advertisement


SBLs from Snaxo to Exakt - A World First!

gidders

hifi enthusiast, golfer & photographer
At the beginning of the year, having been running a pair of SBLs with a pair of 250s sort of bi amped (one for left channel only & one for right - just because I had 2 x 250s & hadn't got round to swapping them for a pair of 135s) I decided that it was time to do the proper thing & go active.

70643072_10157593750100682_2280851719799701504_n.jpg


So SNAXO & Supercap added and this gave a significant improvement in detail, soundstage, PRaT & all the things that SBLs are famous for... But I was still left with the feeling that they are a bit forward & a bit bass light. In fact a non audiophile friend described them as "a bit shouty". And yes - I'd done all the set up things, as close to a solid brick wall as the skirting board would allow, re-sealed the bass loading box etc

Then a couple of months ago I was at Brian & Trevor's in Manchester for a musical evening listening to a Linn/Kudos Exakt active system thinking this system sounds amazing (although I would have liked to hear it with Naim power amps) but the transparency that the Exakt technology brought, which performs the crossover in the digital domain, was breath-taking. In the interval I remarking to one of the other attendees how great Exakt was & that I wished it was available for SBLs! And he said "I write Exakt filters for speakers that Linn doesn't support" OMG!! That person was Neil Hallworth of speakerfilters.com

So we embarked on the Exakt SBLs project. Step 1 was to physically measure the SBL cabinets, the position of the drive unit centres, the front baffle slope etc. These numbers were then plugged into Linn's system to enable Linn's Space Optimisation to be enabled. That is available now for free for anyone using a Linn DS or DSM with SBLs via Konfig!

Step 2 involved Neil calling at my house to take some electrical measurements of the drive units to measure their impedance.

Step 3 Neil then took these measurements together with some information on the crossover frequencies & slopes obtained from Naim to work some digital magic to create a first cut of the digital Exakt crossover filters. In fact he sent me three versions with different crossover slopes, 2nd order, 3rd order & 4th order. It was very quickly apparent that the 3rd order (which is what Naim used) was the best. The 2nd order sounded congested in the mid-range while the 4th order sounded as though there was a hole in it. He then sent me a 3rd order v2 without telling me what he'd done. When I tried it, it was clear that this was an improvement over v1 with better definition & spatial separation. The sound stage was more clearly defined & I was in heaven. The slightly bright & comparatively bass light presentation was gone without losing any of the strengths of musicality & PRaT that SBLs are renowned for. Neil assured me that there would be more to come in step 4 but I was so impressed that my SNAXO & Supercap were already up for sale!

71178756_10157593750645682_6600916446740480000_n.jpg


Step 4 Neil called again to fine tune the filters. Firstly we agreed that in taming the brightness, they had lost a little of their sparkle. Now apparently, the sound waves emanating from all drive units have an interaction with the speaker cabinet front baffle and to compensate for this, within the Exakt crossover Neil had created a 0.8 dB treble filter. Now I'd been advised by John Jackson (ex Naim Senior Test & Service Engineer & currently Witch Hat Lead R&D Engineer) that the SBL cut away foam grill was designed to overcome the treble baffle interaction issue. So Neil removed the filter, updated the Exaktbox & sparkle was back. Hi hats shimmered, room acoustics were tangible, the space around instruments & performers improved.

Next we played with the actual crossover frequency. With v1 the bass/mid roll off & treble roll on were at the same point 2.4khz which we'd got from one Naim source. In v2 the bass/mid was rolled off at 2.3khz & the treble rolled on at 2.5khz (from a different Naim source) So we tried widening the gap by 50hz either way - not as good. Leaving the bass/mid where it was but bringing the treble back again - immediate improvement!

Adding baffle filters for the bass/mid drivers for both the length & width of the cabinet added further improvements and the subtlety of the changes that could clearly be heard was mind blowing; -0.5 dB - too much, - 0.25 dB, not enough so we settled on - 0.4 dB. Adding a 1.5 dB lift in the bass from 0-60 hz - too much, 1.2 dB - still too much, 1.0 dB - bingo! In all we went through 17 variations fine tuning the Exakt crossovers in just over two hours. Imagine trying to do that with resistors & capacitors in the analogue domain!

The end result? Absolutely stunning. If you thought going from passive to active with a SNAXO was a big step... from SNAXO to Exakt is like having seven league boots on! Really, it's that good. If you'd like to listen just get in touch

Oh, the cost? Well I'm going to make a few assumptions. If you're already active with SBLs, then presumably you've got a good front end - LP12 around Akurate spec, a Naim NDX minimum, maybe NDS, or a Linn Akurate DS/DSM Katalyst, Naim 82/Supercap or more likely 52/Supercap, SNAXO / Supercap.

Hardware. You keep whatever power amps you're using - in my case a pair of olive Nap 250s which will retain your signature sound. Ideally you'll need a Linn Akurate System Hub which becomes your pre amp - £2700 new or about £1600 pre owned, & a Linn Akurate Exaktbox Katalyst which is where the majik happens (see what I did there) - £5000 new or I picked one up pre owned with 4 years warranty for £3200; And, you'll need a set of Exakt filters from speakerfilters.com which are just £295. These can be tested for a free 2 week trial period before purchase. Total cost about £5400 pre owned or ~£8000 going new

But you can recoup say £1500 on your NDX, or £2000 on your Akurate DS, £2000 on your 82 /Supercap & £1500 on your SNAXO /Supercap = £5/5500 so approx. cost neutral going pre owned :) And if you've got a 52 or an NDS you might even have some cash left over :)

If you would like to come & have a listen (Birmingham, West Midlands) or a chat, just get in touch - you will be very welcome - just be prepared to radically rethink what active can do for you :)

A huge shout out for Neil of speakerfilters.com for all the work designing & refining the filters.
 
Never heard SBLs but have active Ninkas, I’m thinking of going down the Exakt route with them at some point.

I’m also in the West Mids (Shire Oak, just up from Brownhills) Where about in Birmingham are you?
 
So you spent an additional five and a half large on getting the best from SBLs? You could have bought some fabulous modern speakers for that. I guess you must have found that nothing pleases you as much as SBL's?
Hope you continue to enjoy.
 
So you spent an additional five and a half large on getting the best from SBLs? You could have bought some fabulous modern speakers for that. I guess you must have found that nothing pleases you as much as SBL's?
Hope you continue to enjoy.
It’s amazing how far an old speaker can go with modern partnering kit, especially things like DSP crossovers... which are the most significant advances in loudspeaker technology in history. Dynamic drivers and cabinet materials have gradually improved over the years but DSP allows for an accuracy of control and fine tuning over those components that wasn’t possible previously.
 
Never heard SBLs but have active Ninkas, I’m thinking of going down the Exakt route with them at some point.

I’m also in the West Mids (Shire Oak, just up from Brownhills) Where about in Birmingham are you?

Probably worth the visit. SBLs are very different to Ninkas so best not to consider that Exakting your Ninkas would make them sound like SBLs :)
 
And, to get the best from modern speakers, you could still need to spend the same on the electronics...
So you spent an additional five and a half large on getting the best from SBLs? You could have bought some fabulous modern speakers for that. I guess you must have found that nothing pleases you as much as SBL's?
Hope you continue to enjoy.
 
It's interesting that you ended up with the crossover frequencies you did. Naim docs say the snaxo crosses over at 2.7KHz which is pushing it for the bass driver so any dropping of frequency should help as long as the tweeter can take it. Whether the small drop is enough to be audible, I wouldn't know as the snaxo is fixed. I have active SBLs and do find the midrange slightly odd possibly because of the high crossover frequency.
 
It's interesting that you ended up with the crossover frequencies you did. Naim docs say the snaxo crosses over at 2.7KHz which is pushing it for the bass driver so any dropping of frequency should help as long as the tweeter can take it. Whether the small drop is enough to be audible, I wouldn't know as the snaxo is fixed. I have active SBLs and do find the midrange slightly odd possibly because of the high crossover frequency.
The SNAXO crossover point is a nominal value. It actually rolls the drivers off at different frequencies due to the relatively gentle 3rd order slopes which, of course, should combine to a pretty flat response in a theoretically ideal World. We experimented with a number of slopes and crossover points, centred around 2.4kHz and the differences are clearly audible. As Clive says, even moving them by as little as 50Hz at that frequency makes an easily heard difference.
 
Last edited:
It's interesting that you ended up with the crossover frequencies you did. Naim docs say the snaxo crosses over at 2.7KHz which is pushing it for the bass driver so any dropping of frequency should help as long as the tweeter can take it. Whether the small drop is enough to be audible, I wouldn't know as the snaxo is fixed. I have active SBLs and do find the midrange slightly odd possibly because of the high crossover frequency.

I'm not sure what Naim doc you are quoting. I emailed Naim & they came back with

LF is -6dB at 2.3kHz, 3rd order Bessel
HF is -6dB at 2.5kHz, 3rd order Bessel
Starting levels (before any adjustment for the room) should be with HF set to -2.5dB at 10kHz wrt LF measured at 60Hz.

and John Jackson - ex Naim senior design engineer now working for Witch Hat advised me

2.4KHz at 18dB/Oct
2.4kHz is the -6dB point for both HF and LF
 
I have active SBLs and do find the midrange slightly odd possibly because of the high crossover frequency.
It would be great if you could hear Clive's Exakted version as a comparison. I'm pretty sure the tweeter would be fine down to about 1500Hz and 1700Hz would be a viable crossover point. Does anyone have any documentation as to why Naim made the original choices? Were they trying to keep vocals all in the one driver perhaps?
 
Sounds like an interesting project, that has yielded positive results! I've played with active systems on and off for many years, Saras, SBLs and currently Kans; by the sound of it, there's a lot more potential to be realised in these systems.
 
A very good and interesting write up ‘gidders’. It reminds me, and resonates, of my own experience when I converted the bass drivers of my own speakers to active operation with the crossover accomplished in the digital domain (and also with some room correction thrown in). Although my speakers are three way, rather than the two way of the SBL, I experienced very much the same findings as yourself. I also moved from a second order to a third order crossover point and found this, very similarly to yourself, removed a muddied quality from the lower mid/upper bass; this last only really noticeable once it was removed. Finally, I also found that even small changes in the crossover point made a subtle, but appreciable, difference.
 
This is a massively interesting thread. I have active SL2s with 52/Supercap/SNAXO2-4/Supercap/4x135s and while I like it I do wonder about the SNAXO's ability to split the signal as optimally as it might. I wonder if Neil at speakerfilters would like an SL2 project at some point?
 
I'm not sure what Naim doc you are quoting.

I was going on the Naim Active Manual page 15

Specifications:
SNAXO
...
crossover frequencies
2-4 Credo, SBL 2.7 kHz
..
and later on

ixo2
...
crossover frequencies
credo, SBL 2.7kHz
...

I read this years ago and assumed it was correct. Some mis-direction from Naim?
...or just an honest typo?
 
I always thought the xo on sbls and sl2’s was 2.7k, and snaxo the same. Will get around to measuring it sometime soon hopefully, not that hard to do with such easy access the external xo, surprised it hasn’t been done here already.

I’ve seen some figures thrown about like 1.5 or 1.7k for a lower point. Which of course would be beneficial since 2.7 is way too high for an 8” woofer, and well into break up modes and beaming by that point

Would be very wary of using linns system to drop the xo that low though. Naim surely set the xo point there for a reason, and I wouldn’t lower it very far myself without taking proper distortion measurements and checking what strain it’s under.

Going from 2.7 to 1500 is nearly an octave, which requires 4x the excursion from the driver. It’s a lot to ask from a 19mm tweeter
 
I always thought the xo on sbls and sl2’s was 2.7k, and snaxo the same. Will get around to measuring it sometime soon hopefully, not that hard to do with such easy access the external xo, surprised it hasn’t been done here already.

I’ve seen some figures thrown about like 1.5 or 1.7k for a lower point. Which of course would be beneficial since 2.7 is way too high for an 8” woofer, and well into break up modes and beaming by that point

Would be very wary of using linns system to drop the xo that low though. Naim surely set the xo point there for a reason, and I wouldn’t lower it very far myself without taking proper distortion measurements and checking what strain it’s under.

Going from 2.7 to 1500 is nearly an octave, which requires 4x the excursion from the driver. It’s a lot to ask from a 19mm tweeter

I agree with your caution. The specs on the D2008 suggest 1500Hz would be at the lower end of what would be feasible. I think something in the 1750 to 2000 range is worth experimenting with. But it really would be good if some is aware of Naim's thinking at the time.
 


advertisement


Back
Top