I hate to be *that* guy, but I will be *that* guy. I have six of these, bought four very close to when they were released (maybe even a pre-sale) at $25. Bought two more awhile later for $29, maybe? In any event:
Samples 1 and 2 are from the first batch. I pulled these at random, just a coincidence that the only closely matched pair I have are the first two I swept. Again, coincidentally, 3 and 4 are the earlier ones, and represent the only other reasonably well matched pair. The other two are the newer ones, one of which is a complete outlier to the rest - but might actually be the best of the lot. Who knows? It just isn't worth it to me to dick around with PE at this time, so I will keep 1 and 2 for my personal use, sell 3 and 4 since they would are perfectly serviceable, and Ebay 5 and 6 I guess.
The scale is pretty tight, if I was using Omnimic default resolution (I am aware it does not do impedance, but the out of the box resolution makes *any* driver look amazing - that's my point) things would look much more normal. They are probably within a db or so of each other, and depending on crossover point, the drifting FS and magnitude of said resonance may not even matter two shits. On the other hand, those rear chamber resonances around 2400 or so can sometimes complicate things. The magnitude of those is simply not serious, though. Morel has an amazing reputation for selling this level of Qc for 3-5 times the cost, so there is that to consider.
I may sacrifice 5 and 6 5 to see why they hell they are so different, I mean that is pretty different. The other five are +/-15 Hz and +/-0.65 ohm. This outlier #5 is what pisses me off - over 15% of my samples are simply out of whack - and out of whack in a way that makes it a winner.
Weird tweeter, not sure it is worth the regular price but at the $25 I paid it seems to be ok. I have not listened to them yet, I don't think, so hopefully I'll get something together one of these first days and see what we see.
Note: none of these are broken in. That may change things dramatically, it may not change one fucking thing. Not sure I am going to spend that two hours listening to test tones just to listen to more test tones just to format more data to put into a chart, either.
Same approach as the Dayton, these are the recently on-sale Peerless DA25BG08-06 tweeters.
Not broken in, typical disclaimer blah blah blah.
Barring break-in making a huge difference, not overly impressed with these either. 25% of my samples are clearly off enough to probably warrant a different padding resistor in the crossover. The other three appear to be tight from a sensitivity aspect - but the differing frequency and magnitude of the resonance will push the usefull crossover point up a ways - probably looking at a 2200 or higher XO unless one wants to custom tailor notch filters for the resonance peak.
Le sigh.
Possibly up next - a pair of Monacor DT-99. These appear to be a soft dome version of the Dayton titanium thing they have sold for a thousand years.
Fs is reasonably consistent across the two, and if the scale was a little lower resolution the energy storage at 5k and 8k would probably not be visible.
The wrinkles in the passband of the one is concerning, the other one looks fine. The wrinkly one might benefit from a break-in, but it is a glued together tweeter so my old trick of re-seating the VC and snugging faceplate screws is not an option. It might even have ferrofluid spilling out for all I know. I don't give a shit, I am positive these were just a couple bucks on clearance and I probably needed something to hit free shipping at MCM at the time.
Over the last few weeks I have done impedance sweeps on 14 damn tweeters and all of them exhibit drift. Kind of takes the fun out of it, so I really should run these through a break-in period and re-measure. I know some people don't think tweeters need to be exercised before use, but I have heard them change during measurements so I am a believer.
I know some people don't think tweeters need to be exercised before use, but I have heard them change during measurements so I am a believer.
Many Years ago I had some tweeters from Morel that were giving me trouble. I thought there was something wrong with them. I contacted the Morel rep and he said to break them in slowly, rather than hit them with gobs of power right away. He said that some tweeters take a longer time to settle in. He was right. I think it was the MDT30s, but not sure.
I've been chatting with a few PE guys. The double hump in the impedance seems to be due to stuffing behind the dome and in the pole piece. The comments were that the double hump version has a lower FS over the non-double hump. Some liked the lower FS with the double hump
You can tell after opening these the stuffing behind the dome has a hard cap and is not your typical material. I'll be playing as well. Lots of fun.
I'm using number 2 as the guinea pig, but it should work on both. The key is to attack from the rear. I heated up a small screw-driver to slowly pry off the back cup. In most cases, this will likely chew up and ruin the cup. I found some Ball baby-food jar caps at Walmart to facilitate the replacement parts. They are a touch loose, but will work. 2" PVC flat caps are a touch too small, but would also likely work.
Initial:
No back cup:
Removed the foam plug:
Pulled out string of black adhesive (that was across the front of the pole) from the rear using needle-nose pliers, and then inserted a 1" ball of cotton to flush with backplate. There is still no back cup installed:
And finally added the Ball-back-cup with a thin layer of wool batting over a material similar to Ultratouch, slightly overstuffing the cup, and then wrapping with masking tape to hold and seal the assembly for testing:
I did absolutely nothing in front of the felt pad that covers the vented pole piece, and this looks really good!
I hope those of you with bad RST-F tweeters feel that fixing these will be of benefit. I do at the moment.
Ben, is it possible to get a FR sweep? after and before (with the other non modified Tweeter), but still looks a lot of work, for something that should have just worked off the box...
Mine were free in the bag from MWAF, and it took about an hour of playing around. I can see what I get in reference of the other unmodified but I never took one of the unmodified number 2. The Qts is higher than spec, but I even managed to best the spec impedance ripples. There is no ripple above resonance now.
Okay, did as asked. Not cool enough to do much else today. I had to shut off the ceiling fan and the box fan for these, and the family was nice and kept totally quiet for the 10 minutes it took.
I attempted what I call the 'Jeff Bagby driver measurement setup sans cabinet'. I set the tweeters on a sheet of egg-crate on the carpeted floor, then laid a circle of open-cell 1/16" sheet (with center cut out for the driver) atop of it. Then I surrounded the foamed face with 3-4 pieces of Sonic-Barrier. Jeff uses felt, but I had this handy. The mic was about 2 feet from the driver facing down. A 70uF cap was used in series for driver protection.
Bstock1, unaltered:
Bstock2, modded, same levels and setup:
I think that is a NOTICEABLE difference. Granted, I'll also oblige a graph of the modded Bstock1 when I get there. Ben
Well, I as able to get them a lot closer to each other, and reduce the number of problems they had initially. I have a bit of audible 3rd order and a corresponding broad impedance bump around 250Hz in the Bstock1 driver, but everything else is pretty close. I don't know what is causing it, but it didn't look to be there before. Being I'll roll it off long before there, I'm not worried about it. I also took an HD pop of an old RS28A-4 in the same setup, just for fun. Mod1 then Mod2 for all below...
RS28A-4 HD :
The FR of the Bstock modded drivers is now pretty similar, as is the T/S, impedance, and HD. That's all I was really hoping for. Being that the FR is pretty close, but only applicable to these tests, I'll omit those for the sake of not being notable information.
Comments
Samples 1 and 2 are from the first batch. I pulled these at random, just a coincidence that the only closely matched pair I have are the first two I swept. Again, coincidentally, 3 and 4 are the earlier ones, and represent the only other reasonably well matched pair. The other two are the newer ones, one of which is a complete outlier to the rest - but might actually be the best of the lot. Who knows? It just isn't worth it to me to dick around with PE at this time, so I will keep 1 and 2 for my personal use, sell 3 and 4 since they would are perfectly serviceable, and Ebay 5 and 6 I guess.
The scale is pretty tight, if I was using Omnimic default resolution (I am aware it does not do impedance, but the out of the box resolution makes *any* driver look amazing - that's my point) things would look much more normal. They are probably within a db or so of each other, and depending on crossover point, the drifting FS and magnitude of said resonance may not even matter two shits. On the other hand, those rear chamber resonances around 2400 or so can sometimes complicate things. The magnitude of those is simply not serious, though. Morel has an amazing reputation for selling this level of Qc for 3-5 times the cost, so there is that to consider.
I may sacrifice 5 and 6 5 to see why they hell they are so different, I mean that is pretty different. The other five are +/-15 Hz and +/-0.65 ohm. This outlier #5 is what pisses me off - over 15% of my samples are simply out of whack - and out of whack in a way that makes it a winner.
Weird tweeter, not sure it is worth the regular price but at the $25 I paid it seems to be ok. I have not listened to them yet, I don't think, so hopefully I'll get something together one of these first days and see what we see.
Note: none of these are broken in. That may change things dramatically, it may not change one fucking thing. Not sure I am going to spend that two hours listening to test tones just to listen to more test tones just to format more data to put into a chart, either.
Not broken in, typical disclaimer blah blah blah.
Barring break-in making a huge difference, not overly impressed with these either. 25% of my samples are clearly off enough to probably warrant a different padding resistor in the crossover. The other three appear to be tight from a sensitivity aspect - but the differing frequency and magnitude of the resonance will push the usefull crossover point up a ways - probably looking at a 2200 or higher XO unless one wants to custom tailor notch filters for the resonance peak.
Le sigh.
Possibly up next - a pair of Monacor DT-99. These appear to be a soft dome version of the Dayton titanium thing they have sold for a thousand years.
Monacor DT-99
Fs is reasonably consistent across the two, and if the scale was a little lower resolution the energy storage at 5k and 8k would probably not be visible.
The wrinkles in the passband of the one is concerning, the other one looks fine. The wrinkly one might benefit from a break-in, but it is a glued together tweeter so my old trick of re-seating the VC and snugging faceplate screws is not an option. It might even have ferrofluid spilling out for all I know. I don't give a shit, I am positive these were just a couple bucks on clearance and I probably needed something to hit free shipping at MCM at the time.
Over the last few weeks I have done impedance sweeps on 14 damn tweeters and all of them exhibit drift. Kind of takes the fun out of it, so I really should run these through a break-in period and re-measure. I know some people don't think tweeters need to be exercised before use, but I have heard them change during measurements so I am a believer.
I guess I'll get a crack at them as well.... Whether that is by sledge-hammer or not depends on what I can manage with them.
InDIYana Event Website
You can tell after opening these the stuffing behind the dome has a hard cap and is not your typical material. I'll be playing as well. Lots of fun.
InDIYana Event Website
I'm using number 2 as the guinea pig, but it should work on both. The key is to attack from the rear. I heated up a small screw-driver to slowly pry off the back cup. In most cases, this will likely chew up and ruin the cup. I found some Ball baby-food jar caps at Walmart to facilitate the replacement parts. They are a touch loose, but will work. 2" PVC flat caps are a touch too small, but would also likely work.
Initial:
No back cup:
Removed the foam plug:
Pulled out string of black adhesive (that was across the front of the pole) from the rear using needle-nose pliers, and then inserted a 1" ball of cotton to flush with backplate. There is still no back cup installed:
And finally added the Ball-back-cup with a thin layer of wool batting over a material similar to Ultratouch, slightly overstuffing the cup, and then wrapping with masking tape to hold and seal the assembly for testing:
I did absolutely nothing in front of the felt pad that covers the vented pole piece, and this looks really good!
I hope those of you with bad RST-F tweeters feel that fixing these will be of benefit. I do at the moment.
InDIYana Event Website
InDIYana Event Website
I attempted what I call the 'Jeff Bagby driver measurement setup sans cabinet'. I set the tweeters on a sheet of egg-crate on the carpeted floor, then laid a circle of open-cell 1/16" sheet (with center cut out for the driver) atop of it. Then I surrounded the foamed face with 3-4 pieces of Sonic-Barrier. Jeff uses felt, but I had this handy. The mic was about 2 feet from the driver facing down. A 70uF cap was used in series for driver protection.
Bstock1, unaltered:
Bstock2, modded, same levels and setup:
I think that is a NOTICEABLE difference. Granted, I'll also oblige a graph of the modded Bstock1 when I get there.
Ben
InDIYana Event Website
RS28A-4 HD
The FR of the Bstock modded drivers is now pretty similar, as is the T/S, impedance, and HD. That's all I was really hoping for. Being that the FR is pretty close, but only applicable to these tests, I'll omit those for the sake of not being notable information.
InDIYana Event Website