Next ?

[dynamo]

John with your files in my sim i still need the mids polarity inverted ....

How do your nulls look if you flip your mid polarity and measure with the mic?

[filmslayer]

How do your nulls look if you flip your mid polarity and measure with the mic?

John is sending me a couple of resistors i don't have . just sims at this point .

[rpb]

from 3ft tweet axis and near field ...

[ATTACH=CONFIG]n1364937[/ATTACH] [ATTACH=CONFIG]n1364938[/ATTACH]

Take the 3' measurement, and make it gated. Put it on the same graph as a full range gated measurement like the one in post 173. Match the responses from 100hz to 300hz. Now you can see how high the woofer is crossing over. Better still, do the same, and add the mid response . (with, or without the tweeter.) The x-over point is where the mid and woofer responses cross. Also. The summed output at any frequency is the sum of the woofer, and mid at that frequency. Once one driver is about 12dB below the other, it should have very little contribution to the summation. If one driver is about 6 dB below the other, then it's still having some effect.

[filmslayer]

just did round overs , gotta get it all back together .

[jhollander]

I do too, but these are done following Jeff's instructions to a tee. Mine is with a 10F which if you look at the tech drawings is incredibly similar in proportion to the ds90. Only other difference is my 10F is surface mounted, so take off the flange thickness from the +13mm offset (3.5mm iirc). First pic is +13mm, second is 0mm.

I'm sure your offsets are correct for your build. With a sim there's a lot of guessing so my process is to pick a middle of the road estimate. In terms of importance in a traced simulation I think the the SPL match, baffle response, and driver position, are above the driver's z.

One thing for sure is that my sim is not going to match reality.

[filmslayer]

rpb , here ya go .

[rpb]

How far from walls and floor is the speaker. If you can get the woofer higher, possibly by turning the speaker horizontal, or upside down, then the woofer gating might can be set lower. As it looks now, the woofer response appears to be about 6dB below the mid at 1k. That means that it is going to increase the summation at 1k, and produce the bump in response that you are trying to eliminate. Increase the mid gating a little till it smooths slightly, and see if 1k is humped at all. Doesn't look like it is. Looks like the mid is dialed in for a 550hz x-over point. The woofer appears to be set higher.

[filmslayer]

i think it is in my high pass part of the mid filter ... speaker is in the middle of a 16' x 16' foot room so walls are not the issue but it is close to the floor ...

[rpb]

i think it is in my high pass part of the mid filter ... speaker is in the middle of a 16' x 16' foot room so walls are not the issue but it is close to the floor ...

Flip the speaker. The problem is the woofer filter. Trust me. You'll see.

[filmslayer]

.

[rpb]

[ATTACH=CONFIG]n1365005[/ATTACH] .

See... The mid response is perfect. Now you need a woofer response that doesn't screw up the response at 1k and above.

After you flip it, measure the woofer and mid together, and show the sum on the same graph as the mid above.

The wild swings in the woofer response are caused by floor reflections.

I looked at the woofer close mic, and I think the woofer needs a bigger coil. Can you add a small one?

How high is your ceiling? If it's 10 foot, get the woofer and your mic half that above the floor.

As you can see, measuring a 3-way woofer can be a challenge.

Don't let the speaker fall on you!

[filmslayer]

perfect ? that 10db drop around 2k that is seen in everyone of my measurements is in woofer circuit ?

[rpb]

perfect ? that 10db drop around 2k that is seen in everyone of my measurements is in woofer circuit ?

That's a 2dB scale, so it's a 4dB dip. The mid response is good, bordering on perfect. It might need to be padded a couple dB. We will see with additional measurements I mentioned above, if this forum allows me to participate. I tried for 10 minutes to post, and this software fought me every step. It may turn out that padding is all that is needed. The shape of the mid roll off is ideal.

Does the dip show up in a raw measurement of the mids without any filtering? If it does, then it's not an error in the filter.

[filmslayer]

don't for get the " remember me " box ... i've been logged in all weekend with no issue .... i will say it sounds a whole lot better with the round overs !

[rpb]

don't for get the " remember me " box ... i've been logged in all weekend with no issue .... i will say it sounds a whole lot better with the round overs !

My computer is setup not to accept cookies. That may be causing my issues.

Since I have to rush my posts, I don't always make my meaning clear. I post, and see if it shows up, then I edit, or add to the post. Sometimes I'm not in a hurry, and I write a long post and copy it. I tried to edit the "trust me" post after looking back at some other measurements. If I logged onto this page, the software logged me out. If I logged in, it took me to page one. If I switched to page 13, it logged me back out. I left the site twice, and shut down my computer, and came back. If it was always this bad, I'd not even come to this forum.

I know it's a lot of work making repeated measurements, and I'm just trying to show you how to see what's going on with the speaker. What you hear is more important, but what you measure can help explain what you are hearing. What you sim might not be as accurate as you would like, but a sim can help a lot. Measurements can fool you as well. It sometimes depends on factors you can't control, such as measuring in a small room.