New ProAc
Last week end I heard the new ProAc 2.5 and they were rubbish. At the Heathrow show they had about the worst sound... hope it has something to do with the room. The boxes sounded thin and flimsy and that was reflected in the sound... they have new unots as well as has most of the speakers... best sound.... the big Quad stats....
Michael
Last week end I heard the new ProAc 2.5 and they were rubbish. At the Heathrow show they had about the worst sound... hope it has something to do with the room. The boxes sounded thin and flimsy and that was reflected in the sound... they have new unots as well as has most of the speakers... best sound.... the big Quad stats....
Michael
There are no new 8535 just production variations. Scan Speak let the membranes dry in air without any pressure. That result in variations. Some will be more compact then others, but the weight is the same. The exact same quantity of coating is applied. This doping will suck in differently because of the differing compactness of the membranes. The end result is the same but they look a bit different. Ask Troels he knows this. I don´t know why he persists that there are a new 8535 driver. It is nonsens.
I agree with Wcoil that what Troels has observed are probably batch variations among thousands of driver units. I have seen about 16 drivers in the last 2 years and they all varied slightly in visual appearance with some appearing to be more coated than others by as much as 20% from visual appearance. Judging from the drivers Troels shows pictures of both older and newer ones are different to the ones I have seen. His older diver looks almost like a paper dry coating that I have never seen, and his newer "heavily coated driver" appears to be at the extreme end of appearing too coated.
Today I saw TKTran's drivers which have number 20,000 production numbers, compared to 6 older drivers I still have with 13000 numbers and his are slightly darker coated but basically same as others I have seen.
I think we all appreciate Troels work in offering a alternatives to the design, however for me the sound is not to my tastes. The following is a cut and paste of my comments:
I had the xo lying outside the cabinets so the many minor mods could be done quickly. Firstly, started with the latest suggestions in article 4 and your modified tweeters but no damar coating on the woofer, and then worked back to the standard Jacq mod xo on the website minus the LCR circuit with both tweeters. Equipment used were Spark KT88 50W class A valve amp, Allegro valve preamp, Pioneer PD91 CD (old but battleship class unit), Nordost flatline Gold biwired speaker cable, Tara Labs Prism interconnects. This equipment list is not in the big dollar league but much more revealing than most mid to high-end solid state or integrated and many average to good valve or solid state amp combinations I have heard. The Spark is not well known but has a particular ability to produce very clean treble, project midband/imaging and has great bass extension and the overall character is lean for a valve amp, and the other components are similar in character matched to a warm sounding clone. In addition, I have high quality 0.1uf metal foil capacitors (Aeon) in the clone tweeter xo to lift treble extension and clarity, which is probably making them sound the more extended than many other clones. CDs I used to test were Eva Cassidy, Chris Isaac and Robert Michaels flamenco.
For the benefit of others reading this, Article 4 xo mods include tweeter (HP) at 8 ohm series resistor, 3.9uf replaced 3.3uf cap; woofer (LP) 47 ohm R removed, 3.9 ohm R joined to 6.8uf cap. The treble and overall sound is still similar to the standard xo but considerably more neutral with the treble intensity much reduced by about 20% and some ordinariness in the bass response. My earlier comments about the standard 8513 tweeter versus the modified unit still stand. The modified tweeter is more neutral and technical sounding, but unfortunately for my expectations the life has been knocked out of the sound. It is still a worthwhile option for those who find the treble zing/sibilance a significant issue, although it has not removed all sibilance and the treble balance is still generally on the brighter side. I do not think the sound performance of the tweeter has been enhanced as I perceive that about 10% high frequency extension is lost, imaging is flatter and detailed instruments do not decay as nicely. Sibilance is more controlled but at the cost of other beneficial factors (“you can wash your face in these speakers and they will never dry out” Stereophile 1996). The over-all sound departs significantly from the Proac 2.5 (models from 1996 to 2001) character that I am familiar with. Gone is the richness, texture and flow that defines a Proac 2.5 from other speakers. The balance in my room is not ideal as the reduced treble intensity appears to have resulted in a more even and filled in mid range but now exposes some bass boom and nasality on some voices that were not as evident before.
On the other modifications removal of the 47 ohm R in the woofer xo has a small but subtle and noticeable effect on the coherence and matching with the tweeter response. With the 47 ohm R removed, bass seems a less well defined, expansive, textured and extended and less impressive. Use of this 47 ohm R with the other shunted 3.9 or 3.3 ohm RC appears to help, although not sure what this does electrically.
The use of the tweeter 3.9uf capacitor in place of the 3.3uf does not appear to improve anything except for maybe some perceived sibilance. The 3.9uf cap causes some blurring of the upper mid and lower treble where instruments like guitars, voices and clapping (Eva Cassidy track 4) are less well defined and coherent with much less attack. The brighter sound of the unmodified tweeter has a more positive effect on the speaker as guitars, wind instruments and voices sound much more alive, however this can lead to some sibilance on brighter material and equipment sources.
In my room I found the use of tweeter 8 ohm R instead of 5 ohm is beneficial in reducing some sibilance while still retaining much of the Proac magic and enough attack. Throughout my listening sessions, I have been having difficulty with the subdued treble response of the modified tweeter and the fact that the sound is now noticeably much less impressive with all the final mods suggested in from Articles 1-4. Use of the modified tweeter and 3.9uf HP capacitor, LCR and removal of the 47 ohm R is a dangerous exercise if you do not want the Proac magic to be lost. By reducing too much treble intensity the entire complexion of a lush, textured (technicoloured) and attacking sound is affected and I am no longer impressed in what the design can do best for the lessor benefit of sibilance control on some tracks.
In the end I think one has to accept the limitations of the design (like almost every other speaker) in the very good but not super high end quality, cost and ability of the drive units involved. The sound results derived from the driver cost, construction and model range ratio far exceeds the what could be expected from a floorstanding 2 way design (well done Stewart Tyler, if you need any more accolades) and that of many other DIY or commercial designs within the same cost, context and class. Therefore, while the efforts of many good people to improve on the design are admirable, we should not lose sight that attempting to add perfect neutrality is probably not the answer to this design when more of the best attributes are lost along the way. Troels has mentioned the use of better tweeters like the Hiquhon and Scan Speak 9500, the Seas T25-002 Millennium would also be a good match as it is more detailed than the 8513 yet has the lushness that the 9500 is missing.
I noticed on Troels article pictures that he may not have used spikes. I have found that spikes on the recommended plinth dramatically helps the voicing of the speaker by reducing excess bass, clearer midrange and attack. This is as important as using the correct amount of stuffing.
Sorry to disagree but these are my impressions after using the modified tweeter in all situations over the last several days. After having owned original Proac 2.5s several years ago I have to conclude at this point in time that the various mods are not, for me a substitute to perfect the design if you lose the speakers best aspects.
Today I saw TKTran's drivers which have number 20,000 production numbers, compared to 6 older drivers I still have with 13000 numbers and his are slightly darker coated but basically same as others I have seen.
I think we all appreciate Troels work in offering a alternatives to the design, however for me the sound is not to my tastes. The following is a cut and paste of my comments:
I had the xo lying outside the cabinets so the many minor mods could be done quickly. Firstly, started with the latest suggestions in article 4 and your modified tweeters but no damar coating on the woofer, and then worked back to the standard Jacq mod xo on the website minus the LCR circuit with both tweeters. Equipment used were Spark KT88 50W class A valve amp, Allegro valve preamp, Pioneer PD91 CD (old but battleship class unit), Nordost flatline Gold biwired speaker cable, Tara Labs Prism interconnects. This equipment list is not in the big dollar league but much more revealing than most mid to high-end solid state or integrated and many average to good valve or solid state amp combinations I have heard. The Spark is not well known but has a particular ability to produce very clean treble, project midband/imaging and has great bass extension and the overall character is lean for a valve amp, and the other components are similar in character matched to a warm sounding clone. In addition, I have high quality 0.1uf metal foil capacitors (Aeon) in the clone tweeter xo to lift treble extension and clarity, which is probably making them sound the more extended than many other clones. CDs I used to test were Eva Cassidy, Chris Isaac and Robert Michaels flamenco.
For the benefit of others reading this, Article 4 xo mods include tweeter (HP) at 8 ohm series resistor, 3.9uf replaced 3.3uf cap; woofer (LP) 47 ohm R removed, 3.9 ohm R joined to 6.8uf cap. The treble and overall sound is still similar to the standard xo but considerably more neutral with the treble intensity much reduced by about 20% and some ordinariness in the bass response. My earlier comments about the standard 8513 tweeter versus the modified unit still stand. The modified tweeter is more neutral and technical sounding, but unfortunately for my expectations the life has been knocked out of the sound. It is still a worthwhile option for those who find the treble zing/sibilance a significant issue, although it has not removed all sibilance and the treble balance is still generally on the brighter side. I do not think the sound performance of the tweeter has been enhanced as I perceive that about 10% high frequency extension is lost, imaging is flatter and detailed instruments do not decay as nicely. Sibilance is more controlled but at the cost of other beneficial factors (“you can wash your face in these speakers and they will never dry out” Stereophile 1996). The over-all sound departs significantly from the Proac 2.5 (models from 1996 to 2001) character that I am familiar with. Gone is the richness, texture and flow that defines a Proac 2.5 from other speakers. The balance in my room is not ideal as the reduced treble intensity appears to have resulted in a more even and filled in mid range but now exposes some bass boom and nasality on some voices that were not as evident before.
On the other modifications removal of the 47 ohm R in the woofer xo has a small but subtle and noticeable effect on the coherence and matching with the tweeter response. With the 47 ohm R removed, bass seems a less well defined, expansive, textured and extended and less impressive. Use of this 47 ohm R with the other shunted 3.9 or 3.3 ohm RC appears to help, although not sure what this does electrically.
The use of the tweeter 3.9uf capacitor in place of the 3.3uf does not appear to improve anything except for maybe some perceived sibilance. The 3.9uf cap causes some blurring of the upper mid and lower treble where instruments like guitars, voices and clapping (Eva Cassidy track 4) are less well defined and coherent with much less attack. The brighter sound of the unmodified tweeter has a more positive effect on the speaker as guitars, wind instruments and voices sound much more alive, however this can lead to some sibilance on brighter material and equipment sources.
In my room I found the use of tweeter 8 ohm R instead of 5 ohm is beneficial in reducing some sibilance while still retaining much of the Proac magic and enough attack. Throughout my listening sessions, I have been having difficulty with the subdued treble response of the modified tweeter and the fact that the sound is now noticeably much less impressive with all the final mods suggested in from Articles 1-4. Use of the modified tweeter and 3.9uf HP capacitor, LCR and removal of the 47 ohm R is a dangerous exercise if you do not want the Proac magic to be lost. By reducing too much treble intensity the entire complexion of a lush, textured (technicoloured) and attacking sound is affected and I am no longer impressed in what the design can do best for the lessor benefit of sibilance control on some tracks.
In the end I think one has to accept the limitations of the design (like almost every other speaker) in the very good but not super high end quality, cost and ability of the drive units involved. The sound results derived from the driver cost, construction and model range ratio far exceeds the what could be expected from a floorstanding 2 way design (well done Stewart Tyler, if you need any more accolades) and that of many other DIY or commercial designs within the same cost, context and class. Therefore, while the efforts of many good people to improve on the design are admirable, we should not lose sight that attempting to add perfect neutrality is probably not the answer to this design when more of the best attributes are lost along the way. Troels has mentioned the use of better tweeters like the Hiquhon and Scan Speak 9500, the Seas T25-002 Millennium would also be a good match as it is more detailed than the 8513 yet has the lushness that the 9500 is missing.
I noticed on Troels article pictures that he may not have used spikes. I have found that spikes on the recommended plinth dramatically helps the voicing of the speaker by reducing excess bass, clearer midrange and attack. This is as important as using the correct amount of stuffing.
Sorry to disagree but these are my impressions after using the modified tweeter in all situations over the last several days. After having owned original Proac 2.5s several years ago I have to conclude at this point in time that the various mods are not, for me a substitute to perfect the design if you lose the speakers best aspects.
Disabled Account
Joined 2003
95000
The document was removed from the jpo site, here it is:
https://home.decrep.it/New_2_5_clone_tweeter_crossover_and_speaker_setup.pdf
I've made this setup, it souds really great !!!
hope this help.
bye
--
Felipe
The document was removed from the jpo site, here it is:
https://home.decrep.it/New_2_5_clone_tweeter_crossover_and_speaker_setup.pdf
I've made this setup, it souds really great !!!
hope this help.
bye
--
Felipe
Disabled Account
Joined 2003
"new drivers"
- to Wcoil:
Please read my 2.5 paper. I do not insist there are new coated 8535 drivers. The paper clearly says these are production variations and this is confirmed twice from Scan Speak. However, some of these "off-spec" drivers will perform very differently from what appers to be - furtunately - a majority of 8535 drivers, so my only advise to cloners is: watch out.
Regards
- to Wcoil:
Please read my 2.5 paper. I do not insist there are new coated 8535 drivers. The paper clearly says these are production variations and this is confirmed twice from Scan Speak. However, some of these "off-spec" drivers will perform very differently from what appers to be - furtunately - a majority of 8535 drivers, so my only advise to cloners is: watch out.
Regards
The crossover I used was the one in the paper of Troels quoted by Filipe above.
I really like this combination of the 8535 and the 95000. My musical tastes are for classical music and I look for neutrality.
People should also be aware that with the 8513 tweeter, the female voices can sound pretty metallic. However, after 300 hours, they started sounding much better. After more than 400 hours now, they sound not too bad, but I still prefer my other clones with the 8535 and the 95000 for classical music. But at the same time, I also agree with someone else on this board who mentionned that you loose the "magic" of the 2.5. That was a trade-0ff I was willing to make to get a very neutral speaker.
I really like this combination of the 8535 and the 95000. My musical tastes are for classical music and I look for neutrality.
People should also be aware that with the 8513 tweeter, the female voices can sound pretty metallic. However, after 300 hours, they started sounding much better. After more than 400 hours now, they sound not too bad, but I still prefer my other clones with the 8535 and the 95000 for classical music. But at the same time, I also agree with someone else on this board who mentionned that you loose the "magic" of the 2.5. That was a trade-0ff I was willing to make to get a very neutral speaker.
Disabled Account
Joined 2003
I've managed to take photos of my pair- unit numbers 20495-54 and 20495-55.
Although photos make it look shinier/more coated than it actually is, indeed looks very close to Troel's "coated" woofer.
Anyone interested in these photos please let me know and I can email them.
Alain said:
Have you received the new pair and made the speaker for your froiend? Did you use 8513 or 9500?
Although photos make it look shinier/more coated than it actually is, indeed looks very close to Troel's "coated" woofer.
Anyone interested in these photos please let me know and I can email them.
Alain said:
Have you received the new pair and made the speaker for your froiend? Did you use 8513 or 9500?
Another Cloner's Journey
The R2.5s were my first DIY loudspeaker project and I have spent about a year tweaking, testing and listening to them. I have learned a lot along the way, not least, that everyone has different opinions and that you can read DIYaudio.com all day but in the end you really have build/modify and listen for yourself. So here are my (much condensed) experiences with the clone:
I originally built the clones according to ALM's website instructions on
http://www.geocities.com/diyproac25/
using the crossover with the notch filter
http://www.geocities.com/diyproac25/images/Jacq_Troels.gif
After the initial euphoria of hearing noises from boxes that I had built myself, I had to admit that other than some spectacular bass on certain recordings the sound left a lot to be desired. The bass was boomy, the treble spiky and the mid somewhat variable depending on the program material. The thing which really killed them for me was on the intro to the 2nd track on the Chilli Peppers latest album the two guitars playing were merged into a tuneless one. At his point I was ready to sell the damned things as some sort of bad joke.
Experiment 1 was to play around with the amount and density of the dacron stuffing, too much deadened the sound and too little gave a boomy and echoey effect. Experiment 2 was to disconnect the notch filter - this was a major step forward, the timing improved dramatically, the Chilli Peppers started to play their own instruments, but as many have found, the sound was rather
tiring with excessive zing in the upper mids and sibilance in the treble.
Experiment 3 - I tried all the different crossover designs I could find (except the series ones) and eventually decided that the Chinese one was the best. After a while I grew even more tired of the shortcomings of the speaker and decided they would definitely be sold. I also heard some Proac Response 4.2s (I think) at a hi-fi show and they were so lacking in coherence
that I doubted Stuart Tyler posseses a pair of ears. Since then I have heard some R1.5s in a domestic environment and they were rather good so maybe Tyler was just having an off-day with the 4.2s.
Experiment 4. Then I read Troel's paper about the modified drive units and decided it was kill or cure time. I waited about 6 weeks to get the right brand of Damar varnish and eventually got the stuff, modded the drive units and changed to V6.1 or the crossover. This was the first time that I could honestly say that they were better than my £1K commercial speakers.
Experiment 5. Many people said the clones are not good enough to justify better crossover components so I ordered some Mills resistors and Sonicaps to replace the Solens and ceramic cased resistors. The Mills resistors went in first with a noticeable reduction is graininess and increase in detail.
The Sonicaps seemed a bit soggy sounding at first, since they were in the treble section this surprised me. Maybe they took some time to run-in but I had to swop them in and out a couple of times over a period of a few weeks to be sure they were an improvement, eventually and with a few other changes to
the system there is no doubt in my mind they are a lot better than the Solens. In fact I would go so far as to say that the clones are not worth building
without them.
Experiment 6. Al.M's website states that the stuffing is critical to the sound of the clone, since I now had a "TroelAc" not a Proac clone I decided to try some profiled acoustic foam instead of the Dacron. It took me about 2 minutes to decide this was a bad, bad move. If you like boomy bass and dull mids this is way to go!
I have also dabbled with port tuning and seem to prefer a slightly longer port than the recommended one.
I feel like I have been fiddling with these speakers for ever and I've still got to try different internal cable and the effects of adding bracing (probably by clamping the outside of the box - is this a valid method for assessment?).
So overall these speakers are tuned for my tastes, room and system, but surely that's the whole point of DIY. Finally I now feel that the description on AlMs website "very musical, sweet, coherent, tuneful, lush, and with a very bullet proof bass (for its size)" is appropriate.
Strengths:
Coherent, detailed, tuneful, good microdynamics, deep and tuneful bass, neutral.
Weaknesses:
Occassional slightly loose bass (I've never heard an open ported speaker that does not suffer this), treble could use more "air", cannot play really loud, macrodynamics a bit limited.
For what they cost in materials these are a really excellent pair of speakers, OTOH if time is money....Hmmmm best not think about that one too much.
The R2.5s were my first DIY loudspeaker project and I have spent about a year tweaking, testing and listening to them. I have learned a lot along the way, not least, that everyone has different opinions and that you can read DIYaudio.com all day but in the end you really have build/modify and listen for yourself. So here are my (much condensed) experiences with the clone:
I originally built the clones according to ALM's website instructions on
http://www.geocities.com/diyproac25/
using the crossover with the notch filter
http://www.geocities.com/diyproac25/images/Jacq_Troels.gif
After the initial euphoria of hearing noises from boxes that I had built myself, I had to admit that other than some spectacular bass on certain recordings the sound left a lot to be desired. The bass was boomy, the treble spiky and the mid somewhat variable depending on the program material. The thing which really killed them for me was on the intro to the 2nd track on the Chilli Peppers latest album the two guitars playing were merged into a tuneless one. At his point I was ready to sell the damned things as some sort of bad joke.
Experiment 1 was to play around with the amount and density of the dacron stuffing, too much deadened the sound and too little gave a boomy and echoey effect. Experiment 2 was to disconnect the notch filter - this was a major step forward, the timing improved dramatically, the Chilli Peppers started to play their own instruments, but as many have found, the sound was rather
tiring with excessive zing in the upper mids and sibilance in the treble.
Experiment 3 - I tried all the different crossover designs I could find (except the series ones) and eventually decided that the Chinese one was the best. After a while I grew even more tired of the shortcomings of the speaker and decided they would definitely be sold. I also heard some Proac Response 4.2s (I think) at a hi-fi show and they were so lacking in coherence
that I doubted Stuart Tyler posseses a pair of ears. Since then I have heard some R1.5s in a domestic environment and they were rather good so maybe Tyler was just having an off-day with the 4.2s.
Experiment 4. Then I read Troel's paper about the modified drive units and decided it was kill or cure time. I waited about 6 weeks to get the right brand of Damar varnish and eventually got the stuff, modded the drive units and changed to V6.1 or the crossover. This was the first time that I could honestly say that they were better than my £1K commercial speakers.
Experiment 5. Many people said the clones are not good enough to justify better crossover components so I ordered some Mills resistors and Sonicaps to replace the Solens and ceramic cased resistors. The Mills resistors went in first with a noticeable reduction is graininess and increase in detail.
The Sonicaps seemed a bit soggy sounding at first, since they were in the treble section this surprised me. Maybe they took some time to run-in but I had to swop them in and out a couple of times over a period of a few weeks to be sure they were an improvement, eventually and with a few other changes to
the system there is no doubt in my mind they are a lot better than the Solens. In fact I would go so far as to say that the clones are not worth building
without them.
Experiment 6. Al.M's website states that the stuffing is critical to the sound of the clone, since I now had a "TroelAc" not a Proac clone I decided to try some profiled acoustic foam instead of the Dacron. It took me about 2 minutes to decide this was a bad, bad move. If you like boomy bass and dull mids this is way to go!
I have also dabbled with port tuning and seem to prefer a slightly longer port than the recommended one.
I feel like I have been fiddling with these speakers for ever and I've still got to try different internal cable and the effects of adding bracing (probably by clamping the outside of the box - is this a valid method for assessment?).
So overall these speakers are tuned for my tastes, room and system, but surely that's the whole point of DIY. Finally I now feel that the description on AlMs website "very musical, sweet, coherent, tuneful, lush, and with a very bullet proof bass (for its size)" is appropriate.
Strengths:
Coherent, detailed, tuneful, good microdynamics, deep and tuneful bass, neutral.
Weaknesses:
Occassional slightly loose bass (I've never heard an open ported speaker that does not suffer this), treble could use more "air", cannot play really loud, macrodynamics a bit limited.
For what they cost in materials these are a really excellent pair of speakers, OTOH if time is money....Hmmmm best not think about that one too much.
Disabled Account
Joined 2003
What I don't understand is why there's so much variability in peoples experiences with the 2.5.
Is it the variability in materials (ie. lining, dacron, XO parts etc), or is it because of the ProAc sound (ie. elevated bass and treble, questionable midrange)?
Or is it just the nature of DIY?
I couldn't imagine commercial kits eg. Madisound Solist would need as much tweaking...
Is it the variability in materials (ie. lining, dacron, XO parts etc), or is it because of the ProAc sound (ie. elevated bass and treble, questionable midrange)?
Or is it just the nature of DIY?
I couldn't imagine commercial kits eg. Madisound Solist would need as much tweaking...
And yet another Cloner's journey ...
Hi All,
There has been a lot of discussion on this site and at Madisound about how to improve the hi-end ability of the Proac 2.5 Clone ?? In particular, there is a need to improve the 'midrange dip' that exists with this speaker. People have tried x-o mods; varying the port length; cabinet damping; adding/removing the dacron wadding etc...all with mixed results.
Well I think I've found an easy solution, but like all good things in life - it does not come cheap.
All you need to do is ADD A RAVEN R-1 RIBBON TWEETER (@$800AUD per pair) to the Proac 2.5, and you will find that the end result is superb!! With the Raven, the DIY speaker retains that reknowned Proac bass, musicality and coherence but takes a HUGE LEAP forward in detail retrieval, transient snap and achieves an overall more refined treble quality and tonality.
By way of introduction, I constructed a Proac 2.5 Clone about a year ago (in West Australian Jarrah hardwood, finished with 12 coats of clear lacquer and black piano gloss front baffle, with gold badges etc. - it all looks magnificent,BTW). I was very happy with the sonics too, the x-o used is the original 'Jacq' modified circuit, but with all caps bypassed with expensive AEON tinfoil types. The sound of the speaker is excellent with an all-round virtue of being very musical aligned to an extended bass and good detail retrieval. IMO, the 2.5 clone speaker (and the Original) definitely punches above it's weight (considering the parts used) and will go down in history as a Classic design.
However, comparing the sound of the Proac 2.5 (unfair really, I know?) to my Quad ESL's; and another hi end ultra-thin-membraned Electrostatic loudspeaker; and a 90kg Reference quality double D'Appolito cone speaker using a Raven R-1 ribbon tweeter/ two Accuton Ceramic mids/ two 8" Cabasse 21NDC bass drivers - the sound of the 2.5 was sadly lacking in being too laid back in the mid; lacking the attack and decay of the more expensive drivers; and generally lacking treble quality.
So, how to improve on the sound of the Proac 2.5??
Well, I simply disconnected the existing Scanspeak tweeter and x-o and installed a Raven R-1 ribbon tweeter using the 2,800 Hz third order cross-over suggested by Raven/Orca Design. My own mods to this x-o circuit include bypassing the big 15 & 41 uF Axon/Solen caps with small value Aeon Tinfoil caps and adding a 3 watt 10 ohm variable resistor(this pot is very clean sounding too) on the (+ve) line of the Raven tweeter. This variable pot is ABSOLUTELY ESSENTIAL as it allows you to adjust the treble intensity of the speaker to your liking (I like it set at 3.8 ohm). The result is 'right up there' with hi-end loudspeakers and very good indeed! The upper mid is now nicely filled out, treble quality is sweet, ultra-detailed and very extended (to 40kHz as Raven suggests). Best of all, the tonal quality of the Raven R-1 far exceeds the stock Scanspeak tweeter in the 2.5!
A speaker of this quality thrives on good amplification...
The amplification that I use include the Pass Labs Aleph 3 (there is something inherently 'just right' about this little power amp, regardless of price) and the wonderful Supratek Triode 'Merlot' monoblocs (a 6C33-CB Lamm ML2 style valve amp) the and the unique Supratek 'Cabernet' preamplifier. All pure Class A; Single Ended Triode valves; Transformer Coupled; Zero Feedback; Valve Rectified Power Supplies; Parafeed; Lithium Battery Biasing etc. etc. Suffice to say the sound of Supratek amps is superb, and I would describe it as grainless, textured, pristine and clean like the best of solid state while maintaining the valviness that we all like. Check it out at:
www.supratek.biz (for Triode Amplifiers)
www.eraudio.com.au (for DIY Electrostatic spkrs)
You can also quite successfully use the Raven tweeter and new x-o (mine is in an external box) on any other speaker you have at your disposal, so long as the x-o point is similar, just dial in the required level of treble with the variable resistor.
Trust me on this one guys, the addition of a Raven ribbon tweeter definitely takes the humble DIY Proac 2.5 clone speaker into Hi-End territory, so why not give it a try?
Perhaps we can now call this speaker Steve's "Ribbon Response 2.5" ...
Regards,
Steve M.
Hi All,
There has been a lot of discussion on this site and at Madisound about how to improve the hi-end ability of the Proac 2.5 Clone ?? In particular, there is a need to improve the 'midrange dip' that exists with this speaker. People have tried x-o mods; varying the port length; cabinet damping; adding/removing the dacron wadding etc...all with mixed results.
Well I think I've found an easy solution, but like all good things in life - it does not come cheap.
All you need to do is ADD A RAVEN R-1 RIBBON TWEETER (@$800AUD per pair) to the Proac 2.5, and you will find that the end result is superb!! With the Raven, the DIY speaker retains that reknowned Proac bass, musicality and coherence but takes a HUGE LEAP forward in detail retrieval, transient snap and achieves an overall more refined treble quality and tonality.
By way of introduction, I constructed a Proac 2.5 Clone about a year ago (in West Australian Jarrah hardwood, finished with 12 coats of clear lacquer and black piano gloss front baffle, with gold badges etc. - it all looks magnificent,BTW). I was very happy with the sonics too, the x-o used is the original 'Jacq' modified circuit, but with all caps bypassed with expensive AEON tinfoil types. The sound of the speaker is excellent with an all-round virtue of being very musical aligned to an extended bass and good detail retrieval. IMO, the 2.5 clone speaker (and the Original) definitely punches above it's weight (considering the parts used) and will go down in history as a Classic design.
However, comparing the sound of the Proac 2.5 (unfair really, I know?) to my Quad ESL's; and another hi end ultra-thin-membraned Electrostatic loudspeaker; and a 90kg Reference quality double D'Appolito cone speaker using a Raven R-1 ribbon tweeter/ two Accuton Ceramic mids/ two 8" Cabasse 21NDC bass drivers - the sound of the 2.5 was sadly lacking in being too laid back in the mid; lacking the attack and decay of the more expensive drivers; and generally lacking treble quality.
So, how to improve on the sound of the Proac 2.5??
Well, I simply disconnected the existing Scanspeak tweeter and x-o and installed a Raven R-1 ribbon tweeter using the 2,800 Hz third order cross-over suggested by Raven/Orca Design. My own mods to this x-o circuit include bypassing the big 15 & 41 uF Axon/Solen caps with small value Aeon Tinfoil caps and adding a 3 watt 10 ohm variable resistor(this pot is very clean sounding too) on the (+ve) line of the Raven tweeter. This variable pot is ABSOLUTELY ESSENTIAL as it allows you to adjust the treble intensity of the speaker to your liking (I like it set at 3.8 ohm). The result is 'right up there' with hi-end loudspeakers and very good indeed! The upper mid is now nicely filled out, treble quality is sweet, ultra-detailed and very extended (to 40kHz as Raven suggests). Best of all, the tonal quality of the Raven R-1 far exceeds the stock Scanspeak tweeter in the 2.5!
A speaker of this quality thrives on good amplification...
The amplification that I use include the Pass Labs Aleph 3 (there is something inherently 'just right' about this little power amp, regardless of price) and the wonderful Supratek Triode 'Merlot' monoblocs (a 6C33-CB Lamm ML2 style valve amp) the and the unique Supratek 'Cabernet' preamplifier. All pure Class A; Single Ended Triode valves; Transformer Coupled; Zero Feedback; Valve Rectified Power Supplies; Parafeed; Lithium Battery Biasing etc. etc. Suffice to say the sound of Supratek amps is superb, and I would describe it as grainless, textured, pristine and clean like the best of solid state while maintaining the valviness that we all like. Check it out at:
www.supratek.biz (for Triode Amplifiers)
www.eraudio.com.au (for DIY Electrostatic spkrs)
You can also quite successfully use the Raven tweeter and new x-o (mine is in an external box) on any other speaker you have at your disposal, so long as the x-o point is similar, just dial in the required level of treble with the variable resistor.
Trust me on this one guys, the addition of a Raven ribbon tweeter definitely takes the humble DIY Proac 2.5 clone speaker into Hi-End territory, so why not give it a try?
Perhaps we can now call this speaker Steve's "Ribbon Response 2.5" ...
Regards,
Steve M.
Re: And yet another Cloner's journey ...
Good idea!
You could also then replace the SS mid/woofer with a focal midrange and add an active subwoofer using one of those expensive eton bassdrivers.
I would also make the cabinet egg-shaped.
Oh yeah, and maybe we could also ad one of those supertweeters with a 15th order limpbisquit xover
Now that would be a great proac clown!
Steve M said:
Good idea!
You could also then replace the SS mid/woofer with a focal midrange and add an active subwoofer using one of those expensive eton bassdrivers.
I would also make the cabinet egg-shaped.
Oh yeah, and maybe we could also ad one of those supertweeters with a 15th order limpbisquit xover
Now that would be a great proac clown!
Peter
I have listened to Steve's modified Raven clone and it does sound very good. The suggestion does seem to offer an easy improvement.
But before one can simply "trust" and try it out there are many considerations:
1. The example given includes the use of an older version Raven R1 Mark1 tweeter that is no longer available, with the suggested Raven manufacturers's 2800hz 3rd order crossover for the current Mark 2 version Raven R1. The crossover is probably close judging from the good sound result, but exactly how close and precisely where the best crossover is occuring with the woofer is a little unknown at this point in time. No measurement or data to support as yet. The Raven may actually be crossing over lower or higher than we think. Physically and sonically there are differences between Mark 1 & 2 tweeters
2. Actual cost benefit decision to simply try an enormously expensive Raven tweeter versus other cheaper or more expensive alternative tweeters already mentioned in these forums such as Seas Millenium & T25, SS 9900, 9800, 9700, 9500, Morel, Fountek ribbons, Heil AMT, Hiquphon, Vifa XT25, Newform ribbons, Raven R2 - hell why not a R3?
3. The Raven / clone is a relatively young suggestion and will it outlive and survive the OPs tastes and expectations against other possible speaker configurations, in particular filling the midrange gap in the clone.
4. Will cloners be able to appreciate such a tweeter given that average amplification and source equipment appears to be very commonly used out there.
I have listened to Steve's modified Raven clone and it does sound very good. The suggestion does seem to offer an easy improvement.
But before one can simply "trust" and try it out there are many considerations:
1. The example given includes the use of an older version Raven R1 Mark1 tweeter that is no longer available, with the suggested Raven manufacturers's 2800hz 3rd order crossover for the current Mark 2 version Raven R1. The crossover is probably close judging from the good sound result, but exactly how close and precisely where the best crossover is occuring with the woofer is a little unknown at this point in time. No measurement or data to support as yet. The Raven may actually be crossing over lower or higher than we think. Physically and sonically there are differences between Mark 1 & 2 tweeters
2. Actual cost benefit decision to simply try an enormously expensive Raven tweeter versus other cheaper or more expensive alternative tweeters already mentioned in these forums such as Seas Millenium & T25, SS 9900, 9800, 9700, 9500, Morel, Fountek ribbons, Heil AMT, Hiquphon, Vifa XT25, Newform ribbons, Raven R2 - hell why not a R3?
3. The Raven / clone is a relatively young suggestion and will it outlive and survive the OPs tastes and expectations against other possible speaker configurations, in particular filling the midrange gap in the clone.
4. Will cloners be able to appreciate such a tweeter given that average amplification and source equipment appears to be very commonly used out there.
Peter: sarcasm aside, your point is taken, I agree it is not a Proac 2.5 Clone anymore with the Raven added.
Let's just say that it is now an entirely different speaker now, and that is why at the end of my long post I called it Steve's "Ribbon Response 2.5".
There is a lot of merit in going down the route of the Raven for cloners who have taken the time to build this speaker, but have now come to realize that it is not quite in Hi-End territory. For the cost of a pair of Raven ribbons, you can transform the good work you have already done into something VERY special!!
Albert: Someone on the Madisound discussion forum mentioned that the Proac 2.5 may have an in-built midrange dip because the SS bass driver is rated up to 3,500 Hz and the SS tweeter down to only 4,000 Hz (there is a gap that can't be closed, when ideally there should be an overlap at the x-o point). Perhaps this why the Raven clone sounds so good, because I have it crossing over at 2,800 Hz with the recommended third order Orca Design crossover. This would appear to overlap and close the midrange gap in the original 2.5 clone nicely ??
Anyway guys, the results speak for themselves. The original 2.5 clone is a great speaker, with a high level of coherence, musicality and stupendous bass that is hard to beat (considering it's compact dimensions). Add the Raven tweeter to this recipe, and it becomes a magnificent speaker knocking on the doors of Electrostats and Hi- End Raven/Accuton/Cabasse combinations !!
Regards,
Steve M.
Let's just say that it is now an entirely different speaker now, and that is why at the end of my long post I called it Steve's "Ribbon Response 2.5".
There is a lot of merit in going down the route of the Raven for cloners who have taken the time to build this speaker, but have now come to realize that it is not quite in Hi-End territory. For the cost of a pair of Raven ribbons, you can transform the good work you have already done into something VERY special!!
Albert: Someone on the Madisound discussion forum mentioned that the Proac 2.5 may have an in-built midrange dip because the SS bass driver is rated up to 3,500 Hz and the SS tweeter down to only 4,000 Hz (there is a gap that can't be closed, when ideally there should be an overlap at the x-o point). Perhaps this why the Raven clone sounds so good, because I have it crossing over at 2,800 Hz with the recommended third order Orca Design crossover. This would appear to overlap and close the midrange gap in the original 2.5 clone nicely ??
Anyway guys, the results speak for themselves. The original 2.5 clone is a great speaker, with a high level of coherence, musicality and stupendous bass that is hard to beat (considering it's compact dimensions). Add the Raven tweeter to this recipe, and it becomes a magnificent speaker knocking on the doors of Electrostats and Hi- End Raven/Accuton/Cabasse combinations !!
Regards,
Steve M.
To quote from Madisound.com forum Dennis M:
"Maybe the Raven-modified clone is a first-class speaker. I haven't heard it. But I am suspicious, having listened carefully to the original version. There are two issues I don't understand. First, you refer to the "gap" in the middle as some kind of defect. My understanding is that Proac deliberately voiced the 2.5 in "U" shape fashion to increase the sense of space and air. I never thought it was an uninteneded crossover error. Second, the main problem with the clone was clearly the smearing from
a mid-treble woofer peak. That ruined the speaker for me. Others obviously heard this as well, given the subsequent trap circuit that was offered to deal with the problem. Adding a ribbon tweeter on top of a flawed woofer response would do nothing to correct this coloration."
It is not really a mid gap as if there is nothing there, its just a dip, and it can be closed with the original tweeter, yet the designer did not chose this response. Today I also tried the Raven Mark II as suggested, yes nice but IMO it still does answer all the issues and infact loses some of the original magic, nevertheless a good option, somewhat leaner but smoother treble response, lots of hyper detail. Perhaps the smooth mid transition gives the impression that the alleged "mid gap" is completely fixed compared to the lusher lower treble response of the original tweeter. OP needs to A-B compare this with other correct speakers at his current disposal otherwise relying on aural memory of previous examples.
"Maybe the Raven-modified clone is a first-class speaker. I haven't heard it. But I am suspicious, having listened carefully to the original version. There are two issues I don't understand. First, you refer to the "gap" in the middle as some kind of defect. My understanding is that Proac deliberately voiced the 2.5 in "U" shape fashion to increase the sense of space and air. I never thought it was an uninteneded crossover error. Second, the main problem with the clone was clearly the smearing from
a mid-treble woofer peak. That ruined the speaker for me. Others obviously heard this as well, given the subsequent trap circuit that was offered to deal with the problem. Adding a ribbon tweeter on top of a flawed woofer response would do nothing to correct this coloration."
It is not really a mid gap as if there is nothing there, its just a dip, and it can be closed with the original tweeter, yet the designer did not chose this response. Today I also tried the Raven Mark II as suggested, yes nice but IMO it still does answer all the issues and infact loses some of the original magic, nevertheless a good option, somewhat leaner but smoother treble response, lots of hyper detail. Perhaps the smooth mid transition gives the impression that the alleged "mid gap" is completely fixed compared to the lusher lower treble response of the original tweeter. OP needs to A-B compare this with other correct speakers at his current disposal otherwise relying on aural memory of previous examples.
Five week in and...
My clone is complete. It began with reading the start of this thread and I have done a lot of reading and some experimenting. I don’t know if I am there yet but the sound has come a very long way since ‘first sound’ about 5 weeks ago.
I began with the Troels V6 crossover and was very disappointed. Partly because the boxes were not damped enough but using vita foam (1”) as with my Response 1SCs things improved enormously. I used a double thickness behind the woofer but this deadened the sound far too much. Other than this Al.M has it absolutely right for my ears. Everything he says is true. At least for me. I have worked my way back from Troels V6 using the 3.3uf in place of the 3.9uf and this did clear things up. I have also recently changed a air cooled 1.8mh inductor (0.7R) for the <0.2R ferrite coil and the sound improved in a leap. This I think was due to the lower resistance (small though it seems to be) and the fact that the coils are now at right angles.
Everything else Al.M says I have also found to be true my units are now sounding very similar in voicing to my beloved R1SCs only better stereo deeper sound staging more powerful bass and I have to say more detail, good as the 1Sc are. I don’t know how they sound against original 2.5s but these are improving with each change...... so far.
I plan to include the notch filter having removed the R from the mid/woofer section. There is a bit of recess in mid range (like the R1SCs but more so) and I think a bit of hardness in the upper frequencies. I think the notch filter might help. And they are yet to be put up on their spikes which may help the balance. Also the longer the units run the smoother they seem to get. The R1SCs had a similar hard sound when I got them several years ago but are very mellow now. Other than perhaps a good wine, few things get better with age and it is comforting to think that a great speaker now, will just get better, with almost no extra investment!!!
I may play with the filling a bit more, standardise on the XO caps (using a mixture at the moment) and may make a few more mods to the XO to see what happens, but generally I am very pleased with the sound for my ears and in my room. I think valves would perhaps only improve things (I use Mitchel Alecto mono blocs great but still short of the valve smoothness).
Thinking back on things the place to start is with the original Japanese XO and develop from there, not working back from the later XO versions. I did notice in his papers that Troel’s speakers were heavily braced and deadened and he seems to favour a less lively sound, which is what you get with his mods, each one taking you further from the Proac sound which I have to say is a little 'hard' and not appealing to every one.
Thank you to every one on this forum because it would take forever for one person to experiment with so many variables this way you can make some easy judgments. Particularly Al.M seems to have it about right for my ears. If you like the Proac sound 'as is', this is the advice to start with.
My clone is complete. It began with reading the start of this thread and I have done a lot of reading and some experimenting. I don’t know if I am there yet but the sound has come a very long way since ‘first sound’ about 5 weeks ago.
I began with the Troels V6 crossover and was very disappointed. Partly because the boxes were not damped enough but using vita foam (1”) as with my Response 1SCs things improved enormously. I used a double thickness behind the woofer but this deadened the sound far too much. Other than this Al.M has it absolutely right for my ears. Everything he says is true. At least for me. I have worked my way back from Troels V6 using the 3.3uf in place of the 3.9uf and this did clear things up. I have also recently changed a air cooled 1.8mh inductor (0.7R) for the <0.2R ferrite coil and the sound improved in a leap. This I think was due to the lower resistance (small though it seems to be) and the fact that the coils are now at right angles.
Everything else Al.M says I have also found to be true my units are now sounding very similar in voicing to my beloved R1SCs only better stereo deeper sound staging more powerful bass and I have to say more detail, good as the 1Sc are. I don’t know how they sound against original 2.5s but these are improving with each change...... so far.
I plan to include the notch filter having removed the R from the mid/woofer section. There is a bit of recess in mid range (like the R1SCs but more so) and I think a bit of hardness in the upper frequencies. I think the notch filter might help. And they are yet to be put up on their spikes which may help the balance. Also the longer the units run the smoother they seem to get. The R1SCs had a similar hard sound when I got them several years ago but are very mellow now. Other than perhaps a good wine, few things get better with age and it is comforting to think that a great speaker now, will just get better, with almost no extra investment!!!
I may play with the filling a bit more, standardise on the XO caps (using a mixture at the moment) and may make a few more mods to the XO to see what happens, but generally I am very pleased with the sound for my ears and in my room. I think valves would perhaps only improve things (I use Mitchel Alecto mono blocs great but still short of the valve smoothness).
Thinking back on things the place to start is with the original Japanese XO and develop from there, not working back from the later XO versions. I did notice in his papers that Troel’s speakers were heavily braced and deadened and he seems to favour a less lively sound, which is what you get with his mods, each one taking you further from the Proac sound which I have to say is a little 'hard' and not appealing to every one.
Thank you to every one on this forum because it would take forever for one person to experiment with so many variables this way you can make some easy judgments. Particularly Al.M seems to have it about right for my ears. If you like the Proac sound 'as is', this is the advice to start with.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.