I only ever feed my system from CD / SACD source but there are many CD's (old and new) where the soundstage is so wide that it wraps around - 360 degrees. Instruments often appear to be behind my listening position - and no I'm not using any pseudo 3D encoding. I wonder if anyone else gets this experience from their Rens. Indeed a friend of mine who owns Krell gear driving B&W800S speaker ($$$) claims that the Ren's are somhow creating an artificial 3D experience. I don't buy that though. I believe the Ren's are one of a select few speakers that can actually recreate the actual recorder soundstage. |
I have actually built external crossovers for my Renaissance 90 EMIT tweeters using Mundorf SIO caps and copper foil inductors. There was quite a jump in efficiency and I had to slightly alter the tweeter amp gain to compensate. For single amp drive you could pad the EMIT with a slightly bigger resistor (0.5 > 1 ohm).
The main benefit of the upgraded crossover was improved dynamics. I also modelled the crossover frequency response in LT spice and found that there was a 2.5dB peak at 2.3KHz. I was not surprised to see this because I had always thought that they were a bit too forward in the presence range. I have heard that there were changes made to the crossovers during the Ren 90's lifecycle that corrected this but I'm not 100% sure. However, the Ren 80's are using diferent values and do have a flatter crossover response when modelled.
I subsequently modded my crossover values to achieve a response +/- 1dB from 1Khz to 20KHz. To do this correctly you need to change a 33uF cap in the EMIM crossover to 30uF and also insert a 0.22R in series with the cap. When I did this mod I also noticed that the 10 year old 33uF electrolytics had expanded slightly at each end. I replaced them both with Blackgate 30uF 50V NPO electros and there was an immediate audible improvement. The MF seemed more pure and imaging was much more accurate. The other fine tune necessary is to add a 1uf in parallel with the 5.6uF pp film cap in the EMIT crossover. Both of these changes combined will result in a much smoother sounding upper mid range and a far more seamless transition from EMIM to EMIT.
More recently I have disconnected the EMITS and am using Raal pure ribbon tweeters which sit on top of the Ren 90's. This also necessitated some fine tuning of the crossovers but the result is really stunning. I have also designed a DC bias system for the tweeter crossover so that both SIO caps are voltage biased at 33V DC.
The main improvement from the Raal tweeters stems from the much wider HF dispersion and far more extended treble, but they also seem cleaner, less grainy and sweeter sounding, but then again they are also very expensive. IMO the EMITs used in the Rens are a very good tweeter and with the 2 component changes mentioned above the response will definitely be flatter. |
Melbguy, If you're completely happy with the tonal balance then it's good advice to leave them alone.
FWIW I'm using a 10 year old Meridian 557 to drive the Woofers, Mid Bass's and EMIM's and I designed and built my own amp for the tweeters because I couldn't find anything commercial with the smoothness and refinement I was after. The EMITS can sound a bit metalic with some SS amps and my own design - using a mosfet output stage with bipolar drives in CFP configuration results in a very sweet fluid treble while the gutsy Meridan gives me the dynamics and control from the mids and down to the sub bass.
I find the Ren 90's a tour de force when it comes to 3d imaging. I'm yet to hear another large speaker that can present a 3D image the way these can. |
Hi Okeeteekid,
Thanks for the comments. Yes as indicated in an earlier post I made a dedicated EMIT crossover using Mundorf SIO and copper foil inductors. I have also tried Claricaps. My favourite are the Mundorfs but they need a special DC bias generator to sound their best. I'm sure there are quite a few caps that sound better than the Solens but IMO they are not too bad and if they do err it's on the polite side so they might be more listener friendly for most people. But then it's more about personal choice when it comes to caps and there always seems to be a diversity of opinions. Nice work on the K9's though. They are a great bang for buck speaker.
Cheers,
Timpani |
Re the phase issues, yes you are correct and my adapted tweeter crossover uses some conjugate matching to minimise the sharp phase transitions. Infinity's are not everyone's choice and your friend may have a valid point, however, the issue of coherence can be minimised with careful attention to speaker placement and the right choice of a listening room. Any 4 way speaker will need more distance to generate a coherent wave and the ren's are no exception. One big advantage of planar drivers is that the impedance is flat and resistive not only within but also well outside the operating range, so passive crossover design is far easier to get right and can be achieved without RLC networks that rob efficiency and introduce other side effects. Planar driver impedance also remains constant with power which is a source of non linearity with cones or domes. Every design has strenghts and weaknesses and the 90's are not perfect but they really don't do anything very wrong and what they do very right is difficult to live without once you have lived with them for a while. BTW from my brief experiences with Classe amps I think they are an excellent choice for the Ren's but I'm sure you will hear improved dynamics when you upgrade to the CA-2200. I would like to hear that setup. BTW I live in Sydney and bought my Ren 90'S new back in 1995 from Audio Excellence. The RRP at the time was $7999 but they gave me a nice discount. I have owned many other speakers including B&W 801 but these are my fav. |
Hi Okeeteekid,
Thanks for the comments. Yes as indicated in an earlier post I made a dedicated EMIT crossover using Mundorf SIO and copper foil inductors. I have also tried Claricaps. My favourite are the Mundorfs but they need a special DC bias generator to sound their best. I'm sure there are quite a few caps that sound better than the Solens but IMO they are not too bad and if they do err it's on the polite side so they might be more listener friendly for most people. But then it's more about personal choice when it comes to caps and there always seems to be a diversity of opinions. Nice work on the K9's though. They are a great bang for buck speaker.
Cheers,
Timpani |
Melbguy + others
IMO the only real shortcoming of the emit tweeter is the narrowing dispersion at very high frequencies. This results in a rolloff of high frequencies above 10KHz at > 25 degrees off axis and a rising response - on axis. For this reason the amount of 'toe in' becomes very critical in achieving the right tonal balance which makes then Ren's more of a 'one person' speaker because the sweet spot is very narrow. I have experimented (in the past) with ribbon super tweeters (sitting on top of the cabinets) to supplement the dispersion above 10Khz but I found that it was best to cross over (via a passive network) just above 10KHz, but the additional crossover introduces time alignment complications which can take plenty of listening time to fully optimise. However I was able to acheive excellent results using Raven R1's as super tweeters with a first order crossover at 10KHz. This made the horizontal dispersion really wide and gave the treble more 'air' creating an even bigger soundstage. Unfortunately the trade off was poorer vertical beamwidth (due to the distance between the EMIT and the Ravens) and this made listener height more critical. However I did persist with this setup for some time because it worked better in my room. I'm now using Raal ribbon tweeters which sit on top of the cabinets and the EMITs are no longer connected. This avoids the interactions between tweeters but the larger distance between the EMIM and the Raals does reduce the vertical sweet spot. I will probably (ultimately) go back to using the EMITs with the Raals purely as super tweeters because the EMITs IMO are better than all of the pure ribbon tweeters I've tested in the low range (2 - 6 KHz). The Ravens in particular are very poor performers below 6KHz where they sound very thin and dynamically very lacking. They are a very over rated tweeter IMO but they do work well as super tweeters due to the 50KHz BW and performance above 10KHz is pretty good.
FWIW I'm about to embark on an upgrade of the caps in the Watkins woofer crossover. The bass is sounding a little bit spongy these days which makes me think that the caps are lossy and need replacing. I will provide feedback to this thread on the outcome of that upgrade in the hopefully not too distant future. I do have the schematic but getting the correct value caps from a reliable manufacturer might be a challenge.
Here is a link to the Ren 90 crossover schematic. I would advise downloading this while it's available. Note there is an error. The 5 ohm resistor in series with the EMIT should read 0.5R.
http://www.kalman.org/RENAISSANCE%2090%20x-over.pdf
Timpani |
The xo changes did not seem to find their way into an updated schematic but there must have been errata's sent out to distributors. The only schematics I have been able to find have all had the same values as those in my early gen 90's. The changes I am aware of are in the EMIM LPF (2nd pole) and there were also some updated values in the woofer XO to provide a bit more punch and speed in the bass - at the expense of a little extension. None of these changes make any radical difference but the changes to the EMIM do make the response slightly flatter through the crossover region and address a 2-3dB spike at 2.3kHz. Bear in mind that the values chosen in the EMIM / EMIT crossover are mostly all 5% standard values - which are easy to obtain and result in a lower production cost so there is actually some scope for fine tuning. If you look at the crossover schematics for the IRS Sigma EMIT / EMIM XO you will not find any standard value caps or inductors - hence the higher price tag. Fortunately though the solen caps in the EMIT XO are all very HQ and ditto the EMIM with the exception of C4. This cap is somewhat critical because any capacitance modulation here will result in some audible distortion, which is why I changed it to a BG. The Sigma's address this by using a polyprop cap in parallel with a HQ electro. This is a 50% improvement but I'm sure the Epsilon uses pp throughout.
In summary, the Ren 90's are using the exact same EMIT + EMIM drivers as used in the Sigma's and nearly the same quality of crossover components - save for one cap. The main difference seems to be that the Sigma's do have fine tuned component values for a slightly more linear response - but remember that subsequent revisions of the Ren XO have largely addressed these so if you are not aware of any colourations then any modding is probably not worthwhile. FWIW I have never heard the Sigma's but it's unlikely that they would offer any significant improvements over the Ren's. |
The main issue with the EMIT as I see (hear) it is that they become very directional above 10kHz and they are too hot on axis and too rolled off when off axis. The tonal balance from 10KHz down is best when 30 dgrees off axis and for that reason I have always found them to sound most accurate when firing directly down the room without any toe-in. These emits are the same as used in the Epsilon but in the Epsilon they are used with a specially designed dispersion membrane which increases their horizontal and vertical dispersion and eliminates the hot on axis response, making them more accurate over a wider listening window. As you would have read in my previous posts my ultimate recommendation is to use a supertweeter to augment response above 10kHz but you need to have some technical expertise to implement this correctly.
If you want to experiment another option is to slightly modify the existing EMIT crossover. If you can locate the 5.6uF solen capacitor then solder a 1uF 2% polypropylene film cap in parallel with it. Then locate the 0.5 ohm (2x1 ohms in parallel) closer to the EMIT crossover input, and remove one of the 0.5 ohm resistors so that it becomes 1.0 OHMS. (Do not touch the 5W 0.5 ohm resistor at the crossover output.)
This mod will remove a saddle in the response between 4 and 8kHz and will allow you to toe the speakers in a bit further without them sounding as peaky in the 10 - 12kHz region.
Timpani |
Yes I agree with you re the tilt. None or just a little works best for me too. I also discovered early on that close proximity to side walls kills the soundstage. You need several feet (1m) as a minimum. |
Melbguy, you mentioned how sensitive the rens are to electronics. That is definitely the case and has enabled me to hear subtle changes in the preamps I've designed over the past 10 years. Although the power amp usually offers the biggest contributor to SQ I have also heard great power amps ruined by poor preamps and front ends. One power amp that was really outstanding was a ME1400. I had Trevor Wilson from Rage audio up here one day demo'ing one and I have really never heard any amp since to drive the Watkins woofers the way the ME1400 did. Unfortunately the ME1400 was going to set me back about $8000 so I resisted the purchase - just. That big ME amp could apparently output more than 60 amps so I think there's a message in there - that high current is the way to go to get the best out of the WWs. |
Melbguy, My comments need to be taken in correct context and with allowance for the fact that I have recently been playing with tweeters that have a 120 degree beamwidth at 20kHz. The EMIT's are probably as good as many or even most dome tweeters in terms of dispersion and everything you point out about their develoment is correct but I am pretty sure that they were refined further for the Epsilon and the dispersion was the major criteria. I recall reading this in the Sterephile review about 10 years ago now (how time flies). By adjusting the amount of toe-in you will certainly be able to fine tune them nicely but just don't expect a really huge sweet spot. But as I have said before the EMIT's are fine tweeters and by far and away the best planar tweeter I have ever heard. I have also heard several of the Kappa series and IMO the Rens are more accurate and detailed and because the EMIM and EMIT are of exactly the same design they are able to acheive that rare feat of almost perfect integration from mid to treble which IMO is by far the most important crossover point in any speaker system. In most respects the Infinity multi planar driver approach is better than most of the full range planar designs due to the mentioned beamwidth > frequency issues and also because planar drivers make it possible to implement the near perfect passive crossover - due to their flat and wide resistive impedance. In other words it's almost as though there is no crossover and that's why (when set up correctly) the Ren's + IRS Infinity's sound quite a bit like an electrostatic speaker.
I can understand that the next 2 weeks will be a long time
but all the very best for eternity after that. After all, 'Infinity' is forever.
|
It's an age old discussion point. The safe statement is to say that any system is only as good as the weakest link. However, as a part time amplifier designer I can say that there is infinitely more scope for design variations with amps than there is with speakers, which is the opposite to what most people might expect. It's also a fair point to state the amplifier > speaker interface and the complex load presented by the speaker can and does induce distortions in every amplifier - but very differently depending on the ampifier design. This can be a huge source of confusion for audiophiles who spend a great deal of time changing their power amplifiers for the best possible result only to find after an upgrade of speakers that the amp sounds mediocre when compared with one they had previously discarded. Of course these statements are most true of low to mid priced amps but they can also sometimes apply to very expensive designs. I regard Krell highly because in my experience they produce close to the the same sound regardless of the speaker load presented. When you hear differences between speakers you can be far more certain that they are indeed differences between speakers. Not too long ago I was visiting someone who loved AR preamps. He was demo'ing differences between interconnect cables to the power amp and he believed that his AR must be better than his Krell because it was showing up big differences while the Krell did not. Of course the reality was that the different cables had very different amounts of capacitance which was exposed by the relatively high output impedance of the tube preamp. Unfortunately I was not able to convince him that this was the case. That's not to say that I don't believe that cables can make a difference but from my experince as an amp designer I have found that a small change to an amplifier gain stage - even a small change to a resistor value can result in a far bigger change than you can get by changing cables.
|
Hi Melbguy,
I actually responded to you on AK which you will see before you read this but it was without the benefit of your latest comments here. IMO 4K on a line filter is insane but even more so in your case if you buy the Classe CAP because it is so well designed. Read my other comments and digest them. The mods will work out fine if your AE knows what he's doing but the best approach - be it single or bi-amp may not be apparent until the design work is partly underway. I'll explain further tomorrow. |
Hi Clavil,
I've not heard the latest Evo amps but from my experience the really big Krells usually don't sound quite as nice in the mid / treble as the smaller Krell's. My favourite Krell is the FPB300 which is also pure Class A. I heard that the Evo are now class AB ?.
Sometimes a room can also cause a dry and rather bland treble - if there is too much carpet and soft furnishings as well as room treatments that absorb too much treble.
|
AFAIK, All Ren 90's come standard with a 0.5R resistor in series with the EMIT and 2x 0.5R resistors in series (1R) at the input to the tweeter crossover. The tweeter crossover is identical to that (same values) used in the Sigma / Epsilon but the latter drive 2x EMITs in series. The crossover values are actually optimised for 8 ohms, not 4 ohms and the response exhibits a pronounced peak at 4kHz followed by a saddle at 5-7 kHz before rising above 10kHz. My guess is that this gives a reasonably flat response at 30 degress off axis, although I've not performed measurements to confirm.
The other point to note about the Ren's is that the EMIM crossover is not quite as well optimised as it in the Sigma or Epsilon, thus allowing scope for considerdable improvement which means a flatter response and a less demanding load on the amplifier driving it.
With a few value changes and component upgardes plus fitting of a fabric acoustic disperser to the EMIT, it's possible to gain a ginat lift in accuracy and purity.
|
