Hello there,
I've done some experimenting again but my latest experiment was a bit too successful not to share:
I've been toying with back to back mids for a while to get symmetric front / rear response which I want to achieve. I tried stuffing them in a pipe or running naked or with little stuffing but it just measured horrible, the back driver influenced the front and so forth. So at ~ 700 hz I had a huge null on axis which was to be expected. The off-axis null is awesome though, far better than I've had with the normal array of one above the other and one inverted.
In my craziness I got the idea of: "hey, if the problem is the rear driver influencing the front why don't I just add more stuffing?"
So more stuffing I added and it immediately measured a lot better. This kept going until I had added as much stuffing as the current layout allowed and it measured better for each time I added more.
And there are the results:
Which begs the question if the perfect back-to-back setup is of 2 cardioid mids. So my next experiment is to change such that the drivers are proper cardioids, I'm thinking something like this. And the question is if that setup would measure even better than what I've tried so far.
My tweeters are 2 waveguided DT-300s like this so the extra distance between the mids is not a problem but will rather help me interface with the tweeters.
I tried normal top/bottom setup with a tweeter waveguide mockup in the middle which measured perfect up to 1600 hz but then a huuuge on axis null due to the large waveguide, so the goal of this is to avoid that (on top of that it was a MTM so had worse vertical response).
I've done some experimenting again but my latest experiment was a bit too successful not to share:
I've been toying with back to back mids for a while to get symmetric front / rear response which I want to achieve. I tried stuffing them in a pipe or running naked or with little stuffing but it just measured horrible, the back driver influenced the front and so forth. So at ~ 700 hz I had a huge null on axis which was to be expected. The off-axis null is awesome though, far better than I've had with the normal array of one above the other and one inverted.
In my craziness I got the idea of: "hey, if the problem is the rear driver influencing the front why don't I just add more stuffing?"
So more stuffing I added and it immediately measured a lot better. This kept going until I had added as much stuffing as the current layout allowed and it measured better for each time I added more.
And there are the results:
Which begs the question if the perfect back-to-back setup is of 2 cardioid mids. So my next experiment is to change such that the drivers are proper cardioids, I'm thinking something like this. And the question is if that setup would measure even better than what I've tried so far.
My tweeters are 2 waveguided DT-300s like this so the extra distance between the mids is not a problem but will rather help me interface with the tweeters.
I tried normal top/bottom setup with a tweeter waveguide mockup in the middle which measured perfect up to 1600 hz but then a huuuge on axis null due to the large waveguide, so the goal of this is to avoid that (on top of that it was a MTM so had worse vertical response).
