Lastly, a second attempt at the AN10 classic. This uses an additional Parallel filter (so at least not in series with the driver) to preserve the high end. No super tweeter need apply, but I'll not speculate on the treble from a 10" driver until I've heard it for myself. 
This one MIGHT get away without an impedance flattening circuit.
This one MIGHT get away without an impedance flattening circuit.Attachments
It's hard to say since I don't know how much gain the cabinet is providing; having studied MJK's BLH papers, I know that the gain is very much dependent on the Qts of the driver. If the AN8 truly has a Qts of 0.28 or thereabouts, it probably needs the full amount I gave in my simulations above but if the Qts is higher it might need less. If you actually build that filter and it turns out to be too much, I can easily go back in and adjust the values. It might be as easy as reducing the resistor in the BSC circuit but the contour filter might need additional adjustment as well.
Freddi, thanks for posting a measurement of the super 10. Just to see what would happen I copied your graph and it looks as though a simple 2.2mH/8R BSC circuit would go a long way toward balancing the response. Listening off axis would be enough to compensate for the rising high end, and I wager that ~5db peak at 3k is more annoying, so I tried to filter that out instead. With the filter in place less BSC is needed so the resistor drops to 6R. In total, 2mH/6R BSC, 0.15mH/15uF/12R notch filter, 15uF/15R zobel. PCD7 thinks that 5-10 degrees off axis will balance out the top end.
Attachments
Last edited by a moderator:
Without measurement tools, you really can't; I have the same problem. You'll have to do it by ear. If the full BSC make the speaker too bassy, you'll know that either the Qts is higher than advertised or the BLH is operating very effectively (or both!). What sort of horn do you have it in?
EDIT: by the way, I'm not sure if you have caps and coils just lying around, but if you do and you want me to simulate what would happen with values that you already have on hand, let me know. There's actually quite a bit of flexibility with these filters.
Last edited by a moderator:
Well without actual frequency response charts it's hard for me to say. My suggestion might be to try the less extreme filters I offered above, and if those don't suffice you can go for more compensation. Honestly you could get away with the same values of caps and coils for both and just change the resistors, so trying both would be pretty cheap. And by the way, since I don't have any of these drivers myself I can't guarantee that these filters will be perfect. But given my experience simulating crossovers for my 2-way and how accurate that was, assuming the FR and impedance data being given by Audio Nirvana is accurate I'm confident the filters will work.
Last edited by a moderator:
By the way, if you want to see what I'm on about regarding Qts and BLH performance, jump to page 48 of this paper and read through page 52.
http://www.quarter-wave.com/Horns/Back_Horn.pdf
At worst the AN8 falls closer to a Qts of 0.4 in which case you'll need a little less BSC.
http://www.quarter-wave.com/Horns/Back_Horn.pdf
At worst the AN8 falls closer to a Qts of 0.4 in which case you'll need a little less BSC.
I was asked about a schematic for the Classic 8 filter so I thought I'd post it to the thread. The circles represent nodes where component leads need to be soldered or at least clamped together. The values on the schematic represent the AN8 filter with less baffle step correction, but the layout would be the same for any of the filters shown above (except for the more complicated AN10 classic filter which has four additional components; I'm happy to attach that as well if there's interest). Sorry for the lame drawing but it's my first shot at it.
Attachments
I'll post it tomorrow; too lazy to get the other computer and check the values.
Last edited by a moderator:
I spent some time playing with the simulation this morning and made a breakthrough: my adding a capacitor to the BSC filter I can do away with the second series filter, making the AN10's filter much simpler and cleaner. Here's the final circuit and frequency response chart. It's easy to tune the response: if you increase the value of the series filter cap, you get more treble in the 6-10k range; if you decrease it, you get less. The simulation is for a 120L box 12in wide tuned to about 50Hz, and I factored in about 3db baffle step. I'm going to go back now and see if I can do something similar for the AN8.
Attachments
Success! Here's a simpler and more effective circuit for the Classic 8. This one retains even more HF extension than the old one and doesn't require the addition of a zobel at all. the impedance follows a gentle arc from 8 ohms up to 14 and then back down, while the electrical phase remains very smooth so no problems for tube amps. In one picture you'll see a comparison between the old filter and the new filter (old is in gray).
Attachments
The 10in will work better in a box type cabinet than their smaller units. For e.g., this MLTL: internal dimensions 56in x 13.5in x 10.5in (HxWxD). Zdriver 21in, vent 3in diameter x 1in long. Line the top, back & one side wall from the top 45in down. Brace to suit, and it's a max-flat alignment, so adjust & damp to taste / your own room / system in practice.
This was from a discussion a few years ago about the Audio Nirvana 10" cast driver. I have stumbled onto a pair of AN Classic 10" and wondered if this MLTL cabinet would work for those. What changes might you suggest to make this or a similar cabinet work? I do not really like the cabinet designs listed on the CSA site and want to do something with a little more performance and flair.
This was from a discussion a few years ago about the Audio Nirvana 10" cast driver. I have stumbled onto a pair of AN Classic 10" and wondered if this MLTL cabinet would work for those. What changes might you suggest to make this or a similar cabinet work? I do not really like the cabinet designs listed on the CSA site and want to do something with a little more performance and flair.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.