Coated vs uncoated Scanspeak 15W8530?

[Mingo]

18W8531 is always uncoated. 18w8530 (of which I have 2) is the coated version.
Concerning the K or G: K is the kapton voice coil former (not sure whether this is the correct word) and G which appeared later is the glass fibre voice coil former.
In some cases Scan-Speak kept the K and still changed to glass fibre.
The SS data page have got the wrong photos in some places then. 0 always coated - 1 always uncoated.

[tktran303]

Thanks for that Mingo,

Apparently all the -K drivers now use Fiber Glass voice coil formers. If I recall correctly, the Dupont company no longer sells Kapton to the loudspeaker industry.

[rabbitz]

Originally posted by alan-1-b
Thanks for your replies fellows - much appreciated !

Rabbitz , have you had any success with an 18W8531G00 with a first order cross-over slope ? , and if so was its nominal crossover frequency above or below 2kHz ?
Thankyou ,

I had a real problem trying to implement a 1st order series as that peak after 1kHz was so pronounced. I didn't want to do any correction or go for higher order parallel crossovers and the saviour was using an AR series crossover which basically has the woofer rolling of about 2nd order and crossed over about 2500hz.

[alan-1-b]

Originally posted by rabbitz


I had a real problem trying to implement a 1st order series as that peak after 1kHz was so pronounced. I didn't want to do any correction or go for higher order parallel crossovers and the saviour was using an AR series crossover which basically has the woofer rolling of about 2nd order and crossed over about 2500hz.

Thankyou for your reply indeed !
"that peak after 1kHz" - do you mean simply the double-humped plateau from about 1k2 to 3k5 as shown on Zaph's plot {and which extends further to about 4k on ScanSpeak's own data sheet} ? , or is there worse in your samples ?
I would have thought a large series inductor - of suitable value - would attenuate the plateau and cause sufficient roll-off above 2kHz for a 1st order slope -{above 2kHz}, but I could be mistaken .
Perhaps an R-C Zobel would be needed across the woofer first , to allow an inductor to work in the manner I was hoping .
Partly I am concerned about that resonance - shown on Zaph's plot at about 3k2 . It seems to be low in amplitude and not of excessive duration , thus a 1st order slope may reduce its audibility sufficiently to not be annoying .
I don't like the audible effect of the reverse polarity connection necessary to one of the drivers in second order cross-overs ! - do you hear its effect ? -{some listeners do to greater degree than others it seems} .

Thanks again for your reply !

[fishball79]

The 1khz peak helped when I designed a small bookshelf loudspeaker as it coincided with the bafflestep.

In the end with a stock 6db BSC reccomended by Tolvan's Edge software I was able to listen nearfield.

[rabbitz]

I haven't seen the plot on Zaph's site and I was refering to the hump on the data sheet.

The large inductor will start the roll off further back than the start of the hump and in the graph, the blue line is a 1st order crossover at 2000Hz (0.52mH) and the grey line is at 900Hz (1.2mH). Didn't work for me but I was using a series xo.

I didn't find a zobel necessary as the inductance in the woofer is very low resulting in a gradual impedance increase. A zobel can work well doing just what you have in mind as I have just done that on a Peerless 830875 to tame a peak at 4000Hz.

I've not had a problem listening to reversed polarity in tweeters and BTW, that's needed a lot in first order crossovers as well and some 2nd order have a normal connection.

Cheers

[alan-1-b]

Thanks rabbitz for all your comments and especially for the comparison plots !
I would normally only use a Zobel in an application similar to the one you mentioned , and I'm pleased that one is not necessary with the 18W .
If I go with this project I'll first try a couple inductors of values in between the two you plotted , to get close to what I'd prefer , and have a listening prototype to evaluate .

You are fortunate to not be troubled by the different audible characteristics of same polarity versus reverse polarity ! , and I do know of some other people who aren't too much .
As for which applies when with both 1st and 2nd order slopes , well , within the limits of my experience the requirement seems to depend on the vertical listening axis -specifically the time delay of signal arrivals between drivers -{and slight additional time differences as result of phase shifts in the filters chosen}.
I hope to construct a time aligned and phase coherent loudspeaker - yes , many have tried and few have succeeded , but hey , such is part of the challenge in DIY !

Thanks again , and best wishes for your projects .

Thankyou fishball79 for reporting your finding . Baffle-step effect has not seemed to trouble me much in listening , but as it does for some listeners , this result you've reported may be useful to other designers who want to try the Scan 18W8531 , or the 15W versions - a pair of which I think it is that you have used .

[rabbitz]

Originally posted by alan-1-b

I hope to construct a time aligned and phase coherent loudspeaker - yes , many have tried and few have succeeded , but hey , such is part of the challenge in DIY !

You bet it's a challenge. You might want to consider inverted drivers on a 1st order with the drivers positioned so they give a 15º upward tilt to compensate for the 1st order 15º downward tilt.

Have fun with the project as both the woofer and tweeter are exceptionally good drivers.

[alan-1-b]

rabbitz , that is exactly what I intend to try first , and indeed I was pleased to see you had built the inverted drivers format , plus in a floorstander , which is what I was planning .
Now that old saying about "great minds" , ... uhm , yes I do remember the end of that saying also ... , so for this case we will likely find out !

Thankyou for your advice , - try this one other interested Forum members !

[Salas]

Physical distance of sonic emission centres between drivers translates to phase shift. If it is severe, a polarity inversion that restores it, is indeed in a manner of speaking 'real positive'.