Help desperately needed in order to tame over zealous tweeter!

[Karma Train]

Hello peeps :)

Can somebody a whole lot brighter than myself offer up some advice please? I'm in danger of going down with a bad case of tinnitus here.

I just got hold of a lovely old pair of Rega XELs (photos to follow) and in general they are absolutely brilliant, too bloody brilliant in fact!

The bottom end of their response is absolutely perfect, I have had numerous problems with bass-lift in my horrible listening room in the past but I get no such issues with these speakers, more than likely because they are of a transmission line design. As pretty much every other speaker I've tried in this room causes said bass-lift to occur, I really want to hang on to these if I possibly can.

As ever, nothing in life is perfect and with these XELs, whist the bass is deep, luscious and beautifully tame, the top treble, particularly with female voices can be excruciatingly sharp at times, so much so it renders listening to some of my favourite music a complete impossibility. I have tried different amps, Quad 34/306, Marantz PM-75 and currently an Arcam A85 but the shrillness has persisted, I've tried different cables which have done nothing to alter the level of sickening sibilance and conventional tone controls are completely useless as you have to destroy the entire top end before the problem area is touched. I've messed about with software and hardware equalisers also and whilst they get closer to solving the problem, I hate using them as you spend your entire life fiddling about with settings instead of listening to music.

So I'm looking for another solution if indeed such a thing exists. There is precious little about the XELs out there on t'interweb but I did find a couple of mentions pertaining to their undue brightness and on one such post, another guy had posted a remedy which involved tinkering with the crossover. I shall copy and paste his post in full below so's you can give it the once over. As for myself, I do kind of know how to handle a soldering iron but I know nothing of electronics (I'm a guitarist for Heaven's sake ;)) and wouldn't be confident in doing the work myself but if I could source the parts I do know a man who could do it for me.

I guess what I'm asking here is, do you guys who know about this kind of stuff think the poster's solution might be viable, or if not, are there any other methods of damping down the tweeter from hell which I could try but haven't thought of as yet?

Bless you all for your time, here is the crossover fix I mentioned and some photos of the offending articles.

Kindest regards, Charlie.

Rega XEL tweeter modification.
I already wrote a review some time ago. What I would like to propose here is a quite simple cross-over modification that will smoothen the treble and improve the transition midrange/treble. There is no change of the cut-off frequency, only a slightly different roll-off curve and a slight attenuation in the mid-treble range.
Here goes:
Change the 3.3 microfarad capacitor for a 3.0 microfarad capacitor and add another 47 microfarad capacitor after the inductance pole, just before the tweeter. What you get is a 3° order cross-over instead of 2° order: 3.0 microfarad in series, then a 0.39 mH inductor in parallel to another capacitor of 47 microfarad in series with the tweeter.
I tried at least 10/15 modifications before coming to this conclusion that attenuates the tweeter just where it needs to, without compromising the midrange and keeping the correct phase offset for a still very focused sound in the woofer / tweeter transition. The treble gets less harsh and more refined, especially in combination with my new TAC 34 tube amplifier.
Whoever drives these speakers in a rather reverberating room (wooden floor, long flat uncovered walls, windows) and struggles with a treble that's a bit too crisp, should try this quite easy modification.

http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00215_zpsc605bfc0.jpg (http://s254.photobucket.com/user/Karma_train/media/Ebay%20Uploads/DSC00215_zpsc605bfc0.jpg.html)
http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00219_zps961ad15b.jpg (http://s254.photobucket.com/user/Karma_train/media/Ebay%20Uploads/DSC00219_zps961ad15b.jpg.html)
http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00229_zps7a4a48e0.jpg (http://s254.photobucket.com/user/Karma_train/media/Ebay%20Uploads/DSC00229_zps7a4a48e0.jpg.html)

[istari_knight]

Is the pictured crossover entirely original ? It just doesn't look right to me :scratch:

[DSJR]

One thing to add if you feel like it...

The tweeter Rega used then, as well as Naim (and Linn in Jurassic times), is to my best knowledge the un-damped one, which whistles nicely at 9kHz and sounds it on non-vinyl sources in my experience. Rega like their spit and fizzle and all their current speakers sparkle to a greater or greater extent I've found ;)

Scan do produce a damped version of this tweet I understand, as used by Pro-Ac and Spendor. Similar fitting and certainly better performance, but I don't know how different the sensitivity is not having looked. Scan do publish full specs, so this might make a possible thing to look at?

[brian2957]

I have a pair of Rega EL8s and in the past they could be a little 'raucous ' with certain tracks and types pf music . I sent my crossovers to Paul at RFC ( Reference Fidelity Components ) who is a member here , for refurbishment .
He totally transformed the sound , while still retaining the basic characteristics of the speakers . I'm really happy with the service and the work Paul did for me . He's also very approachable and helpful , a true gent . You won't regret it .
http://theartofsound.net/forum/showthread.php?29244-Crossover-refurb&highlight=rega+el8

[istari_knight]

Looks like a job for Paul [RFFC] to me :D

I suspect fitting a simple L-Pad will sort you out but still cant get my head around that xover, the solen cap & air core inductor are for the tweeter & in a Rega speaker the midbass usually has a 1st order filter which is the ferrite core inductor you can see in the middle... But everything on the left hand side - I've never seen anything like that in a rega speaker.

[Karma Train]

Looks like a job for Paul [RFFC] to me :D

I suspect fitting a simple L-Pad will sort you out but still cant get my head around that xover, the solen cap & air core inductor are for the tweeter & in a Rega speaker the midbass usually has a 1st order filter which is the ferrite core inductor you can see in the middle... But everything on the left hand side - I've never seen anything like that in a rega speaker.

OK, so I now officially have a migraine, I didn't understand a word of that, sorry to be so irreversibly thick! My talents lie in other areas, or so I'm told :)
My gut feeling is that the crossovers haven't been messed around with, based on nothing other than the guy I bought them from was their first owner and my guess would be that he knew even less about this kind of stuff than I do, if that's even possible!
I do really, really thank you guys for having a look at my thread though and for the couple of possible remedies suggested thus far.
I have bookmarked the Reference Fidelity Components site (what a wonderful place that seems to be) and I might give them a ring to see what they suggest. Don't much like the look of £100 starting price to sort out a pair of crossovers as I am perpetually skint but if that's what it takes then sobeit, these speakers have enormous potential and I really don't want to give up on them if at all possible.
I did ring Rega last week and a guy called Simon suggested I checked that the tweeters were the official white label Rega ones, from there he went on to say that some people had replaced the tweeters with the same model number units but these were different to the Rega tweaked versions, I believe he did mention that Rega drained the fluid from the normal versions of the tweeter....good work Rega, thanks for making my ears bleed!
I did try to remove the tweeters to see what I had in there but the buggers are glued in solidly, according to Simon it's just a silicon gasket but I could get them to shift at all, not without wrecking the baffle that is.
OK, so thus far two remedies have been suggested, have the crossover rebuilt or replace the tweeters with a fluid filled one.
Subsequent questions would be....if I replaced the tweeters (favoured option) would anything else need doing to the crossover or would it just be a straight swap?
Or if the crossover was rebuilt would this be a good idea considering that Rega used a particularly nasty tweeter in the first place, wouldn't that just be a bit like compensating for an inadequacy rather than optimising the board?
Thanks again hugely for your participation in this thread and your suggestions thus far, I really am extremely grateful.
Charlie.

[brian2957]

Paul posts on AOS , so you can PM him . I would give him a call or a PM and see what he can do . He's a very helpful guy :)

[Karma Train]

Paul posts on AOS , so you can PM him . I would give him a call or a PM and see what he can do . He's a very helpful guy :)

I certainly will do Brian and huge thanks for the heads up :)

[brian2957]

I would recommend RFFC to any one who was looking for work done on their speakers Charlie . My experience with Paul was exceptional and he will be my first port of call for any future work required .

[Karma Train]

I would recommend RFFC to any one who was looking for work done on their speakers Charlie . My experience with Paul was exceptional and he will be my first port of call for any future work required .

I still think that replacing the tweeters with non-drained ones sounds like a great idea, I love the Spendor/Proac top end and if they had the same effect on these beasts then I'd be in aural nirvana. I will try to PM Paul and point him towards this thread later on, he sounds like the absolutely ideal chap to ask....thanks again, I really am hugely grateful. :)

[brian2957]

I tried to remove the tweeters on my EL8s and they're the same as yours -solid. You have to remember that these were probably built in the days when analogue was king . They were probable sold to give a lift to a budget turntable system . They can be a little ' unpredictable ' shall we say when used with digital sources. Obviously , it's up to you which way you go here , however it's been my experience that Pauls work tamed these characteristics in my Rega EL8s and greatly improved them in all other areas.
Incidentally the Rega bass units are the same as used in the EL8s and in their latest offerings - the RS3s .
Must be a good'un :)

[Reffc]

Looks like a 2nd order crossover in the HF a first order filter for the mid woofers connected to a notch filter. However it doesn't look right if that is the case as the midbass inductor has one leg not connected back to positive output. In that case, the midbass units would be running full range (assuming the red output wire is connected to the input side without going through the inductor in series) and would be sounding horrid in the mid region. The Solen cap is also not a standard one I've seen used by Rega before so I tend to agree that these may have been altered at some point. Hard to tell from the photos though. Are both the same and could you produce a sketch (block diagram would do) of what you see with the incoming wires, outgoing wires and what's connected to each? That would be a help.

[istari_knight]

I can tell you, the unusual looking ferrite encased inductors were commonly used in lots of Sony speakers during the 90's... I've never seen them anywhere else. And yeah, Solen is unusual... often either a BC Comp MKT or Ansar supersound.

Definitely worth asking Rega for a crossover circuit diagram.

[Karma Train]

Right then, I've now had the back off the second speaker, had a look at the crossover and I can confirm that it's exactly the same as the first one, pictures of both below....

http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00229_zps7a4a48e0.jpg

http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00235_zps693f2ffb.jpg

I was also asked if I could supply a block diagram showing the wiring, below is my attempt, please don't laugh too loudly....

http://i254.photobucket.com/albums/hh114/Karma_train/Ebay%20Uploads/DSC00239_zps503191c4.jpg

I will try and speak to Simon again at Rega and see if an original crossover schematic exists plus I'll also see if I can send him a photograph to verify if the crossover has been altered in any way or whether it's still in its original condition.

A couple of people have suggested that the crossover has been altered and although I know absolutely nothing about electronics, my educated guess is that they haven't been tampered with.
I say this for a number of reasons, all pertaining to the 'fix' I mentioned in my original post, I will copy and paste it again below....

Rega XEL tweeter modification.
I already wrote a review some time ago. What I would like to propose here is a quite simple cross-over modification that will smoothen the treble and improve the transition midrange/treble. There is no change of the cut-off frequency, only a slightly different roll-off curve and a slight attenuation in the mid-treble range.
Here goes:
Change the 3.3 microfarad capacitor for a 3.0 microfarad capacitor and add another 47 microfarad capacitor after the inductance pole, just before the tweeter. What you get is a 3° order cross-over instead of 2° order: 3.0 microfarad in series, then a 0.39 mH inductor in parallel to another capacitor of 47 microfarad in series with the tweeter.
I tried at least 10/15 modifications before coming to this conclusion that attenuates the tweeter just where it needs to, without compromising the midrange and keeping the correct phase offset for a still very focused sound in the woofer / tweeter transition. The treble gets less harsh and more refined, especially in combination with my new TAC 34 tube amplifier.
Whoever drives these speakers in a rather reverberating room (wooden floor, long flat uncovered walls, windows) and struggles with a treble that's a bit too crisp, should try this quite easy modification.

Firstly as I mentioned before, these speakers sound fantastic in every respect other than the overly sharp tweeter, so there's no question that all drivers are working as intended, the mid-range is absolutely superb, so to my mind at least this suggests that they are wired entirely correctly.
With regard to the modification above, as you can see, the writer also complains about the sharpness of the tweeter and indeed it must have troubled him greatly if it caused him try 10 different modifications before he found one which worked. Also you might notice that the first step in his fix involves "changing the 3.3 microfarad capacitor for a 3.0 microfarad capacitor" and as you can see in my photographs, the capacitor installed is indeed rated at 3.3uF, to me at least, this suggests that what greeted him when he examined the crossover is the same as that which greeted myself.

I really wish I possessed the skill to try this guy's modification as he seems convinced that it solves the problem but sadly I wouldn't have a clue where to start :(

Anhoo, hopefully this further information throws a bit more light on the subject and once again, thank you most kindly for all your help.

Charlie.

[Reffc]

Good job Charlie

If that sketch is anything to go by, those crossovers look to be wired up incorrectly or some wiring isn't visible because there's not output voltage to the bass units if wired that way.

I suspect that there may be some connections not shown, perhaps the other side of the board. There should be a connection from the input black wire direct to the output black wire without passing through the small inductor to the bass units negative terminal. There should also be another connection from the input red wire to the large (long ferrite cored) inductor. The notch filter is the odd looking inductor to the left together with the resistor and square caps. They should be connected after the main bass inductor to the series leg output, and wired in series with each other with the end leg connected to the negative terminal. The way it's wired up, it wouldn't work like that and it more resembles a baffle step correction filter in series with the main series bass inductor, albeit with a few connections missing!

Also, I would be wary of "fixes found on the interweb"

What that fix tells me is that the chap stumbled across something that worked for him without any form of measurement or calculation other than Mk1 earlugs. It wouldn't have taken 10 attempts to "find" something that worked with proper measurement/calculation. That approach is like shooting in the dark until something is stumbled across but even then, the summed response is still not understood as a result of the fix, other than it sounds better to him (which in fairness it probably does). I have no idea how the 47uF value was arrived at as for the crossover point chosen, the expected values for the inductor would be closer to 0.22mH and the 2nd cap at 9 or 10uF (if the tweeter is an 8 Ohm model). His description of attenuation is wrong. The tweeter is not strictly speaking "attenuated". Reducing the value of the capacitor raises the crossover point reducing summed response and 3rd order increases roll off steepness.

Changing from a 2nd order to a 3rd order without adjusting for phase changes will possibly work because the bass is 90 degrees phase and 3rd order will be 270 degrees or -90 (whichever way you look at it) so is more phase correct as the chap says. My guess is that the summed response of tweeter and woofer(s) peaked at crossover previously, or the harshness was due to tweeter resonance below crossover point (high distortion) summing with woofer response (needs a notch filter if that is the case but only measurement of the tweeter will confirm that). From your comments Charlie, something is amiss there as standard (if that is what it is). If you're handy with a soldering iron, you could try the "fix" above to see if you like it, but best bet if possible would be to undertake both acoustic and electrical measurement to see what was going on there before making any decisions.

The fix will not change the tweeter output level above crossover. It will only increase the steepness of the roll-off from 12dB/Octave to 18dB/Octave and shift phase by 90 degrees as well as lifting crossover by around 300Hz but if it works, it works.

[Karma Train]

Thank you most sincerely for your post Paul, that must've taken ages to type up (if you're typing is anything like mine that is) and so I really do thank you for your time and your consideration of the matter - just such a shame I barely understood any of it, like I said, my knowledge of electronics is on a par with knowledge of brain surgery, so come again! :scratch:
Today I spoke to Simon at Rega and his suggestion was to leave the crossover well alone and perhaps try fitting a pair of their new ZRR tweeter, the one they currently utilise in their RS5 and RS7 models, however, he couldn't say with any assurance whether the crossover would or wouldn't need tweaking after the change, so that looks like a bit of a non-starter. He also said that there were no longer any schematics available for the XELs.
I then spoke with Jerry at Falcon Acoustics and sent him a link to this thread, his conclusion was that he couldn't make head nor tail of the crossover either and therefore could offer no solution to the menace of overbearing sibilance.
All in all, I think I'm backing a total loser here, even if anything could be done to sort the problem out it would probably cost a fortune and as I'm not overly endowed with funds I just think that perhaps I should set about finding some speakers which do work in this room as opposed to fiddling about with these shrill monsters, the words flogging and dead horse are springing to mind here. It's a shame because they do have the makings of a wonderful sounding speaker but enough is enough I guess :(
Thanks again for all your help.

[Reffc]

Hi Charlie

don't give up on them just yet. I'm rather astonished that Rega think you can swap tweeters out for a different model without attending to the crossover. The fact that they didn't know I think sums it up! If you could get them over to me, and the fix is just a matter of sorting the wiring out (It may check out ok on inspection), I would do it free of charge for you. I can also test both speakers and provide the acoustic measurements so we can see more accurately what's happening here. It depends I guess what you paid for them whether the trip would be worth considering? My experience is that the tweeter is rarely to blame. Its more usually the crossover implementation as our hearing is most sensitive close to the point at which these cross over.

[technobear]

It sounds like a tweeter damping resistor might help here. You need a non-inductive wirewound resistor rated for 10 or 12 Watts and about 50 ohms. Place it in parallel with the tweeter. Cheap to try and may be all that is needed.

[Yomanze]

Square of toilet paper over the tweeter, some Sellotape, done. :)

[Yomanze]

Also, Google Jim Goulding and think about some wool anti-diffraction pads for your tweeters.

[Karma Train]

It sounds like a tweeter damping resistor might help here. You need a non-inductive wirewound resistor rated for 10 or 12 Watts and about 50 ohms. Place it in parallel with the tweeter. Cheap to try and may be all that is needed.

Huge apologies for not responding to your kind, helpful post but I thought that my thread was dead in the water, I still have the Rega XELs but have since bought an old pair of Rega Elas which are a bit more room friendly (size wise) and slightly less skin-shredding in the high treble department.

However, every time I look at the XELs stood over in the corner looking all forlorn I find myself hankering after their beautiful tight bass response and wishing that I could have made them acceptable to the local bat population, then today I see your post which has started me thinking again....D'oh!

I had a long, informative conversation with Paul at Reference Fidelity Components who told me that essentially all Rega crossovers are pants and whist not disputing that for a second I've always thought that there was possibly a simple 'electronic' solution to my problem. Of course rebuilding the crossovers would be the electronically 'pure' way to go but I'm not in a position to afford such major work.

I've seen the resistor solution mentioned numerous times on other forums and that's always struck me (although I know absolutely nothing about electronics) as the most simple and less bankrupting method to try and I'm up for giving it a go but I will need your help please.

Could you please provide me with a link to the actual resistors I would need to buy. I've searched for the items but as with anything on t'interweb the choice is dazzling and as I don't know my resistors from my capacitors, I really have zero idea of which ones to obtain.

Secondly, could you please also tell me how I would 'wire them in parallel', in layman's terms, where I would physically put the resistor. If as I imagine in my dumb-ass way the recommended location might be from positive to negative tweeter terminals, then a further problem I might have is that the sodding tweeters are glued into the cabinets, according to Rega, they are only in there with a silicone gasket but I've tried to prise them out with zero movement thus far, however I will persevere if I have a viable reason to.

So, thanks again for your suggestion and if you could possibly provide me with a link and the further information I'd be more than grateful.

Charlie (not a very techno bear!)

[Karma Train]

Also, Google Jim Goulding and think about some wool anti-diffraction pads for your tweeters.

Massive thanks for your suggestion, I did lots of reading about this possible solution prior to posting on AOS but from what I can make out, the anti-diffraction pads, whilst possibly improving the imaging of the speakers would do little to affect the actual distortion or sibilance that the tweeters suffer from, I could be wrong on that, I am with most things but that was the thrust of what I gleaned from checking them out.

Charlie.

[YNWaN]

No, you are correct. Reducing edge diffraction won't alter the intrinsic behaviour of the tweeters. Depending on the off-axis output of the tweeter it may subjectively lower output, but this will be a relatively minor effect (this is not a criticism; I have used ant-diffraction rings on my speakers for many years).

[Karma Train]

No, you are correct. Reducing edge diffraction won't alter the intrinsic behaviour of the tweeters. Depending on the off-axis output of the tweeter it may subjectively lower output, but this will be a relatively minor effect (this is not a criticism; I have used ant-diffraction rings on my speakers for many years).

Blimey, I wasn't wrong about something....now there's a first :D

Possibly if the resistor tweak suggested by TechnoBear worked then the anti-diffraction pads might be an added improvement which I'm more than willing to try.
Somehow though, if I'm to ever live with these monsters, I have to stop them frightening dogs at 50 meters first of all!

[Reffc]

Bear in mind that a damping resistor whilst lowering tweeter output will also shift the crossover point and that in itself can raise other issues. At 50R, it will drop an 8 Ohm tweeter to 7 Ohms which has the same effect as lifting crossover point (other components remaining the same) by 4 or 500Hz and creating quite a dip across the upper mid band.

There's essentially two ways to tackle the problem:

1). have a mess about and try to tune things by ear until you start ignoring speaker problems and start listening to music (you have it right when that happens) or
2) have the things measured, find out IF there is a lift somewhere happening and tackle that more surgically with a one-off fix.

Just for the record, my comments on Rega crossovers were that most of the speakers using the same drive units shared the same crossover design, and that it is a design that can be improved upon (as with many mass produced speakers). The midbass units are actually nice little units on these.

[Yomanze]

Blimey, I wasn't wrong about something....now there's a first :D

Possibly if the resistor tweak suggested by TechnoBear worked then the anti-diffraction pads might be an added improvement which I'm more than willing to try.
Somehow though, if I'm to ever live with these monsters, I have to stop them frightening dogs at 50 meters first of all!

The anti-diffraction panels will smooth out the response though potentially making it less peaky. So whilst not a direct cure for sibilance still something to use that can really improve the quality of the top end. :)

[Karma Train]

For Reffc

Honestly I could cry, I really could.:(
First there is hope, then there is no hope, then there is hope and now....well you get the picture!
I just know these speakers would be wonderful if their shoutiness could be tamed, it's a bit like having a teenage child who is a considerate angel 6 days of the week but loves going out mugging on a Saturday! So much potential but he keeps stabbing himself in the foot.
I kind of understand what you're saying Paul even if I don't know what an 'R' is (I always thought in ohms myself) but as I'm incapable of fiddling about with the speakers (in an electronic way) myself, nor am I that flushed with cash that an expensive crossover rebuild would be a drop in the ocean, in layman's terms, I'm buggered!
And while we're talking in layman's terms, you describe Rega crossovers as 'a design that can be improved upon' which to my less than subtle way of thinking does indeed place them firmly in the pants drawer!
Oh well, I guess I'll have to stick with these miserable little Elas I bought (which have sadly reintroduced the 73Hz fart problem to this room although the treble isn't quite so terrifying) as it appears you either have to have cloth ears or inestimable wealth to run a pair of XELs....such a shame.
Thanks again, I really do appreciate your comment even though it is heartbreaking.

Charlie.

[Karma Train]

The anti-diffraction panels will smooth out the response though potentially making it less peaky. So whilst not a direct cure for sibilance still something to use that can really improve the quality of the top end. :)

Definitely worth a try then although I'm not sure where one would buy them. Just to experiment with the principle, do you think it would be possible to make my own? I was looking at this stuff on FeePay!

http://www.ebay.co.uk/itm/261476618247?_trksid=p2055119.m1438.l2648&ssPageName=STRK%3AMEBIDX%3AIT

[istari_knight]

You could always try what the pro's do with NS10M's... Tape or blutak a bit of tissue paper over the tweeters :D

http://www.gearslutz.com/board/attachments/so-much-gear-so-little-time/105129d1230618415-correct-tissue-ns-10s-ns10tissue.jpg

[Karma Train]

You could always try what the pro's do with NS10M's... Tape or blutak a bit of tissue paper over the tweeters :D

http://www.gearslutz.com/board/attachments/so-much-gear-so-little-time/105129d1230618415-correct-tissue-ns-10s-ns10tissue.jpg

Ah NS10s, the original shouty box....I actually tried sticking some acoustic foam over the tweets which didn't make a scrap of difference, bit low on boggit at the mo, so I might give that a try when the Andrex puppy has been bought new supplies :)

[walpurgis]

Failing the use of measuring equipment, my inclination would be to look online and try and find a response graph from a review or from the manufacturer. That may give you an idea of what the tweeter is doing and you may be able to use a notch filter if there's a peak or a simple 'L' pad circuit to add to the crossover if you just want to drop the tweeter output a bit. There are many free calculators that can be Googled up to give you component values to start with.

[Reffc]

Failing the use of measuring equipment, my inclination would be to look online and try and find a response graph from a review or from the manufacturer. That may give you an idea of what the tweeter is doing and you may be able to use a notch filter if there's a peak or a simple 'L' pad circuit to add to the crossover if you just want to drop the tweeter output a bit. There are many free calculators that can be Googled up to give you component values to start with.

I have a customer who lives nearby who owns a set of Regas with the same tweeter. If I can persuade him to drop off one them off, I'll do an acoustic measurement plot of the tweeter response on-axis and post it up on this thread. It should be close enough to figure out if there's something significant going on here (it uses the RFC modded crossover but that wont affect upper HF response).

[technobear]

http://www.maplin.co.uk/p/wirewound-10-watt-68-ohm-resistor-h68r

http://www.maplin.co.uk/p/wirewound-10-watt-47-ohm-resistor-h47r

Your local Maplin in Taunton may well have these (adjacent to Sainsburys car park in the town centre).

The effect on the crossover frequency is likely to be far less audible than the effect on tweeter resonances.

Yes, they should go across the tweeter terminals. If you can't reach them, you will have to identify the wires and put the resistors at the other end from one wire to the other. The important point is that no other components come between the resistors and the tweeter.


Oh, forgot to add: 47R = 47 ohms. If you see 4R7 that would be 4.7 ohms, etc. The R takes the place of the decimal point until you get up to 1000 ohms where a lower case 'k' takes over, e.g 1k8 = 1800 ohms.

[Gordon Steadman]

I'd try a very simple solution first. If that doesn't work it would be easy to reverse it. I had a balance problem with one pair of loudspeakers and so planned to drop the output of the tweeter with an 'L' pad as I had always been told you should. As I only had one of the resistors to put in parallel, I just added the one in series as a temporary measure on one of the speakers. Blow me down if I didn't prefer the sound of that one. OK it might not be the most scientific of methods but it solved my particular problem.

It might just fix yours for very little effort. Worth a try.

[istari_knight]

http://www.troester.org/ls/lpad.html

The Scanspeak D2008/851100 is 8ohms nominal BTW.

[Reffc]

Tweeter resonance is usually only an issue at tweeter Fs resonance point. A 50R isn;t enough to really put much of a dent in resonance output but will pull down tweeter response across the board as well as shifting crossover point. If it is a resonance issue, then a notch filter is the better way of dealing with it (and this wont affect crossover). All speculation though without knowing the impedance curves or acoustic response.

I've dug up my files on the tweeter impedance tests and have two different plots (early and later tweeters). The worst case, which I suspect this is, is for the tweeter with the higher resonance point and this is at approximately 1150Hz with just under a 10R impedance at resonance for a nominal 6Ohm tweeter (it may claim to be an 8 Ohm tweeter but it actually measures as a 6Ohm nominal). Arguably a typical response.

A notch filter using the following values will squash the tweeter resonance: L= 6.7mH C = 3uF R = 18R(total...so the actual resistor value will be 18R MINUS the DCR of the inductor).

Whether this will cure the harshness I dont know. As the crossover point is above 3KHz, and therefore 1.5 octaves above resonance, my gut instinct is that this may be a woofer issue (response towards cone creak up) or a summed response issue at or slightly above crossover. The tweeter on both the Juras I looked at didn't seem to have an audibly unpleasant hf lift and in fact seemed to lack a little sparkle higher up.

L-pad is the cheapest thing to try first of all, but you may find that just kills off whatever HF detail you have. Suck it and see.

[technobear]

The theory is all well and good but the proof is in the listening. I have fixed several pairs of speakers now with tweeter damping resistors and they sounded a whole lot better afterwards. I haven't tried one yet where I had to take it back out!

If the 47R does too much then try the 68R instead.

[fatmarley]

It sounds like a tweeter damping resistor might help here. You need a non-inductive wirewound resistor rated for 10 or 12 Watts and about 50 ohms. Place it in parallel with the tweeter. Cheap to try and may be all that is needed.

Sorry for bringing up an old thread but I just simulated this In LspCAD and It does indeed drop the tweeter level without affecting the crossover slope or phase (In my speakers at least). It would be well worth a try If you have too much treble.
If It were me I'd buy some 50ohm 100ohm and maybe 150ohm wirewounds to try across the tweeter terminals.

[Antinchip]

But I thank you for it Mat.

I have been troubled by excessive sibilance on the centre channel of my AV set-up for a while. I have changed a lot of kit of late in an effort to address this but must conclude that poor quality broadcast material on HDTV and SkyHD is the root cause. I am on the brink of buying yet another centre channel speaker to try but I had been wondering if there might be some tweak to the centre channel speaker, tweeter and/or crossover, that might ameliorate this.

The centre speaker is now a Tannoy Revolution RC (legacy model). How is the tweeter damping resistor mod performed? Is it a case of buying a 40p resistor from Maplin, removing the crossover and soldering the resistor in place or...? Any guidance would be appreciated. Many thanks.

[walpurgis]

A parallel or series resistor will affect the crossover characteristics.

Instead, use a simple 'L' pad or adjustable 'L' pad to attenuate tweeter output.

http://www.erseaudio.com/CrossoverCalculators/L-Pad-Attenuation

If you want a noticeable but minor cut in output, go for -6db.

[fatmarley]

A parallel or series resistor will affect the crossover characteristics.

Instead, use a simple 'L' pad or adjustable 'L' pad to attenuate tweeter output.

http://www.erseaudio.com/CrossoverCalculators/L-Pad-Attenuation

If you want a noticeable but minor cut in output, go for -6db.

6db Is a massive drop in level! You only need to drop 1db across the whole of the tweeter output to notice quite a large difference. This tweak would work like an L-pad in a lot of situations because anything in series with the tweeter would have some resistance, so no need to add the series resistor.

I'm only using my own speakers for the sims, so I can't guarantee It wont affect the crossover slope In other designs but It's cheap and easy to try (my guess Is it would work the majority of the time). If the problem Is cone breakup from the woofer/mid or the impedance peak of the tweeter then obviously It wont work.

[walpurgis]

6db Is a massive drop in level! You only need to drop 1db across the whole of the tweeter output to notice quite a large difference.

Did you say that as a joke?

Some people may just perceive changes of 1db on steady mid tones. But a 1db drop is inaudible on music and 6db is only slight. It's reckoned that steps of around 3db are the average threshold of audible sound level change.

[fatmarley]

Did you say that as a joke?

Some people may just perceive changes of 1db on steady mid tones. But a 1db drop is inaudible on music and 6db is only slight. It's reckoned that steps of around 3db are the average threshold of audible sound level change.

Not a joke at all.

I design my own speakers from scratch using passive crossovers, so know what a db sounds like over the whole bandwidth of the tweeter.

[Macca]

I disagree - My preamp works in 3dB steps, and 3dB is a noticeable difference and 6dB is a lot of difference.

Sibilance always gets blamed on the tweeter when it is rarely the problem, nor is the source material. I'd look at the amplification and cabling first.

[walpurgis]

http://www.audioholics.com/room-acoustics/human-hearing-amplitude-sensitivity-part-1

[57charles]

Shove a thick woollen sock over the tweeter.

[Macca]

http://www.audioholics.com/room-acoustics/human-hearing-amplitude-sensitivity-part-1

Over the years I've had my hands on plenty of gain controls and EQ sliders and often experimented with minimums. My experience has been when real music (as opposed to pink noise, burst tones, etc ) is played back in real (as opposed to laboratory) settings that the minimum level differences I've been able to discern is in the range of .75 dB to 1 dB.

This held true whether I was in a 3-chair living room experimenting with an EQ or a 3000-seat venue where I have enough power at my fingertips to light up a small city.

[DSJR]

I could just hear half a db on one channel recently when two otherwise identical excerpts were compared using headphones. The half db lift on one channel was all that was done and nothing said until after the experiment was concluded. Not everyone could hear any difference though.

[struth]

I can barely hear the phone ring now :wheniwasaboy:

[walpurgis]

Pardon? :)

[Macca]

I think we have our explanation for 6dB not being much. Bloody old codgers. I bet you think you've got it on at background levels when in fact they can hear it in the next parish.

[walpurgis]

I think we have our explanation for 6dB not being much. Bloody old codgers. I bet you think you've got it on at background levels when in fact they can hear it in the next parish.

Saucy young sprout!!:lol:

[rdpx]

From the title I thought this was going to be a Donald Trump thread.

[emoji1]

Sent from my ZX81

[fatmarley]

But I thank you for it Mat.

I have been troubled by excessive sibilance on the centre channel of my AV set-up for a while. I have changed a lot of kit of late in an effort to address this but must conclude that poor quality broadcast material on HDTV and SkyHD is the root cause. I am on the brink of buying yet another centre channel speaker to try but I had been wondering if there might be some tweak to the centre channel speaker, tweeter and/or crossover, that might ameliorate this.

The centre speaker is now a Tannoy Revolution RC (legacy model). How is the tweeter damping resistor mod performed? Is it a case of buying a 40p resistor from Maplin, removing the crossover and soldering the resistor in place or...? Any guidance would be appreciated. Many thanks.

Yep, those cheap ceramic wirewounds from Maplin are fine. You can either remove the tweeter and solder one end of the resistor to the positive terminal and the other end to the negative terminal (the terminals on tweeters are notoriously flimsy and very easy to overheat, so be very careful If you do it this way) , or you can find the crossover and put the resistors across the tweeter output (Remove the tweeter and use the wire colour as a guide or use a multimeter set to continuity to find the correct wires to solder across).

The smaller the resistor value, the more attenuation you'll get - Try a 50ohm first and if it's too much try 100ohm.

[Antinchip]

Matt, thanks for the explanation. Not sure I'm up to the task but will have a look at the innards. Woolly sock treatment might be the place for me to start!

Martin, thank you for your observations. I went through an extensive process of elimination swapping the kit;

Sources - Bluray, SACD player, HDTV and SKYHD.
Amplification - Swapped between Arcam AVR300 receiver and Primare SP31 pre/processor with three different power amps including Croft 7, Amptastic and Primare.
Speaker - Swapped between Castle Keep, Q Acoustics Q1010i, JPW Mini Monitor Gold, Tannoy Revolution RC.
Various interconnect cables including Mark Grant, Canare and Ecosse.
Various speaker cables in both single and bi-wire configuration including Cable Talk, Avondale Black Link, Van Damme and Ecosse.
I have sorted all cabling to ensure that power and signal cables are separate and kept AC and DC on separate sockets and distribution blocks.

In any of the above combinations sibilance was present on HDTV and SKYHD but not the others. The issue is present on broadcast material and is variable in extent. The issue does not occur on downloaded material.

The configuration that I have settled with is; Primare SP31 pre/processor, Ecosse The Composer IC, Primare A30.3 power and Tannoy Revolution RC centre speaker single wired with Ecosse MS 2.15 solid core speaker cable. I have just added an Isotek EVO3 Mira mains conditioner for the HDTV (Panasonic GT30B Viera plasma) and SKYHD box and, shock horror, felt this helped with both sound and picture quality! This is the most satisfactory combination and the instances of sibilance are less frequent and less severe but still undesirable and a little irritating. Front and two channel duties are performed by a pair of Croft 7R mono blocks feeding Tannoy Turnberry SE's. Overall the sound can be superb, Westworld last night was most enjoyable!

My conclusion is that broadcast sound quality is variable and that the system is revealing of the deficiencies. Hence my thought of a "line of least resistance approach" (pun not intended) to try to treat the symptom in an effort to moderate the occasional intrusion of sibilance. Otherwise I will just have to put up with it as I am not inclined to change all the kit yet again.

Thank you all for your various comments and the discussion is interesting in any event. Cheers.