Its funny you should say you like the original the best. I made two pairs, and the one I like better is the original, without the notch filter. But not being satisfied, I've put the SS9500 in the other pair (which are mine, the others are not) using Troels' crossover and am now ready to tweak these.
I guess my point is, if you are satisfied stop there. My friend with the other pair is completely satisfied (and thinks I'm some sort of wizard with this). You will only drive yourself crazy thinking there is one more tweak you can do. Plus, it starts to get expensive...
BTW, does anyone have any ideas for lowering the xover point. With the 9500, I can go a little lower, which would put a little less strain on the 8535 I would think. And what does everyone think about the damar varnish on the dust cap? Does it do the job of taming the (alleged) breakup? I treated the 8513 tweeters with it and it did tame the sibilance some, but obviously not enough as I replaced the tweeters after. The tweaking goes on and on...
I guess my point is, if you are satisfied stop there. My friend with the other pair is completely satisfied (and thinks I'm some sort of wizard with this). You will only drive yourself crazy thinking there is one more tweak you can do. Plus, it starts to get expensive...
BTW, does anyone have any ideas for lowering the xover point. With the 9500, I can go a little lower, which would put a little less strain on the 8535 I would think. And what does everyone think about the damar varnish on the dust cap? Does it do the job of taming the (alleged) breakup? I treated the 8513 tweeters with it and it did tame the sibilance some, but obviously not enough as I replaced the tweeters after. The tweaking goes on and on...
hello! i'm lost with all this filters!!
Which one his the best??? The v3 or the v6 ???
i ve never heard the real 2.5, so, i don't matter to have this 'clone' sound different.
thx for your help!
Which one his the best??? The v3 or the v6 ???
i ve never heard the real 2.5, so, i don't matter to have this 'clone' sound different.
thx for your help!
and about the replacement of the tweeter by the 9500.
the 8535 must be damar coated???
and the 9500 must have is ferrofluid removed??? (its write on the Troels PDF)
the 8535 must be damar coated???
and the 9500 must have is ferrofluid removed??? (its write on the Troels PDF)
an other question about the v6 filter.
it is write that resistors must be 5W.
is there a problem by using MOX 4 watts??
(and sorry for my poor english ;( )
it is write that resistors must be 5W.
is there a problem by using MOX 4 watts??
(and sorry for my poor english ;( )
Have you read Troels' latest work?
http://members.chello.se/jpo/The_2_5_clone_papers_by_Troels_Gravesen_v3.pdf
I don't think you can find anything more expert in advice than this. I would say that if building one today, your choice would be between the V6.1 and V6.2 crossover.
http://members.chello.se/jpo/The_2_5_clone_papers_by_Troels_Gravesen_v3.pdf
I don't think you can find anything more expert in advice than this. I would say that if building one today, your choice would be between the V6.1 and V6.2 crossover.
😀 yes 4w are ok to use [ not sure what mox is?] but it will be allright AFAIK..... cheers Tomcat
thanx for the enlightenment, after all I'M JUS' AN IGNORANT PUMPKIN so I wouldnt have known what mox meant for all I knew it could have been a different strain of pox😀 cheers Tomcat
Hello 2.5 cloners,
The 2.5 Clone Papers and other speaker projects are now available from:
http://home1.stofanet.dk/troels.gravesen/
regards
Troels
The 2.5 Clone Papers and other speaker projects are now available from:
http://home1.stofanet.dk/troels.gravesen/
regards
Troels
Disabled Account
Joined 2003
Scanspeak Ref Plus vs Proac 2.5 clone
The 2.5, being a 2 way will always be torn between reproducing the mid and bass freq. The single Scanspeak 8535 driver is great, but is stretching very hard to reproduce the freq spectrum between mid and bass. As a result, it will never be 100% accurate.
The 2.5 design is fabulous, and will beat all 2 way speakers in its ability to reproduce clean, musical and big sound stage. High is accurate, with plenty of headroom. Bass is grounded and deep. But mid could be better, its a notch above normal human voice. On a good amp system, all weaknesses are revealed.
However, its still cheap & easy to DIY!
Ref system:
Vac CPA1 Mk3 preamp, Graaf 100 poweramp.
Sold my diy proac after 12 months, because the voice reproduction is never grounded despite all tweaks....wires, caps,
After listening to the 3-way Proac 3.8, I am convinced the 3 way speaker is the way to go. It has Pristine mid, good bass, but Proac 3.8 's high is disappointing.
I came across the Scanspeak Reference Plus http://www.audiocomponents.nl/ after 4 weeks of searching late last yr.
The design and driver choice grabbed my attention immediately.
Was a difficult project due to the slanting/angled box, but the drivers recommended were the best from Scanspeak. Revelator 9900 tweeter, plus 8545, 8546 for mid & bass.
Completed the speaker at around A$1800 (US$1400). Parts from Parts Express (US) and WES Comp (Aust).
The music from the Scanspeak Refplus is miles above Poac 2.5. It sounds more like the Sonus Faber Cremona, (at A$18,000).......for a mere A$1800 outlay plus 2 weekends in the garage.
My advise for Proac2.5 DIYer is to move on, and save yourself the aggro . The Scanspeak Ref is the best kept secret, Put the 3-way Proac 3.8 and DIY Scanspeak side by side, and the Proac 3.8 will lose out big time even before the Scanspeak ref has time run in.
Testing source:
Sheffield Labs: Lincoln Mayorca - Perfect Song , America
Norah Jones: Come away with me, Lonestar
Jacinta: Danny Boy 45 rpm
Linda Rosenthal: Oh My Stradivarius XRCD
Result: Big sound stage.
High: The Revelator 9900 is crisp, accurate, and non fatigue
Mid: Scanspeak 8546 Kevlar - fast, grounded, vocal in front and clean, with plenty of air, saliva and wetness. Yes its all there.
Bass: Scanspeak 8545 - Deep & tight, but much better extension than 8535 in Proac.
The freq shift - between the drivers from bass, mid to hi is totally velvety.
Good luck.
The 2.5, being a 2 way will always be torn between reproducing the mid and bass freq. The single Scanspeak 8535 driver is great, but is stretching very hard to reproduce the freq spectrum between mid and bass. As a result, it will never be 100% accurate.
The 2.5 design is fabulous, and will beat all 2 way speakers in its ability to reproduce clean, musical and big sound stage. High is accurate, with plenty of headroom. Bass is grounded and deep. But mid could be better, its a notch above normal human voice. On a good amp system, all weaknesses are revealed.
However, its still cheap & easy to DIY!
Ref system:
Vac CPA1 Mk3 preamp, Graaf 100 poweramp.
Sold my diy proac after 12 months, because the voice reproduction is never grounded despite all tweaks....wires, caps,
After listening to the 3-way Proac 3.8, I am convinced the 3 way speaker is the way to go. It has Pristine mid, good bass, but Proac 3.8 's high is disappointing.
I came across the Scanspeak Reference Plus http://www.audiocomponents.nl/ after 4 weeks of searching late last yr.
The design and driver choice grabbed my attention immediately.
Was a difficult project due to the slanting/angled box, but the drivers recommended were the best from Scanspeak. Revelator 9900 tweeter, plus 8545, 8546 for mid & bass.
Completed the speaker at around A$1800 (US$1400). Parts from Parts Express (US) and WES Comp (Aust).
The music from the Scanspeak Refplus is miles above Poac 2.5. It sounds more like the Sonus Faber Cremona, (at A$18,000).......for a mere A$1800 outlay plus 2 weekends in the garage.
My advise for Proac2.5 DIYer is to move on, and save yourself the aggro . The Scanspeak Ref is the best kept secret, Put the 3-way Proac 3.8 and DIY Scanspeak side by side, and the Proac 3.8 will lose out big time even before the Scanspeak ref has time run in.
Testing source:
Sheffield Labs: Lincoln Mayorca - Perfect Song , America
Norah Jones: Come away with me, Lonestar
Jacinta: Danny Boy 45 rpm
Linda Rosenthal: Oh My Stradivarius XRCD
Result: Big sound stage.
High: The Revelator 9900 is crisp, accurate, and non fatigue
Mid: Scanspeak 8546 Kevlar - fast, grounded, vocal in front and clean, with plenty of air, saliva and wetness. Yes its all there.
Bass: Scanspeak 8545 - Deep & tight, but much better extension than 8535 in Proac.
The freq shift - between the drivers from bass, mid to hi is totally velvety.
Good luck.
Strange that you like the reference plus so much. Most people here who have heard it, don't seem like it much. Not that they say it's a bad speaker, just not worth the money.
Continued from Post #72
After reading elsewhere about preference for the Chinese XO, I built another and spent a few days comparing it to V6.1 (note that my speaks have modded drive units as per Troels' recommendations).
I can certainly understand why some people prefer the Chinese XO - it is quite a bit livelier and faster sounding with more apparent detail. In comparison V6.1 is fuller sounding and more coherent through the midrange. As I said before though, it does lack some "air" and is perhaps a bit unexciting sounding. If I had to choose only one it would certainly be V6.1 because female vocalists are an uncomfortable experience when the warp factor is increased a bit.
Anyway, I wanted the best of both worlds so I analysed the response of the various crossovers and investigated the box tuning a bit more, with the following conclusions:
1) The box is too small for the SS 8535 resulting in about 3dB boost to the bass at 60Hz.
2) Proac deliberately designed 3 or 4 dB of boost at 2KHz into the XO.
3) Troels' treble XO has relatively less high freqs (above 6KHz) than the Chinese XO.
All we can do about 1) is too increase the port length a bit and tweak the box stuffing - giving a bit more of an EBS alignment.
2) I believe Troels reduced the mids level too much (my speaks measure with a 3 dB dip at 3KHz as well - both XOs). The 47R resistor in \\ with the 1.5mH inductor helps to damp the 2KHz peak but Troels idea to put 3R3 or so in series with the bass inductor is a better way to do this.
By increasing the mid and treble energy a bit the bass seems to tighten up and become more tuneful.
In summary here is my suggestion to get a response (and sound)half way between the Chinese and v6.1 XOs:
Starting with the Chinese XO...
Remove the 47R across the 1.5mH choke
Change the 8.5uF cap to 6.8uF (I actually used 6.9uF)
Add 3.9R (I actually used 3.3R) in series with the 6.8uF cap
No changes to the treble XO - make sure the 47R is across the inductor not the tweeter.
I have yet to measure the changes and do fine tweaking on the freq and phase response but listening tests so far have been favourable - IMHO it is undoubtedly better than the Chinese XO and I think it may well be better than V6.1.
To put this into perspective, most of the Proacs I have heard lack coherence and have a manipulated sound quality. Probably the most coherent are the Tablettes. So if you like the Proac sound I suggest you ignore me completely.
After reading elsewhere about preference for the Chinese XO, I built another and spent a few days comparing it to V6.1 (note that my speaks have modded drive units as per Troels' recommendations).
I can certainly understand why some people prefer the Chinese XO - it is quite a bit livelier and faster sounding with more apparent detail. In comparison V6.1 is fuller sounding and more coherent through the midrange. As I said before though, it does lack some "air" and is perhaps a bit unexciting sounding. If I had to choose only one it would certainly be V6.1 because female vocalists are an uncomfortable experience when the warp factor is increased a bit.
Anyway, I wanted the best of both worlds so I analysed the response of the various crossovers and investigated the box tuning a bit more, with the following conclusions:
1) The box is too small for the SS 8535 resulting in about 3dB boost to the bass at 60Hz.
2) Proac deliberately designed 3 or 4 dB of boost at 2KHz into the XO.
3) Troels' treble XO has relatively less high freqs (above 6KHz) than the Chinese XO.
All we can do about 1) is too increase the port length a bit and tweak the box stuffing - giving a bit more of an EBS alignment.
2) I believe Troels reduced the mids level too much (my speaks measure with a 3 dB dip at 3KHz as well - both XOs). The 47R resistor in \\ with the 1.5mH inductor helps to damp the 2KHz peak but Troels idea to put 3R3 or so in series with the bass inductor is a better way to do this.
By increasing the mid and treble energy a bit the bass seems to tighten up and become more tuneful.
In summary here is my suggestion to get a response (and sound)half way between the Chinese and v6.1 XOs:
Starting with the Chinese XO...
Remove the 47R across the 1.5mH choke
Change the 8.5uF cap to 6.8uF (I actually used 6.9uF)
Add 3.9R (I actually used 3.3R) in series with the 6.8uF cap
No changes to the treble XO - make sure the 47R is across the inductor not the tweeter.
I have yet to measure the changes and do fine tweaking on the freq and phase response but listening tests so far have been favourable - IMHO it is undoubtedly better than the Chinese XO and I think it may well be better than V6.1.
To put this into perspective, most of the Proacs I have heard lack coherence and have a manipulated sound quality. Probably the most coherent are the Tablettes. So if you like the Proac sound I suggest you ignore me completely.
This is a never ending story. As time goes by and you listen more and more I think we all hanker for something ‘different’ not necessarily better! I have mentioned amongst this thread that when I began my clone that my Response 1SC was very good and I wanted at least to match it.
Now I like the Proac sound and my R1sc was the reference to my listening so for those who like the Proac sound this may be interesting. I have gone through the whole gamut of tweaks and trials and swapped components around and moved them from here to there. The ultimate problem is of course we all have different equipment and rooms and preferences of how we like something to sound so what I end up with may not be what you like.
I have lived with the R1sc for about 8years (it has been upgraded through the different models) and I have been very happy with the sound at every point so, as it is all relative, if you like Proac in your setting you may find similar results as me with the clone.
At first I started with the unmodified drivers and the V6 cross over with an un-braced cabinet and Dacron filling. It might be unfair as the drivers hadn’t ‘run in’ but the sound was not very good at all and only about 60%, if that, of my 1sc. Very disappointing. I have now ended up with a coated dome on the tweeter, 1” foam damping (no Dacron) and a modified XO (the most interesting part).
It is possible to ‘kill’ the life in the speaker by over damping and a single layer of foam over 5mm of dead sheet is enough. I want to retry the Dacron but using more dense packing than before but things are so good at the moment I’m reluctant to ‘play’.
The XO is where I have notice most changes. Now I don’t profess to know anything about the art of xo design or understand what is going on about slopes, db per octave, phase or anything like that (despite Troles help) but I know what I like and my current system (re-listening to the 1sc) is about 10 – 15% better than the 1sc (subjective I know!).
With the above mods I revisited the Chinese XO. The V6 was too ‘flat’ for my tastes and ‘boring’ compared to my reference and in my room (quite heavily furnished). So looking at Al.M’s earlier comments I replaced the 47R across the inductor moved the tweeter 47R across the inductor and not the tweeter used 3.3 uf (as per the R1sc) instead of the 3.9uf, used a 5.6R attenuating resistor (8R was too much for my ears), upped the 6.8uf to 7.8 in the woofer (although I may see what it is like back down to 6.8uf) and removed the 3.3R which was in series with the cap in the LP section. Inserted the notch filter then took it out.
Now the interesting part and this may seem totally illogical but I’ve checked it several times with the same result, the make of components I used had a significant effect on the sound. I first used a set of cheep components but finalising the circuitry I thought I would use better quality components and purchased Solen capacitors and mox resistor (metal oxide, to replace the wire wound ceramic cased type).
The sound was significantly more detailed but I now had some very harsh frequencies which grated and made some music too brittle. I think others have mentioned this. However this brittleness wasn’t there before! So after changing components and modifying the tweeter with only minor effect, I started looking at the components. Now I found and many may disagree, that Solen capacitors with mox resistors caused the problem. Replacing the resistors with the wire wound partially resolved it but I felt I lost some ‘detail’ or presence or life in the sound. So I replaced the Solen caps in the HP section with SCR caps (similar price to the Solen) from a French company purchased through Wilmslow Audio. Using the MOX resistors suddenly the harshness was virtually gone and yet the imaging depth sweetness, detail and life were all there (enhanced when placed on spikes).
It was sounding perfect apart form some emphasis on some records of S’s and T’s. Now I think I’m a bit over critical of this (listening too hard) because my1sc (I’m thinking of changing the components in those) are similar and so were many units at the recent Bristol HiFi Show (I specifically listened as a comparison). At that show in fact Proac still caught my attention above most others, at any price (Even the JM Labs I found quite hard, harsh and fatiguing and what is the cost of those!).
As a total novice I owe a great deal to Troels who has been very helpful as I built this, and to the contributors of this thread. At times I felt the project hardly worth the money but when you get to the sound you like, it is everything that has been said of it. If you are thinking of building one - do it. All the things about XO and filling etc mean that you can change the sound to suit your ears in your home and if you move or what something different it only means a change in value of a cap or resistor or internal cable or moving a component.
I’m now considering the Dacron filling and maybe a 6R (8R is still too much) or there abouts instead of the 5.6R and this may finally tune the system to perfection. But at the moment, for my ears, in my room the speakers are perfect – at least for the time being!
Now I like the Proac sound and my R1sc was the reference to my listening so for those who like the Proac sound this may be interesting. I have gone through the whole gamut of tweaks and trials and swapped components around and moved them from here to there. The ultimate problem is of course we all have different equipment and rooms and preferences of how we like something to sound so what I end up with may not be what you like.
I have lived with the R1sc for about 8years (it has been upgraded through the different models) and I have been very happy with the sound at every point so, as it is all relative, if you like Proac in your setting you may find similar results as me with the clone.
At first I started with the unmodified drivers and the V6 cross over with an un-braced cabinet and Dacron filling. It might be unfair as the drivers hadn’t ‘run in’ but the sound was not very good at all and only about 60%, if that, of my 1sc. Very disappointing. I have now ended up with a coated dome on the tweeter, 1” foam damping (no Dacron) and a modified XO (the most interesting part).
It is possible to ‘kill’ the life in the speaker by over damping and a single layer of foam over 5mm of dead sheet is enough. I want to retry the Dacron but using more dense packing than before but things are so good at the moment I’m reluctant to ‘play’.
The XO is where I have notice most changes. Now I don’t profess to know anything about the art of xo design or understand what is going on about slopes, db per octave, phase or anything like that (despite Troles help) but I know what I like and my current system (re-listening to the 1sc) is about 10 – 15% better than the 1sc (subjective I know!).
With the above mods I revisited the Chinese XO. The V6 was too ‘flat’ for my tastes and ‘boring’ compared to my reference and in my room (quite heavily furnished). So looking at Al.M’s earlier comments I replaced the 47R across the inductor moved the tweeter 47R across the inductor and not the tweeter used 3.3 uf (as per the R1sc) instead of the 3.9uf, used a 5.6R attenuating resistor (8R was too much for my ears), upped the 6.8uf to 7.8 in the woofer (although I may see what it is like back down to 6.8uf) and removed the 3.3R which was in series with the cap in the LP section. Inserted the notch filter then took it out.
Now the interesting part and this may seem totally illogical but I’ve checked it several times with the same result, the make of components I used had a significant effect on the sound. I first used a set of cheep components but finalising the circuitry I thought I would use better quality components and purchased Solen capacitors and mox resistor (metal oxide, to replace the wire wound ceramic cased type).
The sound was significantly more detailed but I now had some very harsh frequencies which grated and made some music too brittle. I think others have mentioned this. However this brittleness wasn’t there before! So after changing components and modifying the tweeter with only minor effect, I started looking at the components. Now I found and many may disagree, that Solen capacitors with mox resistors caused the problem. Replacing the resistors with the wire wound partially resolved it but I felt I lost some ‘detail’ or presence or life in the sound. So I replaced the Solen caps in the HP section with SCR caps (similar price to the Solen) from a French company purchased through Wilmslow Audio. Using the MOX resistors suddenly the harshness was virtually gone and yet the imaging depth sweetness, detail and life were all there (enhanced when placed on spikes).
It was sounding perfect apart form some emphasis on some records of S’s and T’s. Now I think I’m a bit over critical of this (listening too hard) because my1sc (I’m thinking of changing the components in those) are similar and so were many units at the recent Bristol HiFi Show (I specifically listened as a comparison). At that show in fact Proac still caught my attention above most others, at any price (Even the JM Labs I found quite hard, harsh and fatiguing and what is the cost of those!).
As a total novice I owe a great deal to Troels who has been very helpful as I built this, and to the contributors of this thread. At times I felt the project hardly worth the money but when you get to the sound you like, it is everything that has been said of it. If you are thinking of building one - do it. All the things about XO and filling etc mean that you can change the sound to suit your ears in your home and if you move or what something different it only means a change in value of a cap or resistor or internal cable or moving a component.
I’m now considering the Dacron filling and maybe a 6R (8R is still too much) or there abouts instead of the 5.6R and this may finally tune the system to perfection. But at the moment, for my ears, in my room the speakers are perfect – at least for the time being!
Disabled Account
Joined 2003
I am unbelievably lucky that my clones have not required any tinkering or modifications.
I used Al M's cabinet specifications but Troel's 9500 tweeter and crossover.
I have also made A/B comparisons with the original clone. It was brief (1 hour), and I don't want to do a song and dance just yet, but the preliminary verdict looks very promising.
I have not heard the V6, V6.1 or V6.2, so cannot comment on these. But with the 9500 and approprite XO, the midrange is improved, but without compromising too much of crispness and interesting treble.
I'll stick my head out here and say "For builders wanting to avoid modifying their units, (and don't mind extending the budget)- I believe this is the best option"
I won't replicate what I wrote about the sound, because what I said here still holds.
I used Al M's cabinet specifications but Troel's 9500 tweeter and crossover.
I have also made A/B comparisons with the original clone. It was brief (1 hour), and I don't want to do a song and dance just yet, but the preliminary verdict looks very promising.
I have not heard the V6, V6.1 or V6.2, so cannot comment on these. But with the 9500 and approprite XO, the midrange is improved, but without compromising too much of crispness and interesting treble.
I'll stick my head out here and say "For builders wanting to avoid modifying their units, (and don't mind extending the budget)- I believe this is the best option"
I won't replicate what I wrote about the sound, because what I said here still holds.
I have played a lot with various crossovers, including the original XO and Troels mods. None of them really satisfied me, so started to tweakening a new layout.
The idea is coming from Loudspeaker Lab 2, so that I use two different coils in series for the baffle step compensation and for the first coil in the LP filter. For the first sight it might look a bit complicated or non-sense but works for me.
I think with this XO the midrange is considerably better and the sound lacks the boxy sound which sometimes characterize the clone.
The best thing of all that you can set the BS step compensation very easily, by changing the resistor value only. I haven't heard this mod on other clones, so some feedback would be welcome.
There are some minor mod in the HP section as well. This is to flatten the 8513 raising response and also changes the slope a bit.
Please feel free to try and comment.
Zozo
The idea is coming from Loudspeaker Lab 2, so that I use two different coils in series for the baffle step compensation and for the first coil in the LP filter. For the first sight it might look a bit complicated or non-sense but works for me.
I think with this XO the midrange is considerably better and the sound lacks the boxy sound which sometimes characterize the clone.
The best thing of all that you can set the BS step compensation very easily, by changing the resistor value only. I haven't heard this mod on other clones, so some feedback would be welcome.
There are some minor mod in the HP section as well. This is to flatten the 8513 raising response and also changes the slope a bit.
Please feel free to try and comment.
Zozo
Attachments
Bricolo,
If you want me to answer this question, I think the answer is clear : because we want to keep the original design and stick as close to it as possible.
Active BS compensation is great, but I wouldn't call it after a 2.5 clone. Even with the XO mods I am in doubt if it is.
Zozo
If you want me to answer this question, I think the answer is clear : because we want to keep the original design and stick as close to it as possible.
Active BS compensation is great, but I wouldn't call it after a 2.5 clone. Even with the XO mods I am in doubt if it is.
Zozo
But who cares? (maybe you do...)
The crossover is already different
The drivers are the standard versions (and not the custom ones that Proac seems to use)
They're not 100% proacs clones, so, why not going a little further. Especially if it's for optimising them 😉
The crossover is already different
The drivers are the standard versions (and not the custom ones that Proac seems to use)
They're not 100% proacs clones, so, why not going a little further. Especially if it's for optimising them 😉
😀 hi zozo, szia kedves magyar testvérem, ha ez neked jó akkor ne törÖdjél a masok véleményukrÖl , most angolul fojtatom- as a very famous person once said; critics are like eunuchs, they can tell you how to do it but can't do it themselves [ H. COHN: former head of columbia pictures{ deceased}] cheers from down under... OZ T.C
- Status
- Not open for further replies.