I did the second speaker. Traces from top to bottom are first speaker, second speaker, and the difference. Traces separated for clarity.
Not exactly sure why the second speaker turned out so much better (audibly so). Maybe it has something to do with the way I stuffed the cabinet.
I constructed the crossover differently in each box but they both still buzz at the same frequencies and volumes, so I think I was just hitting the SPL limit for these speakers.
As for the 6KHz notch, it seems to have gone away on it's own...
Not exactly sure why the second speaker turned out so much better (audibly so). Maybe it has something to do with the way I stuffed the cabinet.
I constructed the crossover differently in each box but they both still buzz at the same frequencies and volumes, so I think I was just hitting the SPL limit for these speakers.
As for the 6KHz notch, it seems to have gone away on it's own...
Attachments
Last edited:
The buzz at some notes is the spring connectors in the back of the speaker. It happens because I use them much louder than they would typically be used. I don't know of a way to quiet buzzing spring connectors.
It's the spring inside the connectors that is buzzing at it's resonant frequency. I don't know if I could get to it to pack it with hot glue, but the application would have to be minimal to keep the connectors functioning as such. I have thought about stuffing them with foam, but again not sure if I can even get to the springs.
So, it's been a while, and it seems the speaker has been breaking in, a lot.
Either the woofer is much more sensitive now or the tweeter is much less sensitive. And the 3KHz woofer resonance seems to have reduced in intensity.
Woofer and tweeter measure with the same DCR that they did when I designed this crossover.
Either the woofer is much more sensitive now or the tweeter is much less sensitive. And the 3KHz woofer resonance seems to have reduced in intensity.
Woofer and tweeter measure with the same DCR that they did when I designed this crossover.
Both speakers measure the same way, so either my microphone has shifted or maybe both AMTs lost some magnetization. I'm guessing the mic...
My microphone does not seem to have problems at LF. It seems both tweeters whether they were being used or not have lost 5db or so of sensitivity. I don't know why this might be. They are AMTs.
Funnily enough, I changed the tweeter resistor back to the original value of 4ohm to match the sensitivities. Maybe Dayton already knew this would happen, but is there anyone discussing it?
Funnily enough, I changed the tweeter resistor back to the original value of 4ohm to match the sensitivities. Maybe Dayton already knew this would happen, but is there anyone discussing it?
I've decided that due to the tweeter's higher dispersion, it makes sense for the treble to be lower. The frequency response isn't as flat, but it really sounds incredible this way.
Therefore my original crossover mod stands. I have done days and days of crossover mods, shuffling drivers around in boxes, checking with other drivers etc. In the end my original crossover still sounds so much better.
I was sure that after fixing the response the speaker would sound better, but clearly this is a case where sound doesn't correlate to one simple measurement.
So the crossover in post #19 still stands. I may do some further work to optimize the crossover since the drivers have parameter drifted. I thought they were already fairly broken in by the time I made the crossover, but maybe not.
Therefore my original crossover mod stands. I have done days and days of crossover mods, shuffling drivers around in boxes, checking with other drivers etc. In the end my original crossover still sounds so much better.
I was sure that after fixing the response the speaker would sound better, but clearly this is a case where sound doesn't correlate to one simple measurement.
So the crossover in post #19 still stands. I may do some further work to optimize the crossover since the drivers have parameter drifted. I thought they were already fairly broken in by the time I made the crossover, but maybe not.
It is good that you could break away from having a flat response, now that you have discovered reasons.
I always knew there were reasons, but I figured I would just start with a flat response and then deviate from that as necessary. And for a long time, I was satisfied with a flat response. I'm not sure what changed between then and now.
Some things come through clearer with a flat response, and with the reduced treble it can occasionally sound a bit muted (perhaps the subject of further work). But to have the whole range integrated so that music comes through compellingly requires some reduction of the treble, and that is not really felt as a concession when you are able to hear it.
Some things come through clearer with a flat response, and with the reduced treble it can occasionally sound a bit muted (perhaps the subject of further work). But to have the whole range integrated so that music comes through compellingly requires some reduction of the treble, and that is not really felt as a concession when you are able to hear it.
Last edited:
This makes sense. You discover the big changes needed. You also need to find the little changes needed. These happen for different reasons and you will find them at another time 😉
Sorry comin’ soo late to this party...
1. What did you do about the binding post buzz? PE sells an inexpensive model banana BP in a pkg to do a pair of speakers.
2. I’m intrigued by this speaker. PETT has a thread upgrading the mid woofers. Together w/your R&D, one could DIY something pretty decent
3. Off topic: BSC corner frequency for the 652 Air? I understand BSC is dependent upon the front baffle width. The 652 and my towers have the same front. I have good DSP for the BSC EQ, I want to tweak it. My approach WAS a shelving boost but I plan to change that to a shelving cut. Please recommend a slope & frequency.
Thnx in advance. Tony
1. What did you do about the binding post buzz? PE sells an inexpensive model banana BP in a pkg to do a pair of speakers.
2. I’m intrigued by this speaker. PETT has a thread upgrading the mid woofers. Together w/your R&D, one could DIY something pretty decent
3. Off topic: BSC corner frequency for the 652 Air? I understand BSC is dependent upon the front baffle width. The 652 and my towers have the same front. I have good DSP for the BSC EQ, I want to tweak it. My approach WAS a shelving boost but I plan to change that to a shelving cut. Please recommend a slope & frequency.
Thnx in advance. Tony
I just improvised everything based on someone else's design. I can't recommend anything. I can only say I think this mod turned out well.
As for a woofer, I think it should not be any shallower than the stock woofer so that the dispersion pattern of the tweeter and woofer are better matched. I have an old shallow Radioshack woofer I tried in it which works well except for the dispersion effect and the old woofer's poorly controlled cone modes (for most things I still prefer the stock woofer).
I haven't done anything about the binding post buzz yet.
As for a woofer, I think it should not be any shallower than the stock woofer so that the dispersion pattern of the tweeter and woofer are better matched. I have an old shallow Radioshack woofer I tried in it which works well except for the dispersion effect and the old woofer's poorly controlled cone modes (for most things I still prefer the stock woofer).
I haven't done anything about the binding post buzz yet.
Remove the stock spring clips and cover with piece of wood with two proper 5 way screw threaded binding posts.
Thanks for posting the BSC mod on post 19. Looks like a nice speaker now as measured and with listening tests.
Thanks for posting the BSC mod on post 19. Looks like a nice speaker now as measured and with listening tests.
So quick question, just received my B652-AIRs and after some use, I'm noticing a lot of what appears to be the tweeter fluttering with higher frequency harmonics at all volume levels. Extremely apparent in Home by Aurora. Is this a known issue in the stock config and @keantoken can you test to see if your crossover addresses this?
If it was the woofer producing this fluttering I would assume it was the woofer breaking up as it approached its upper limit but by cupping the tweeter I can isolate where the majority of the distortion is coming from in, at least with this record.
Appreciate the work you've done on this so far!
If it was the woofer producing this fluttering I would assume it was the woofer breaking up as it approached its upper limit but by cupping the tweeter I can isolate where the majority of the distortion is coming from in, at least with this record.
Appreciate the work you've done on this so far!
Are you hearing it at 1:40 here?
AURORA - Home - YouTube
It might just be part of the music. This tweeter works up to 40KHz and is boosted from stock to begin with, so details like that can jump out. I often hear electronic clicks or speaker clicks that were recorded and used in the music and they sound like it's my speaker. It's not, it's just that the tweeter reproduces them so well it sounds like something is wrong.
If there is a problem I would check that the nonpolar electrolytic capacitor in series with the tweeter is not shorted, as that could allow bass content to thrash the tweeter. My crossover uses a much smaller tweeter capacitor, so if it is a feature of the stock speaker the mod will probably reduce it. I listen to these at extra power due to quirks of my listening setup.
Since I actually measured it a few days ago, here is my B652 frequency response. You can see it goes to 40KHz. But the 6KHz notch is back... I really have no idea what it up with that. Note that this is just a measurement at the optimal listening angle which tilts the speaker up a bit to better time align the tweeter and woofer.
AURORA - Home - YouTube
It might just be part of the music. This tweeter works up to 40KHz and is boosted from stock to begin with, so details like that can jump out. I often hear electronic clicks or speaker clicks that were recorded and used in the music and they sound like it's my speaker. It's not, it's just that the tweeter reproduces them so well it sounds like something is wrong.
If there is a problem I would check that the nonpolar electrolytic capacitor in series with the tweeter is not shorted, as that could allow bass content to thrash the tweeter. My crossover uses a much smaller tweeter capacitor, so if it is a feature of the stock speaker the mod will probably reduce it. I listen to these at extra power due to quirks of my listening setup.
Since I actually measured it a few days ago, here is my B652 frequency response. You can see it goes to 40KHz. But the 6KHz notch is back... I really have no idea what it up with that. Note that this is just a measurement at the optimal listening angle which tilts the speaker up a bit to better time align the tweeter and woofer.
Attachments
Last edited: