I’m not sure what’s going on at PE, but getting a pair of decent tweeters has not easy. A few years ago it was the ND28F-6. I ordered 2 and they didn’t match. One was close to the spec sheet. They sent me another one and it didn’t match either of the original pair.
I ordered 2 of the silkies last month to build JR’s latest Cabrini design and a pair of HiVi RT1.3WE to see if the aluminum SB woofer would work in Craig’s Keramiska design. All of them look wonky. I must have angered the gods of HiFi. I already have a ticket in the works for the silkies. I need to get some more measurements of the HiVis before talking with them about those.
Here are the DC28Fs after break-in with pink noise for 26 hours
I realize this could be obvious, but have you re-calibrated your test leads and impedance in DATS? (I only ask, because I have got some weird results before and that was the reason.)
Checking the back cup on that tweeter would be worth checking as well.
Recently I had a tweeter doing something simalar and I pinched it from back cup to faceplate and worked my way around untill the impediance responce changed. For me it was a leaking gasket on the back cup.
That is one thing I liked about ARTA. If you have it do a bunch of consecutive cycles you can poke and prod the tweeter while watching any changes occur essentially in real time.
I used that to tweak the positioning of diaphragms in the horn tweeters I used in the Faux Pas and T-tuxes.
I had that problem with HiVi Q1R’s. One of the first two had way too much ferrofluid. They sent me a replacement that apparently had no ferrofluid. Had to get another replacement that they measured before sending to ensure I received a decent tweeter
I always fire up DATS and let it warm up for a few minutes, then perform the shorted leads calibration, then the impedance calibration with a 1kohm resistor, every single time I use it. Since I started doing that I haven’t experienced any issues or questionable measurement. Only takes a few minutes.
Update: PE is going to send me 2 new Silkies and they didn’t ask for the problematic tweeters back.
I just checked a pair of ND28’s I bought las year and although the Fs is higher than the spec sheet, they both match reasonably well. I think I got those from Brad, who may have matched them before sending them out. I only ran pink noise through them for a few hours and I probably need to give them a good 24 hr break-in to see where we end up.
Follow up report - the replacement Silkies were delivered today. Even without any break-in, they are pretty much spot on. I was comparing the first ones to these and notice the faceplates are slightly different. The new ones have a very subtle texture, while the out of spec pair are more matte.
