I can't really explain the difference, except to say that both of mine measured identically, and it couldn't be due to a mic calibration problem. When I measured the tweeter running free without the crossover cap, the maximum output was 8 dB below the woofer breakup peak at 4 kHz. So there is no way the tweeter could attain the output shown in the other two sets of measurements, because the woofer runs free. Plus, they sound exactly like they measure. I'm constantly checking my measurements with Jim Salk, and while they may differ by a dB or two at the very top, there's nothing that could explain the difference between my measurements and, say, Stereophile's. I took this plot on the tweeter axis, but going further down or off axis horizontally didn't really help. There was a little less cancellation between the tweeter and woofer when I measured nearer the woofer axis, but the woofer peak got higher (as it naturally would). I also tried backing up to the maximum possible mesuring distance (around 50"), but it was still the same banana. If I get a chance, I'll try and get a measurement using the Parts Express mic and software. But I've always got reasonable agreement in the past.
I understand that both of yours measured identically, but it's quite amazing that Zaph & SP appear to be in closer agreement and that with being done a few years apart. I'm not trying to call you out Dennis, I hope you understand that. I'm looking for reasons. A few that come to mind, and by no stretch of the imagination am I an expert here, are as follows...
Could it be that the microphone was positioned differently in your case? Zaph says he measured midway between the tweeter and woofer but IIRC, he didn't give the distance. Don't know about SP but what about you?
Could it be the microphone? As you said, you'll be looking at using the PE setup.
Could it somehow be related to to any smoothing in the software? I don't know what anybody used here.
Could in be in fact that your speakers are different from theirs? By that I mean is it possible that the values for your crossover components are different or perhaps defective? What about cancellation effects that may be a result of the drivers being mounted differently like not flush mounted with the baffle?
Could one of the drivers, even in both speakers, somehow be different or defective from Zaph's or SP's? I mean at this price point, I wouldn't expect an inordinate amount of production QC going on. More like get them out of the box and stick it in. Rinse, repeat.
Now, as Zieglj01 said above, regardless of which measurement one takes, they still measure poor. However, if one is going to base an improved crossover design on measurements, then I think reconciling measurements takes on a certain degree of importance.
I once worked for a firm that had a few manufacturing sites around the country and overseas. Chemicals if you will. Mother solutions would be prepared made with known amounts of elements, all NBS traceable. This mother solution was then divided and sent to the respective laboratories for analysis for the purpose of comparing the results and then determining if the numbers were in reasonable agreement. None of the laboratories had the same equipment, meaning the exact same model number of whatever. Nor did they use the same reagents, water, etc. Sometimes a problem was found and it was then incumbent to determine the reason for the discrepancy and correct it. As you said, you were looking for a check against SP here and what was found was the agreement was not so good. Perhaps you can contact JA if needed and then look to resolve the discrepancy.
