I suspect it's a resonant peak, but I'm not seeing it right away. Bending down the Vifa response might point you in the right direction.
Hi Pano, thanks for your response.
Yes, at the listening position and one speaker.
There is a bump between 1 and 2 kHz, could it be the harshness?
I think I can't export the data, just make plots, but I must check it. I also think the system can do waterfall, I will repeat the measurement soon.
Thanks for the advice, changing the resistors will take time, I have to order some. Sadly I have no spare parts, I'm only since a couple of months into building speakers, this is just my third one (after the FrugelHorn and the FAST TL).
Today I will try to apply some felt on the back around the TC9's cut out. Will see.
Yes, at the listening position and one speaker.
There is a bump between 1 and 2 kHz, could it be the harshness?
I think I can't export the data, just make plots, but I must check it. I also think the system can do waterfall, I will repeat the measurement soon.
Thanks for the advice, changing the resistors will take time, I have to order some. Sadly I have no spare parts, I'm only since a couple of months into building speakers, this is just my third one
(after the FrugelHorn and the FAST TL).Today I will try to apply some felt on the back around the TC9's cut out. Will see.
I tried the felt and as Arne @nouvolare wrote earlier, it is now - calmer, as it was with felt directly in the cut out, but to a lower degree. The stage don't shrinks, but I miss the drive.Yes it could be that peak, but usually harshness is perceived higher than that, like 3K-5K. If you can get waterfall working we might see a long ringing in the upper mids somewhere.
in the meantime try the felt!
On the other hand I noticed - independently from the felt -, that the "harshness" or "sharpness" is volume dependent and it begins to occur just at the volume level where i usually listen to music (at about 75-79 dB(A) peak).
Out of couriosity, I set up my 3000 euro commercial speakers - and they were pushed aside again after about 7 minutes - it's hifi, but I don't like this kind of smooth and somewhat borig sound anymore
So what next? I will order some resistors, birch ply and will play with the Manzanita, because it is definitely worth it. In the meantime I will have a surgery next week, so no progress to expect
Many thanksBest wishes and best of luck for the surgery! You can enjoy some good music while you heal.
I found something interesting now, @John Busch wrote it back in 2012:
"It seems the Latest Vifa's are a bit hotter than the earlier production runs, so it seems 5.0 ohms for R2 and 15-25 ohms for R1 seem to be a better balance for most listeners."
This is a big difference to the values in the circuit diagram, where R2 is 8 ohms and R1 4.5 ohms. Why this big difference?
"It seems the Latest Vifa's are a bit hotter than the earlier production runs, so it seems 5.0 ohms for R2 and 15-25 ohms for R1 seem to be a better balance for most listeners."
This is a big difference to the values in the circuit diagram, where R2 is 8 ohms and R1 4.5 ohms. Why this big difference?
Because the design keeps evolving. In the latest post of crossover values, that is what John was recommending at the time. That's what I built. But I have also used rathr different values when the room acoustics called for it. OB has so much sound going out of the back and bouncing back that tuning to your room becomes even more important than with box speakers.
Same thing with the wings. Originally wings were asymmetrical, later they were changed to be the same on both sides for ease of building. This is a speaker build that can be done with minimal tools and woodworking skills.
Many thanks Pano. Regarding woofer, as I will order many new parts and baffle material, should I go for the GRS 15"?
John wrote, that it works better in the Manzanita. Or is it just the personal taste?
Which values for R2 could be fine? I know, it's difficult to answer, but... should I go for 12 and 15 ohms? Or higher, or maybe for 10 ohms too? Should I think about changing/increasing R1 too? Or just try and error, because the design is forgiving?
Yesterday, out of curiosity, I listened to my 4 speakers and none can deliver the spatial imaging that the Manzanitas are capable of. That's really amazing.
John wrote, that it works better in the Manzanita. Or is it just the personal taste?
Which values for R2 could be fine? I know, it's difficult to answer, but... should I go for 12 and 15 ohms? Or higher, or maybe for 10 ohms too? Should I think about changing/increasing R1 too? Or just try and error, because the design is forgiving?
Yesterday, out of curiosity, I listened to my 4 speakers and none can deliver the spatial imaging that the Manzanitas are capable of. That's really amazing.
Last edited: