Perhaps an explanation? Very different results when tested...

[altoncustomtech 773]

Anyone who's read through my build log knows I had a set of Bravox CS60K components in there for awhile before going to the FR/MB setup.  I tried a lot, and I do mean A LOT of things to get the speakers in the component set to sound right and was never able to succeed.  In my pursuit of having all the handier tools at my fingertips to use and play with I have been trying to get a DATS V2 from PE for most of the summer and it was finally back in stock this past weekend so I finally got it ordered.

 

It showed up yesterday and to play around with it I connected one of the old midwoofers from the Bravox components just to give it a shot.  I got a weird result on the numbers but wasn't able to play with it anymore.  Last night before going to bed I tossed the numbers it pulled from the woofer into WinISD and it couldn't even calculate for it.  It couldn't show a useful alignment and tried to automatically calculate a -1.606 cuft enclosure (apparently the driver needs to be ran in a wormhole??) so I knew something was up. 

 

So today I throw the other one on the DATS and get a perfectly normal looking, very nice set of numbers and when I plugged them into WinISD everything looks pretty damn good and normal there as well.  Figuring it was probably a fluke anomaly I threw the first driver back on it and got very consistent results with the first time I tested it after running it five times.  So I threw the second driver back on and did the same thing with it.  Attached are the numbers it pulled and you'll see what I'm referring to, there appears to be some sort of issue with the first driver.  At this point it has me curious as to whether or not the issue with the sound I had was either partly, or entirely due to some sort of problem with the driver itself.

 

 

Here's the 2nd, more normal looking driver results.

 

 

 

 

 

 

 

Here's the 1st, funny resulting driver results.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Look at the differences between the Re, Qts, and Qes.  Would love to be able to figure out if it happened during use or if the differences existed from manufacturing.

[Quentin Jarrell 671]

First off the dats is a hunk of junk. The specs it's showing is proof of that.

Assuming you used the added mass method it's calculating a mms of ~10 percent more with a Cms that's nearly identical and a motor force factor that's enormously different. It not physically possible.

[altoncustomtech 773]

It's no Klippel by any means, but I figured for messing around it's not completely horrible.

 

Well, I would have agreed it's not possible except that I've tried it now on nine other drivers, pulling numbers on each one 5 times.  Each of those five times the numbers are very close.  Also on eight of those tested, half were the same types of drivers (four Bravox marine speakers, and four Fountek FR83's) and they too are also all pretty damn close to each other.

 

I do believe this one Bravox mid driver has some kind of defect.  What it is, I have no clue, but everything else so far has been pretty consistent except for it.

[95Honda 486]

One is dampening completey different than the other.  Almost makes me wonder is something is wonky with one of the suspension components and/or gap alignment...

[altoncustomtech 773]

Something is definitely wonky with the one, I also grabbed my DMM to verify the Re measurements, which happened to be spot on.  The voice coil on the one definitely has a DC resistance that's over 10 ohms.  I'm about half tempted to tear them both apart and compare the coils, gaps, suspensions and everything else I possibly can to try to find the cause of the wonkiness of the screwy one.  The voice coils, visually through the vents on the frame, look like they're both good so I believe it's going to come down to a closer inspection.  It could be my eyes playing tricks on me, but it appears that there may be a slightly larger gap between the point where the cone glues to the coil former and where the spider glues to the coil former on one over the other.  If there is any difference between the two its super slight, perhaps a millimeter or two.  Not sure how much of a difference that would make on response/measurements, surely not that much and it sure wouldn't affect the DC resistance.

 

I did notice I got CONSIDERABLY more consistent results with the DATS v2 when I tested a driver with it in a super quiet environment with no air moving and no background noise.  It's definitely no Klippel, or even the LMS LX500, but for getting some numbers for drivers that I would otherwise never be able to know the T/S of without spending thousands on at least the LMS system it's got to be better than nothing.  I have a pair of old ten and twelve in woofers that came out of some old console cabinet stereos, along with a fair number of other miscellaneous drivers that either the manufacturer is no longer in business or the data was never publicly available, and many other reasons like that a device like the DATS could come in handy for.  I figured as long as it's consistent then it's good enough for some of that old crap or even other stuff that I can't find/get data for, or at least better than just shooting in the dark.

[///M5 2,833]

I'd be tempted to play some subsonic frequencies near excursion to see if you hear/see differences between the two.

You are measuring a difference so there is one. Our group of 5 testing showed the relative uncertainty of the measurement.

[altoncustomtech 773]

Okay, so I played them at full tilt on 100, 80, 60, 40, and 20hz tones.  The bad one with the high Re didn't move as much as the good one at lower frequencies.  There was no audible differences in the sound/noise from full excursion movement in one from the other though.

 

So after that I took them apart, wanting to see if there was any noticeable physical difference between them that I could see.  Visually the coils look great with nothing unusual to note about them.  I'm thinking there's an issue of some sort with the connection of the tinsel leads to the coil causing the high Re on the bad one.  It looks as though the coil leads are wound around the tinsel with some sort of compression/crimp on sleeve that attaches them to the tinsel lead itself, then glued down into the joint where the spider meets the coil former.  As best I can tell the coil heights are identical and I can't imagine that just one or two extra or lost windings around the former could account for so much of a difference in the dc resistance.  I also can't visually see a difference in the gaps, pole height, top plate thickness, or anything else that would have an effect on the measurements.  At least visually I not noticing anything but at this point I'm super curious as to what's going on so I'm going to go over it all with a fine tooth comb.  I'm going to grab a set of feeler gauges and a micrometer and see what, if any, differences there are in things and go from there.

 

Here's some pics of the disassembled drivers (marked good and bad) just for giggles.