Threshold NS10 Lineamp PCB

[apassgear]

I decided to open yet another NS10 thread, this one is about building a lineamp as close as possible to the original thing with some update ideas.

This has been possible thanks to R-K (Rolv Karsten) who posted a schematic he kept for 27 years that appeared to be faithful to the original in words of Nelson Pass, besides that there were many other valuable contributions such as close up pictures of the naked boards (sans cans) of the line and RIAA PCB’s by Malotron and others. Thanks to all you great guys again.

My Take on Threshold NS10

The idea behind this layout might not be for everyone since I took some “non audiophile” route regarding input and output caps which are not MKP’s but had my reason to it. Laying film caps on the board crated an awful appearance and did not look neat and ended with a big PCB for each channel. So I decided to keep the original spirit and go with lytics and thought that today’s caps are way better than 25 or 30 years ago, so why not?... use some BG or Panasonic FC’s which are way better than the old ones and in some cases being preferred to film caps by some DIY’s. The only big film cap is at the rail decoupling.

Since I wanted short traces and a compact layout I also decided to use the smaller ½ W resistors with an over all length of 0.360” (9.15mm) so no Dales RD-55’s, though I know in some places on the layout it is possible to add longer lead resistor and could add some pads, but for the time being only short resistors are possible.

All 3 signal BJT are as close as possible and the overall signal path is as short as it gets with these type and size of components which was one of the premises I started with, so I expect that the overall performance of this layout be quite good.

The overall size of the PCB is 3.110 X 2.125” (79 X 54 mm) one sided and uses no jumpers so its easy to etch and build. Most of the traces are 0.035” (0.89 mm) smaller pads are 0.080” (2mm) really don’t know how difficult could be to etch this by oneself with the regular photocopy transfer process. Any ideas?

Related to the bias opamp I decided it should have double rails for best stability and this enable to use any type with standard pinout you may have at hand, rails then should be 18/15 +-V. It’s a single unit DIP so one on each channel as the Master did.

I will welcome all suggestions so please chip in to make this layout better.

[apassgear]

Two components have been added to the original Schematic posted by R-K, these are C11 and R15 which can be deleated and jumpered if desiered.

[Zen Mod]

whoa !
another sheeeeeeeeetttty sounding preamp!



just go for it

[steenoe]

Great work Tony Thats a nice board layout.
Even though our other NS10 "clone" sounds really great, I agree that its much more fun to have a closer clone, like yours.
What do you have in mind for the PSU?
I am getting a Threshold NS10 this week. I convinced the owner that it would be better off at my place. Cant wait to get my hands on it. I am lucky enough that it comes with the moving coil box and the manual.

Steen

[Zen Mod]

Originally posted by steenoe
Great work Tony Thats a nice board layout.
Even though our other NS10 "clone" sounds really great, I agree that its much more fun to have a closer clone, like yours.
What do you have in mind for the PSU?
I am getting a Threshold NS10 this week. I convinced the owner that it would be better off at my place. Cant wait to get my hands on it. I am lucky enough that it comes with the moving coil box and the manual.

Steen

it's always nice to have piece of history,especially when it's working and when it works jolly good

I hate ya,ya lucky bstrd!

[Ipanema]

Hi,

Would NS10 be able to drive a 32ohms headphone? Any modification needed?


Thanks.

[apassgear]

Originally posted by Ipanema
Hi,
Would NS10 be able to drive a 32ohms headphone? Any modification needed?
Thanks.

The NS10 has about 1K output impedance and low power so it wont drive any headphone. For this you would need to add a buffer to the output.

I don't have a ready solution for this, but will think about it.

If you only need is a headphone driver it will be best to tackle an amp that was specificaly design for this purpose, there are many projects on the forum you could try.

Hope you find a solution

[apassgear]

Here is an updated version of the layout.

I have added R20 to test loading a bit the Opamp, which of course is optional. This ideas come from Walter Jung and John Curl.

R19 has change position to clean traces arround the Opamp.

All traces are now 1mm which would be easier to etch.

I will follow up with a corrected schematic

[lumanauw]

Even though our other NS10 "clone" sounds really great, I agree that its much more fun to have a closer clone, like yours.

Anyone who has built the "original" NS10, please inform how it sounds

I've been thinking about the original NS10 schematic that RK has posted (Thanks to RK). I think it will sound quite different from the "clone". I haven't built it, though.

First is the CCS (Q4 in the above schematic) as load. This will give less "distortion". Wheter you will like sound with more or less distortion, you should try both of them

The value of R8. With bipolar as Q1, 3k3 will give quite clean "noise" characteristic. I've been confused about this myself.
I refer to using Jfet as Q1. John Curl said that there existed phenomenon of "starving for current" when using Jfet. I experimented with this and appears that when Jfet biased with very low current (uA), it gives very noisy operating characteristic. You can read this in any small Jfet datasheet, in the first uA, Jfets always the noisiest, and very high peak noise in this area.

I use 10K for R8 and 2SK30 for Q1. For me (not speaking for anyone else ) this sounds very good. I've wonder why is this "wrong" implementation (to textbooks ie: noisiest) seems to sound better to me? The "correct" way to use Jfet for Q1 is biasing it very close to it's Idss (that is below 2k2 for R8). Later I don't care if textbook blames me of how I use this Jfet (starving for current), cause I feel it just sounds good
Then I think, LP's have very big noise too (compare music to "silence" noise), and still many still take LP as the best sound source, eventhough it is the most noisiest.

[steenoe]

Anyone who has built the "original" NS10, please inform how it sounds

That would be Threshold
Interesting points on the J-fet implementation, Lumanauw. I think I will try and test some of your ideas this weekend. I have some of the 2sk30's on hand.
Tony that board looks great, nice and tidy layout. You are more than welcome to post a pdf to scale of the copperside

Steen