advertisement


How to run SBLs?

One crucial question is whether there is anything inherently special in the fact the the woofer box sits on spikes. Sure, they minimise the contact area, but does this really make that much of a difference? Would any stable coupling of the woofer box to the contact points on the frame be equivalent?

I was thinking about this too. The complication with the SBL is that although Naim says the box B must rest on the spikes and not the mastic, the mastic nevertheless performs an important role not only in sealing the join, but also in sticking the boxes together. Without the mastic the least knock would dislodge it. So if one could devise a way to attach the boxes firmly together and then seal the gap it would be a more robust design.

Not sure about magnets in proximity to woofers?

Tim
 
Without the mastic the least knock would dislodge it.
Yet another reason why things need improving!
So if one could devise a way to attach the boxes firmly together and then seal the gap it would be a more robust design.
Working on it, I just need some input so I don't wreck things in the process...
Not sure about magnets in proximity to woofers?
Most of them work in reasonable proximity to other magnets in tweeters...?
 
So if one could devise a way to attach the boxes firmly together and then seal the gap it would be a more robust design.

But wouldn’t that effectively turn SBLs into BLs?

Not sure how it’d sound, but I’m pretty sure the design aim was to have the middle and bottom boxes as decoupled from one another as possible. Or at least more decoupled than directly coupled.
 
But wouldn’t that effectively turn SBLs into BLs?

Not sure how it’d sound, but I’m pretty sure the design aim was to have the middle and bottom boxes as decoupled from one another as possible. Or at least more decoupled than directly coupled.

This is the exact point. What is the engineering goal in precise terms and is it achievable without fragility?

Tim
 
But wouldn’t that effectively turn SBLs into BLs?

Not sure how it’d sound, but I’m pretty sure the design aim was to have the middle and bottom boxes as decoupled from one another as possible. Or at least more decoupled than directly coupled.
This is the thing... or at least one of them... by what definition are they currently decoupled? The base box would appear to be very firmly attached to the frame, but the woofer box? Conventional wisdom would suggest the spikes are there to decouple the woofer box from the frame while allowing minimal movement of the box.

I think this is the first question which needs answering: should the spikes be retained? If so, an alternative needs to be found for the aluminium plates. If not, then should we attempt to emulate their mechanical properties or not? Should we rigidly couple the woofer box to the frame, or attempt some other way to decouple it?
 
I'm picking up a pair of SBLs from south Wales when the lockdown allows me, so if anyone would like any further discussion on this, I'm happy to continue :cool:
 
I think we are keen to hear the results of your experiments :)
Hmmm... yes, well... they may be a long time coming, but I'm certainly not short of ideas...!

What I'd really like to do is put accelerometers (or at least piezo pickup type things) on the frame and boxes to measure the effect of any changes I make, and see how that correlates with subjective listening.

Using piezo pickups would give more more bandwidth but zero amplitude calibration - they're often peaky in the treble - but I would at least be able to make relative measurements of vibration levels across most of the audio band.

Accelerometers that cost under £750 typically have bandwidths of 1-1.5kHz at best, but will give absolute g-force measurements with 10-14 bit resolution. Noise can be a problem, but 'no music or footsteps' measurements can be taken, averaged and subtracted. I'm still searching for a sub-£100 accelerometer breakout board which will interface with a Raspberry Pi or Arduino and give a several-or-more kHz bandwidth. If improvements are worthwhile and obvious with a lesser accelerometer, it may be worth investing more. If I can find a piezo device with a well-characterised response, I could use a cheap accelerometer to calibrate it where the responses overlap, perhaps.

However, let me get them home, and perhaps replace the tweeters, and I'll take things from there.
 
Standard industrial piezo accelerometers will be flat enough in the audio band. Resonance is more like 30kHz plus. Well under £750 this is true. Largest limitation will be the mounting. Magnetic mounts, glue, all have an upper limit in frequency that will limit you to below audio range but that isn't a piezo accelerometer limit.
 
Just listen to them?
Where's the fun in that? Admittedly, this thread is probably more suited to the DIY section now, kinda...
Standard industrial piezo accelerometers will be flat enough in the audio band. Resonance is more like 30kHz plus. Well under £750 this is true. Largest limitation will be the mounting. Magnetic mounts, glue, all have an upper limit in frequency that will limit you to below audio range but that isn't a piezo accelerometer limit.
Yeah, I'm just learning about mounting. Any pointers for good piezo accelerometers? I've just found this, which will do full audio band, but I'd need three of 'em coming to £120... and then there's the soundcard interface and power supply to consider...

https://www.digikey.co.uk/products/en?keywords=EVAL-ADXL1005Z
 
Hmm that looks good, it isn't piezo but MEMS, they have come on a lot since I last looked at MEMS, but I suppose that was (mumble mumble) years ago. You will not get a piezo that cheap that has that bandwidth.
 
Why do you need 3? Tri axis is interesting but if you repeat the same test and mount one transducer 3 ways ... time spaced measurements are valid if nothing else changes.
 
The hierarchy seemed to be a best amp was better than 2 lesser ones, 135's better than active 250's, 250 better than pair of 180's, haven't done it myself so its hearsay from me.

Years ago I was at a few Naim evenings in a dealers. Twice I saw the Naim bods choosing to run active 250s instead of 135s when the latter was an option. They tried both and went with the 250s. I think one was introduction of the SBL, the other was the CDS. I did ask on one of the occasions why they'd done it but I can't remember what they said. I didn't get it and I still don't. The 135s are literally twice the price. Surely their should be zero contest?
 
Why do you need 3? Tri axis is interesting but if you repeat the same test and mount one transducer 3 ways ... time spaced measurements are valid if nothing else changes.
First rule of government spending: why have one when you can have three for thrice the price? Having three will save me a heck of a lot of time, probably a lot more than £80's worth... although thinking about it, impulse tests aren't going to take very long...

Also, with 3 axes, at least the axes will be consistently oriented in space.
 


advertisement


Back
Top