Cases for NS1000 Cross-Overs

Hi-Fi discussion
Post Reply
MTPapaDan
Posts: 436
Joined: Fri Oct 25, 2013 7:52 pm

Re: Cases for NS1000 Cross-Overs

Post by MTPapaDan » Sat Mar 16, 2019 5:33 am

Wow! This is extremely powerful information. I have decided to go with the fixed values you calculated. No doubt it will sound better. I always left the Lpads at Normal and adjusted FR with DSP so I probably wasn't shifting phases? I hope..... At least the sweeps didn't indicate it. I was able to achieve sweeps closely resembling your final result +- 2 dB. Most likely a difference in room furnishings. I have a lot of soft surfaces. Couches, carpet, and window coverings. I'll get these values ordered ASAP. Will keep all informed of progress/result. Thanks for all your hard work!

reffc
Posts: 94
Joined: Thu Aug 30, 2018 8:40 pm

Re: Cases for NS1000 Cross-Overs

Post by reffc » Sat Mar 16, 2019 12:02 pm

The values which to my ears sound the most balanced, born out by the measurements, are the last ones posted, or as close to as you can get. I played around with several values yesterday and between 5R1 and 5R4 series with a 4R7 parallel sounds best with the mids and brings them down equal to the bass average envelope. HF is tailed off a little due to room furnishings but sounds natural and unfatiguing at 3R3 series, and between 5.6 and 6R parallel. Phase response is near perfect at those HF values so the tweeter is very well integrated.

The effects of taming the mid range is to raise the perception of bass. Whilst it hasn't got any higher, you've effectively sacrifice some mid level efficiency for a more balanced sound and the bass is definitely more prominent. This leads to claimed sensitivity. With these levels, it is slightly less I think than the 90dB figure but this needs testing for an input signal of 1W and measured at 1m. Talking of specifications, the revised specs, as opposed to factory claims are as follows:

Frequency response (tested) with bass inductor value of 5mH/0.25DCR: 47.5Hz (-6dB) to 15KHz (-6dB) IN ROOM FREE SPACE

Nominal Impedance: 5 Ohms (NOT 8 Ohms as claimed)
Minimum Impedance: 4.1 Ohms (100Hz) rising to 5.8Ohms (2KHz).

The colouration in sound in the upper bass is partly the break-up mode of the woofer causing a slight peak above crossover and partly that the cabinet is slightly under-damped. Adding a little more stuffing should improve things a little but there remains some boxy colouration which may be reduced a little with improvements to cabinet wall damping and a little extra bracing.

Valve amp users are advised to use the 4 Ohm taps. Phase swing is -62 degrees @55Hz rising to 45 degrees @45Hz so these are not a particularly easy load irrespective of an overall system sensitivity of between 88 and 90dB.

Valve amp users would be advised to use an amplifier in push-pull configuration of at least 40 to 50wpc...the more the better. I'd love to try Will Moores new (and rather fabulous) Radford design out with these. He's just completed the prototype of a self-biasing highly stable brand new valve power amp (stereo monoblocks) which kick out 100wpc and are stable enough to drive a 1 ohm load with lower distortion than even the STA25 and a very low output impedance (0.2 Ohms). He's configured the amp to use any power valve (or even a mix of valves) from EL34 to I think KT120s. Don't ask me how...the guy's got genius on his side. I may even persuade him to hold a bake off at my place so we can try it out on various speakers....watch this space.

As the 'speaker impedance isn't especially wildly fluctuating, SE amps may also work but would be best at slightly higher power ratings and with some negative global feedback. Forget 15wpc single ended amps...they's drive the speakers reasonably loud but would be breathless and soon run into clipping. In general, higher power amplifiers would work best and those able to provide decent amounts of current and cope with reasonably large phase swings.

reffc
Posts: 94
Joined: Thu Aug 30, 2018 8:40 pm

Re: Cases for NS1000 Cross-Overs

Post by reffc » Sat Mar 16, 2019 2:40 pm

Just out of interest I measured the original caps for the mid section (the 6 tied together). I'd have expected each of these to be around 3.5uF/0.3Ohms ESR for a low loss electrolytic tested at 200KHz. The average capacitance tested as 3.8uF (ie within 9%) but the ESR had risen to around 1.2 Ohms. The ESR shift isn't as serious as it sounds since in parallel this equates to around 0.2 Ohms total (another reason in favour of using bridged capacitors). The conclusion is that they were good quality electrolytics in the mid section. If the bass 'lytics were further out than this I'd expect that also to contribute to a shift in crossover and a slight change of response profile. I may take one out and test it as it would be interesting to see how the years have aged it.

Updated: pulled the bass ones out just to see. They were 6% over with an ESR of 0.3Ohms...spot on. That's as good as it gets for those so they've aged very well. There's still a big difference in SQ between factory and new crossovers so the L pads and other quality factors (inductors and spacing) probably account for much of the improvement.

YNWaN
Site Admin
Posts: 3940
Joined: Wed Apr 27, 2011 2:02 pm

Re: Cases for NS1000 Cross-Overs

Post by YNWaN » Sat Mar 16, 2019 4:36 pm

That parallels my subjective experience when updating my crossovers Paul. I felt replacing the inductors made a much more worthwhile improvement that the capacitors.
Mark

karma67
Posts: 222
Joined: Sat Sep 08, 2018 12:19 pm

Re: Cases for NS1000 Cross-Overs

Post by karma67 » Sat Mar 16, 2019 5:20 pm

paul can i ask what sort of cost for all this work is please? pm if you prefer.

reffc
Posts: 94
Joined: Thu Aug 30, 2018 8:40 pm

Re: Cases for NS1000 Cross-Overs

Post by reffc » Sat Mar 16, 2019 6:03 pm

PM'd

MTPapaDan
Posts: 436
Joined: Fri Oct 25, 2013 7:52 pm

Re: Cases for NS1000 Cross-Overs

Post by MTPapaDan » Sat Mar 16, 2019 8:07 pm

YNWaN wrote:
Tue Mar 12, 2019 6:54 pm
Oh, really, scratchy when adjusting them! Mine definately didn’t do that, so it seems like a thorough clean would be a good idea.

I mainly removed mine as a statement of purity rather than a genuine belief that they harmed the sound. Having said that I can easily make you some blancking plates if you want to remove them and replace them with fixed value resistor networks (I created blanking plates for mine but filled the holes in instead - I still have the CAD drawings though and can easily laser cut acrylic replacements.
Yes, please Mark. I will need two sets (4 skrs). PM me if you've lost my address. I really should build the external boxes and replace the inductors. Something to do in the fall. Too busy with car/shows. I've finally achieved a level of completeness I registered for 3 MT shows.

MTPapaDan
Posts: 436
Joined: Fri Oct 25, 2013 7:52 pm

Re: Cases for NS1000 Cross-Overs

Post by MTPapaDan » Sat Mar 16, 2019 8:26 pm

Paul, I really don't have time until fall to build external XOs. Can I set the Lpads to a particular value (after cleaning) to achieve a best case scenario for now? For series measure org to yell and org to blk? I don't think that's the case because when you adjust one you also change the other. If not I'll have to find time to install the new resistors. Who needs slept anyway?

reffc
Posts: 94
Joined: Thu Aug 30, 2018 8:40 pm

Re: Cases for NS1000 Cross-Overs

Post by reffc » Sun Mar 17, 2019 9:08 am

The pot doesn't work that way. As you say, you alter the position of the series, so the parallel shifts. I wouldn't mess with them internally until you have a chance to take them out of circuit. Best response to leave them at is the mids to -3dB (or about an 1/8th turn under) and HF to "Normal" minus a little bit...perhaps a 1/16th. That's about the best approximation you can get with the switches but even that' won't achieve the better results of proper L-pads.

MTPapaDan
Posts: 436
Joined: Fri Oct 25, 2013 7:52 pm

Re: Cases for NS1000 Cross-Overs

Post by MTPapaDan » Mon Mar 18, 2019 4:27 am

That's what I thought. Circuit cleaner arrives tomorrow. I'll clean the Lpads and set per your instructions. Resistors researched. Will try to get ordered asap.

Post Reply