Jims 3 way speaker build

[johnny5jz]

Do you have the FRD playing all 3 drivers in parallel in order to derive the driver offset and phase?

[jimbones]

No I am trying to get a basic set of frd for each driver first. If I cant get that than I figure that 3 in parallel wont help me lol.

[johnny5jz]

The 3 in parallel is used to get the correct driver offset and phase data in XSIM. You load the FRD file running all 3 drivers in parallel and adjust your individual drivers to match by adjusting the mod delays. Without this your crossover will not be correct.

[Chris Roemer]

"Absolute" SPL doesn't really matter (for XO development).
Also if your mic is omnidirectional it won't matter much where you "point" it. If you use the same setup (mic position) and drivers are mounted on your final "box" (baffle) then you can use "measured phase" for your XO work.
Also if your term "port resonant freq." means your tuning freq. (Fb), it is NOT <20Hz (the lower peak), it is the "valley" between the 2 peaks, which looks to be 35-37Hz or so (depending on which woofer .zma I look at). Basically right on.

With your driver in the center of a room that's an 8' cube (4' from every surface), 3ms (3.3) is about "perfect", and is accurate down to around 300Hz.

[jimbones]

Im talking about the first peak usually called F1.

[DeZZar]

yes gated at 5ms, 1 m out with mic between mid and tweeter.On axis.Yes mic config file is loaded and running.I can not move these boxes they are heavy and I had back surgery so they are staying put. I'll try gating at 3ms and remeasure

Yeah you will definitely need to pull the gating back to 3ms then to avoid those first reflections. Each driver measurement should be directly on axis with that driver - not half way between the two. If half-way between the tweeter and midrange is your intended listening axis then you can measure on that basis to verify the crossover when built - but not when taking the measurements for each driver,

A timing reference loopback and mic distance information in REW will take care of phase so long as each driver is measured at the same precise distance exactly on-axis. Measuring half way between will cause all sorts of little phase issues between the simulations and reality. With consistent drive level and mic distance you will have accurate relative SPL between the drivers as a start - regardless of absolute SPL.

Driver offset can be taken care of in software like VituixCAD when simulating the crossover. You can work with whatever listening axis you intend then - directly on tweeter, half way between, on midrange etc..

[Wolf]

"Absolute" SPL doesn't really matter (for XO development).
Also if your mic is omnidirectional it won't matter much where you "point" it. If you use the same setup (mic position) and drivers are mounted on your final "box" (baffle) then you can use "measured phase" for your XO work.

Not with Omnimic you can't. You have to take the 3 measurement for 2way and 4 measurement for 3way scheme to gain the reference from plot to plot. If using Holm, REW, Arta, etc; those would be fine with a 2 channel mic system and time lock, but Omnimic is a single channel system.
Wolf

[jimbones]

Wolf so you are saying one measurement for each driver then all three together? Another question, the amp I am using (NAD) is not stable at very low impedances. nominally that would be 2.5 ohms. Can i put a 1 ohm resistor in series to make it about 4 ohm?. I will also gate down to 3ms.

[jimbones]

Gated at 3 ms. woofer, mid a tweet screen shots. frd will be in following oist

[jimbones]

its 3 ms gating, ignore note that says 2 ms.

[jimbones]

frd. measured 3 ms gating, 36 from front baffle on axis, between mid and tweet. 33 uf cap on tweeter.

[jimbones]

Chris Roemer I input your values into XSIM, Did I interpret your post correctly? See jpgs

[Wolf]

As someone else told me at the power levels required, you really don't have to worry about the impedance at the amplifier. This is also a short-term sweep, and that helps not overload things. If you are very unhappy with running in parallel, you can run them in series instead for a decent load.
Wolf

[jimbones]

Got it~! Ill do measurement now

[jimbones]

3 drivers together Fr and Impulse

[Chris Roemer]

Your XO looks good. Did the impedance hold up?

[jimbones]

very good It measured about 3 ohms. I thought it was going to be about 2.4.Wjat is the next step now that I have the frd for all three drivers. I remember a convo (not sure if it was you) about correcting for phase error or z axis in the XO.

[johnny5jz]

I ran a quick sim on your data, and I came up with the following. It is similar to yours.

[johnny5jz]

I used the FRD file with all 3 drivers to set the delay of the mid to 2.18" and the woofer to 10.21".

[jimbones]

Thank you for running a sim. Since I need to learn this if you could expand on how you came about with your results. For example, did you use Omnimic to adjust delays and what did you look for? Sorry to be a pain. Im a novice and I only know basics. I would like to know how to account the drivers being in all different setbacks on the baffle

[johnny5jz]

I import the 3 FRD and ZMA files into xsim and hook them up with no crossover. Then go to curves and click get file. Import the file with all 3 drivers. Now open the mid and adjust the mod delay until the curves match the best. Then go into the woofer and do the same. Without doing this, your crossover will not be correct.

[jimbones]

johnny5jz OK I did that. It worked pretty well for the mid. It takes a little imagination it's not like you get them to match really well but I got the same number as you so I will say that is good. However I am struggling with the woofer. I am trying to position the 2 troughs in Fr to align with each other and I get a bit over 11 inches.Can you double check the woofer for me. So if I now go and develop the XO with this new info that means I will have taken into account the drivers delays correct? Much Thanks everyone.

[johnny5jz]

Going to a mod delay of much over 10.5" gets worse for me.

[jimbones]

johnny5jz Does this look right? Its ugly Lol