In a previous post, I described how I performed some troubleshooting on a number of crossover (XO) networks for the EAW KF650 — specifically KF650isfx. At the conclusion of that post, I described what I thought were the failed components of each XO assembly.
Since then, the good news is that EAW is still stocking the required parts to effect repair. Almost is good is that I received all the necessary parts, delivered, for less than a Franklin.
So I repaired the XOs with the replacement parts.
After repair, I used SMAART to measure the transfer function (TF) of each of the XO assemblies.The set setup was the same as in the previous post.
Here is a TF of all five LF modules, superimposed:
The lack of correlation at ~1.25KHz is puzzling. However, all five demonstrated the same issue, and the level is already attenuated by about -60 dB at this frequency.
Here are TFs of all five MF XO modules, using the same IO Panel 3203 for a control:
Excellent! And all five IO Panels, using the same MF module 3203 as a control:
Again, excellent.
Lastly, we have TFs for all five HF XO modules:
Unit 3204 has a slight dip at 2K. Note however, that the phase at that frequency is spot-on. Other than this slight attenuation, they match almost exactly.
The upshot of all this is that the repairs have been effected, and the XO’s match extremely well with each other.
Three-way operation
In order to visualize the operation of the crossovers as a whole, the following is a graph of representative LF, MF, and HF modules superimposed:
Note that the mids and highs are horn loaded. Accordingly, they are acoustically more efficient than the lows. In terms of acoustic output, it would help to visualize the MF and HF curves as boosted by 7dB from their current values.
Here is a high resolution look at the magnitude curves:
And a look at the phase curves when unwrapped: