advertisement


Upgrading Yamaha NS-1000Ms

Having now had the pleasure of listening to James' modified NS 1000s, I have to agree that Troels' crossovers absolutely transform the old Yams.They sound like a completely different speaker, which effectively they are. Notable qualities are a very natural tonal balance, unforced detail, and the ability to follow separate musical threads with ease. Seemed to me that the new crossover allows those remarkable Be drivers to really shine. Listening to the mighty NS 2000s shortly thereafter, the latter sounded very coloured by comparison!
 
Crossovers make or break loudspeaker performance. Not that the original XO broke the NS-1000Ms, but it didn't bring out their full potential either. These Troels designed XOs with Jantzen components are worth the time, effort and money to invest in.
 
A comparison between the RFC and Troels' version would be interesting, but I would need directions to Middle Earth.

I've bought in a pair of NS2000s now but don't consider them in any way coloured in comparison to my NS1000s. They did need the Lpads dialling down significantly though, something I'm not used to having to do now having had them removed in the RFC crossover.
 
The L-Pads on my then-unmolested NS-1000Ms needed dialing down as well to get barely enough bottom end. The Troels XOs do away with the L-Pads, which worried me at first. I needn't have, because his recommended resistor values are spot on for my room.

Directions to Middle Earth? Head to 41.0375° South, 174.8855° East.
 
Rusty I'm intrigued by the comment about the NS2000 being coloured in comparison to the Troels' NS1000, can you expand?
 
Paddy said:
Rusty I'm intrigued by the comment about the NS2000 being coloured in comparison to the Troels' NS1000, can you expand?
Click to expand...

Nothing to expand on really. I'd never considered the 2000's as particularly coloured, but listening to them very shortly after James' NS 1000s I was taken aback at the contrast. I guess it just goes to show how neutral Troels' crossovers are.
 
Thanks Rusty, I guess without hearing the Troels crossovers I will thankfully be deaf to the colouration then :)
 
200w.gif


 
sq225917 said:
Mmm 8db FR? Errr they're within 3db from 50hz-15k. That pair are shagged.
Click to expand...
The perils of designing a crossover upgrade for a 40-year old speaker using a sample size not even of one pair but half a pair! :rolleyes: I have enough experience of vintage speakers to know they are more than likely to drift off factory spec over time and the drift can be very different from one pair to the next...
 
Hard to argue with his measurements of the raw drivers in the box.

Seems like the standard crossover has minimal BSC. He doesn’t mention speaker positioning with and without the revised crossover :(
 
Anyone who has heard a decent pair of NS1000 knows that they remain a top-drawer speaker. I am sure that spending a grand on the crossover can change the sound, and good luck to those who wish to go down this route.
 
Has he actually made a crossover or just modelled it in software and then run it active to replicate how it sounds?
 
ToTo Man said:
The perils of designing a crossover upgrade for a 40-year old speaker using a sample size not even of one pair but half a pair! :rolleyes: I have enough experience of vintage speakers to know they are more than likely to drift off factory spec over time and the drift can be very different from one pair to the next...
Click to expand...

It would be very interesting to see some measurement data for multiple examples of the drivers. Off spec is a very real thing. Sadly I know it well from some of my favourite speakers (Tannoys, BBC monitors etc). As an example designing a crossover for my JR149s as found would have resulted in two entirely different schematics as the Kef drivers had drifted so far apart from one another. They all “worked” in as much as sound came out, but that was about it. The speakers were certainly incapable of generating a central image.

The video struck me as roughly what one would expect from a salesman pitching an $800 replacement crossover for a speaker. I was impressed how he managed to entirely ignored the level controls too. Anyone familiar with this speaker knows they are there to be used and while ‘flat’ may work soffit-mounted in a typically over-damped 1970s studio control room it sounds all kind of wrong in a typical home.
 
oldius said:
Anyone who has heard a decent pair of NS1000 knows that they remain a top-drawer speaker. I am sure that spending a grand on the crossover can change the sound, and good luck to those who wish to go down this route.
Click to expand...

And that’s the key word - “change”.

This does not always necessarily mean “improve”.
 
Rug Doc said:
Has he actually made a crossover or just modelled it in software and then run it active to replicate how it sounds?
Click to expand...
AIUI Danny designs a passive crossover circuit and swaps out components on-the-fly to observe the effect on the measured FR and phase and keeps tweaking until he obtains his desired target FR and phase using the least possible number of parts in the signal path. He warns against modelling crossovers in software because it doesn't account for the drivers' acoustic response in the baffle and enclosure they've been installed in.
 
^glad to hear. I really can’t understand people who tune by graphs.

I would like to hear what the original owner has to say about the changes from original. It would be FAR more engaging to see plots, in room, from a pair of speakers before and after rather than just a mono.

when we used to tune in car, you could tune one side by ear (and an Audiocontrol spectrum analyser (which I still have btw!)) and then the other side, both in mono, and then when running the two together would then need more adjustment as the two sides combine to increase the output too much at certain frequencies.
 
ToTo Man said:
AIUI Danny designs a passive crossover circuit and swaps out components on-the-fly to observe the effect on the measured FR and phase and keeps tweaking until he obtains his desired target FR and phase using the least possible number of parts in the signal path. He warns against modelling crossovers in software because it doesn't account for the drivers' acoustic response in the baffle and enclosure they've been installed in.
Click to expand...

That does make some sense. It's incredibly easy to make mistakes when measuring and modelling, and there are so many variables - I'm not sure software is 100% foolproof either.... Saying that, VituixCAD (when used properly) is supposed to be the far better for modelling than previous options like LspCAD.
 
I'm still amazed that a company with the resources on hand to design and build beryllium mid and treble drivers from scratch found a passive notch filter beyond their capabilities...
 
You must log in or register to reply here.


advertisement


Back
Top