FWIW, this kind of ripples are normal for a narrow baffle, the impact of the baffle edge is quite large (because SPL at the edge is still close to source SPL). But, in some people's opinion they don't matter much for the perceived end result.
To do your crossover, just mentally draw a straight line through the ripples.
Now, none of the above sounds very clean and perfectionist. But in my experience it matters less than the measurements suggest. And the unseen is often more important. For instance, the woofer doesn't show the same ripples. Why? Because (assuming it is a 6.5") at 1.5k+ it has already become directional. So its sound output hardly"sees" the baffle edge.... and no ripples. But the trade off is that now the total power output at these frequencies is lower (off axis). Thinking of it you could use a waveguide for the tweeter (one account of how to do this is on Zaph's site). That will make the tweeter directional and gone are the ripples...
You can also improve the situation by doing a more pronounced roundover (requires a hefty router bit, because the radius would have to be large, say 1" = 25 mm). No router? No problem - you can glue another panel to the sides of the box and glue a 1/4 rounded rod at the end (available wherever you find wood mouldings). That will give you a rounded edge. I used this method for my baffles.
In theory even a 1" edge roundover is not enough to prevent ripples at higher wavelengths, but I noticed that it is enough to make the ripples look a lot better.
To do your crossover, just mentally draw a straight line through the ripples.
Now, none of the above sounds very clean and perfectionist. But in my experience it matters less than the measurements suggest. And the unseen is often more important. For instance, the woofer doesn't show the same ripples. Why? Because (assuming it is a 6.5") at 1.5k+ it has already become directional. So its sound output hardly"sees" the baffle edge.... and no ripples. But the trade off is that now the total power output at these frequencies is lower (off axis). Thinking of it you could use a waveguide for the tweeter (one account of how to do this is on Zaph's site). That will make the tweeter directional and gone are the ripples...
You can also improve the situation by doing a more pronounced roundover (requires a hefty router bit, because the radius would have to be large, say 1" = 25 mm). No router? No problem - you can glue another panel to the sides of the box and glue a 1/4 rounded rod at the end (available wherever you find wood mouldings). That will give you a rounded edge. I used this method for my baffles.
In theory even a 1" edge roundover is not enough to prevent ripples at higher wavelengths, but I noticed that it is enough to make the ripples look a lot better.
MBK said:
Well, I have a different opinion. The ripples are very audible, and since the reflections from the edges come so close in time to the direct sound, they are perceived as being part of the sound source, the brain does not sort them away as reflections (as it can with reflections that arrive later).
The trick is to find a smart placement on the baffle and possibly to use a smart shape of the baffle (usually, rectangular is smart). I would say that this is even more important for the frequency response and the resulting timbre than the choice of driver in many cases.
Also, the Edge simulates the directionality of the drivers*, so the effect that the edges of the baffle is not hit by sound due to directionality is not true for frequencies as low as 1.5 kHz (assuming a standard 1" tweeter). The The Edge curve i showed included directionality of the driver.
*It does so by having several point sources distributed acress the driver membrane. It does not consider cone break-up, but this is hardly an issue at 1.5 kHz for a tweeter.
Michail Bugaria said:
Ah, there it was I knew I had seen it. Yes that is what I meant about the peak at 8 kHz. Actually it seems on those measurements as if such mounting would give effects even further down, I did not remember that.
And no, I don't think that the bolts are important either.
From looking at Zaph's measurements for the 27TBFC/G, and the baffle diffraction simulation, the 8khz dip does not appear out of the ordinary.
From modelling baffle diffraction effects of center mounted tweeters in narrow baffles, I found the golden ratio really works. Take the width of your baffle, divide by 2, and multiply by 1.618. This should be the distance between the top of the enclosure and the center of your tweeter.
With rounded edges and an offset tweeter, you can completely flatten out the on axis baffle diffraction effect in the operating range of the tweeter. Below is an image of what I came up with.
Dan
From modelling baffle diffraction effects of center mounted tweeters in narrow baffles, I found the golden ratio really works. Take the width of your baffle, divide by 2, and multiply by 1.618. This should be the distance between the top of the enclosure and the center of your tweeter.
With rounded edges and an offset tweeter, you can completely flatten out the on axis baffle diffraction effect in the operating range of the tweeter. Below is an image of what I came up with.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Dan
Wow! Thanks for all of the help. This is making me think that to really get the best out of my drivers I should probably make a big "infinite baffle" to measure the drivers on, then take the whole project through the design phase in Soundeasy before cutting more wood. I did not realize that baffle layout could make such a big difference, I should take advantage of SE's baffle modeling system.
jarkaa: To remove the cover just grab on with an appropriate pair of pliers and pull... it doesn't take much. Make sure you have a good grip as the cover is ferris and will get pulled directly twords your precious dome if you are not carefull. After that you are left with some ugly glue, but it comes off easy enough with a scraping with a tooth pic.
Thanks again... it has been very informative!
Chad
jarkaa: To remove the cover just grab on with an appropriate pair of pliers and pull... it doesn't take much. Make sure you have a good grip as the cover is ferris and will get pulled directly twords your precious dome if you are not carefull. After that you are left with some ugly glue, but it comes off easy enough with a scraping with a tooth pic.
Thanks again... it has been very informative!
Chad
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.