My Take on Threshold NS10

[Zen Mod]

I am confused again - my bad habit.

Thanks for the reply, tube man Choky, and I believe yours must be good.

I was thinking to use NPN transistor and other components already in a junction box . . .

And, I was thinking that the arrangement of a zener regulation combined with an emitter follower will reduce zenner resistance, i.e. due to Rz/beta, so to reduce the ripple in a better way . . . In addition, the following source followers will purify the output voltage once again . . .

Could you give me a hint how yours reduces the noise?

Regards
jh

you don't need to reduce ripple from zenners,noise is what you need to reduce;
in Oly's PSU ,zenners are in colector leg of Q5 (generic BD PNP 100V min) and zenner noise can't be in any way amplified through that or any other transistor.besides that,with mentioned BD ,zenners are fed with CC,and consequences are simple and logical-less noise ,less voltage and temp variations
in your implementation,zenners are in base of NPN.....looks like no-no ?

 

[jh6you]

. . . and consequences are simple and logical . . .

Not to argue, but I do see the logic not clearly yet.
Maybe, I need somewhat detailed explanation . . .

Regards
jh

 

[Zen Mod]

Not to argue, but I do see the logic not clearly yet.
Maybe, I need somewhat detailed explanation . . .

Regards
jh

just build both versions and look with your CRO,or hear with your ears on your fullrange drivers on your horns in your favorite room........

or I'm wrong?

you know-I'm tooboolar in my head,SS is just a hobby for me,toobz are passion

anyone else want to comment and cut the chat of two greenhorns

 

[jh6you]

. . . just build both versions . . .
. . . I'm wrong?

Ya . . . maybe build both . . .
Why you are wrong . . . ? You are a good man.

It seems that other diYers have no interst in this.

. . . Anyhow . . . I redraw my thinking.
After the bridge rectifier, CRC and zener regulator follow. By the way, there will be still some of ripple voltage at the enterance of the zener regulator. But this ripple will be once again reduced by a voltage divider formed by the 47K5 in series with the small internal zener resistance. Maybe, it is not enough yet, so the 220uF cap in parallel with the zeners will clean up the remaining ripple further.

Meanwhile, the Emmiter follower will reduce the output impedance of the zeners so that this can hold the load voltage almost constant because the voltage source looks stiff. Finally, the additional Source followers will increase the current handling capability of the zener follower regulator.

I hope I am not speaking on behalf of a lunkhead . . .

Any comment to the details . . . ?

Regards
jh

 

[Zen Mod]

if we assume that your zenners are 1W3 each,then you must recalculate your resistor in series with zenner string;
1W3/9V1=142mA max per zenner and I recommend you to put some 5 to 10 mA through zenner chain;

in case that your input voltage is ,say,80V then we have :
80-(7x9,1)/0,01 =1K63 ,so you can use 1K6 ,not 47K as in your schematic.
with that value I doubt that zenners will be biased even slightly

gimme exact values for input voltage etc.......

I'm still for CCS-ed zenner string,hey-that's your reg,not mine

 

[jh6you]

Thanks for your calculation.
It should have been 4K7 instead of 47K5.
Sure it will be my reg . . .

OK . . . I will try the CCSed zener regulator.
I have to visit a shop for PNP . . .
I will make a drawing and post it.

Regards
jh

 

[Stabist]

Hi Stabist I just read through this wonderfull thread and think I missed your question Well I actually did compare the preamps with a ZenV4! Well, I have used my BoZ and my BosoZ and the X-CC-CCS-BosoZ as well as the NS10! Let me tell you, they all sounds great The BoZ family is a bit more romantic sounding than the NS10 I totally agree with Choky on that! The NS10 is pretty brutal as Choky stated! I like the X'ed BosoZ best But I guess thats a matter of taste, in the end I might add that the NS10 is way more powerfull than one would expect from a small transistor circuit like that It drives the Zen and the F2 with ease

Steen

Hi!

Cool, sound like NS10 would be ideal - I'm in a quest of a more brutal and not so romantic preamp Really! Not kidding!!

So - since I'm at home and have some stuff at hand ... Just to check some things:
- probably I'll go in with +/-25V PSU version - or maybe 0-50V - must check which transformers I have at hand ...
- I think I have most parts - except - the transistors I allready have are:
- mpsa42/92 "pair" and also some
- BC550C and mpsa56 (but no mpsa06).

So for Q2 I'll use mpsa 92 and for Q3 mpsa42 ... But what about Q1? I can use mpsa42 or bc550c ... Is it better to stick with mpsa or maybe better to use bc?

Oh and yes - at this moment I want to build it with those parts available since electronic store is quite far from me ...

Btw - which capacitors did you use (or others that have allready built NS10) in your versions?

 

[GeorgeBoles]

As I understand it:

Dear Stabist,

As I understand it, you can use nearly anything for Q1: whatever suits your ear and taste best.

I have used BC546 and 549 just to get the thing working. BC550, MPSA's all hae been suggested, plus some suggestions early in the thread coming from Nelson Pass himself, and others describing which has subjectively the best sound.

Here are the highlights I have plucked out. These relate to Q1 as well as the PNP transistor, if I recall correctly:

Nelson has tiped that the NPN devices should now be MPSA18 (On-Semi) I check on this and it is a modern replacement of the original MPS6571

Q1, MPSA06, MPSA42, 2n5401, BC550C, BC546, they sound differently. The best sound is 2sk30y (you can also buy 2sk30gr suffix).

I don't advise the following, but I do this : I pick up transistors with lower gain (HFE) in position of Q2 and Q3. I use MPSA42 and MPSA92 there. I don't use R3.

I just checked with my normal supplier, he has some of the sk30's. I also had some sk389, sk369 and the sk170

The post shows a single rail without resistor values using MPS6571 for Q1 and Q3 and 2N4250 for Q2. Then Nelson tipped a replacement for Q1/Q3 would be a MPSA18. I checked on the last one and it is an almost identical replacement from On-semi with better voltage ratings.

Originally I used a 2n4250 ... Nelson Pass

MPS4250 simms to be the replacement transistor. It's an On-Semi active part

I received 2N5087 and MPSA18.

No fancy parts yet. The first test will be to see, if this works at all Dont let Q2 confuse you, I have used a BC560C, which has a different pinout than the MPS4250.

I just used MPSA06/56.

Regards,
George.

 

[Zen Mod]

I use two of BC413 and third is .....?

search function is in this moment......

 

[Stabist]

@george - as you see from your quote (btw - thanks to gather together those infos) - the mpsa42/92 are declared to have a bit too low hfe ... so that's why I guess it's better to use bc550C for Q1 ...

I'll try to make a functional preamp in few days with what I have and we'll see

Btw - if going to make single supply version - isn't e.g. PSU from AL good for NS10 too if I "rearange" it to give 50V?

 

[GeorgeBoles]

Hello Stabist,

I cannot help you much more. I only built the circuit so I could test my LM3886 gain clone amps. Even with the BC546 as Q1 there is buckets of gain in the whole circuit (about 5 times voltage gain, but I understand so little about the real "workings" of the circuit that I cannot say whether low hfe itself is a problem. If Nelson Pass suggests MPSA18 as Q1 and Q3 and the 2n4250 (MPS4250) as Q2, that would be a good balance of parts to start with and experiment from there. I am sure that his original Threshold NS10 was a beautifully balanced and engineered device, with lots of subtle thought put into it.

I did not build the circuit properly into a shielded case with properly positioned shielded wires ... everything was just in a great big mess on the floor with mains wiring, power amp wiring and power supply, a boring LM317/LM337 power supply (from a transformer with too low a voltage: 21 V AC centre tapped - regulated to +/-25 Volts DC) - all the wires lying mixed up ... and there WAS quite audible noise. But that, I hope could have come from the mess and not having a "low noise" transistor as Q1.

Unfortunatley, I don't know what an "AL" is, if that question was directed to me.

Regards,
George.

 

[Zen Mod]

FOR LOCALL THREAD BOYZ AND GIRLZ

http://www.diyaudio.com/forums/showthread.php?postid=831920#post831920

and few posts lower


any comment?

 

[steenoe]

any comment?

Good thinking, Choky Are you gonna build the Aleph-J?

I don't know what an "AL" is,

It's an Aleph-L preamp.
Stabist, if you will use the devices you have on hand, the best choise for Q1 is the BC550 device. (I think I would use it for Q3 also) If you have a BC560 you can use that for Q2 (thats what I did) I think your AL psu will work, if adjusted to 50V. I did try a BosoZ supply (adjusted to +/-25V) with the NS10. Worked just fine. I will soon take up the shunt reg. again. I will try and make one with the 317/337 CCS.

Steen

Edit. Choky take a look here: http://www.diyaudio.com/forums/showthread.php?postid=795972#post795972 You mean to put capacitance
before the LM's right? I figured something like 2x1000uF or something. Not a CRC, is that correct?

 

[Zen Mod]

Good thinking, Choky Are you gonna build the Aleph-J?





Edit. Choky take a look here: http://www.diyaudio.com/forums/showthread.php?postid=795972#post795972 You mean to put capacitance
before the LM's right? I figured something like 2x1000uF or something. Not a CRC, is that correct?

yup,steen....

have you few pieces of 2SJ109V?


regarding caps before LMs
steen-that is your choice
whatever you have :graetz ,caps,chokes resistors....even series regs,zenner followers are maybe best solution


you know that you must have smooth caps
2x 100.000uF ?

 

[Stabist]

@george - as said - what I have at hand is a complementary pair mpsa42/92 and few bc550c.
Sure - ASAP as I go to our capital I'll have a chance to buy some other transistors - but - that's probably in the next week ... And I don't want to wait anymore
Well - mpsa18 and mps4250 are more tricky to buy here - that I'll have to order from farnell ...
Yep - AL is AlephL as allready explained ...


@stenoe - yep - that's what I thought - when choosing between bc550c and mpsa42 - the bc is probably a better choice ...
Hmm - are you sure it's ok to use mpsa92 for Q2 and bc550 for Q3 - aren't they a bit "too much different" (not complementary pair) ...

Well I'll see - maybe the protoype will be done allready today

 

[Zen Mod]

Primoz-there is no need for complimentary pair-ing

 

[Stabist]

Hmm, now I haveanother Q ... How to figure it out who's the manufacturer of transistors - I mean - I've checked datasheets for mpsa42, 56 and 92 and bc550 - and - it depends from the manufacturer that for same transistor the legs are or C-B-E or E-B-C ... (e.g.: On semi, Motorola, National etc have same marking, but Philips, KEC, etc the opposite????)

- for mpsa92 and 42 I'm quite certain that they are On Semi ones (since in on semi datashhet there ia an example how they mark them),

- then mpsa56 is probably Motorola (because of typical "M" on it)

- but I don't have a clue for bc550c ... a bit similar marking to mpsa92 and 42 but not sure ...

What is the way to figure it out?

 

[Zen Mod]

bc550 is always ebc
legs to you
flat side to down

 

[Stabist]

hehe - I've realized from where my confusion Silly me

Ofcourse - I should be more careful with datasheets where transistor is drawn 3 views (ground plan, side view, ...)
Geee, feel sooooo stupid right now

 

[steenoe]

have you few pieces of 2SJ109V?

Nope, not at this time. I might have a source though
With NP declaring the Aleph-J the best sounding of the Aleph bunch, I better start gathering parts

I should be more careful with datasheets where transistor is drawn 3 views

Stabist, you are not alone on that one Yes the BC's have different pinout than those on Russ's board.
Happy building.

Steen