advertisement


Regulating NAP front end

A

AR_sound

Trade: AR Sound.
Hi chaps,

Inspired from Teddy's "NCC Front end regulating" thread (http://www.pinkfishmedia.net/forum/showthread.php?t=20798)

I have implemented the front end regulation in my revised NAP.
The results are superb, It's remindes me the effect of Super regulating my NAC two years ago.

For the benefit of all the "bodgers", Here is the recipe:

1- Open your NAP and measure the voltage on the power supply capacitors-
This should be between 32v to 41v, depends the model (Nap 90 ~32v to
NAP135 ~41v). The Regulated voltage would be set to the rail voltage-5v

2- Second step- ingredients: -two LM1086 and two LM337
-four 150 ohm resistors (Out-Adj legs)
-four resistors (Adj-Ground), 3300-4500 ohm
according to the desired output regulated
voltage (3300 will give 28.7v and 4500ohm will
give 38.7v,(I chosed 4120ohm for 35.6v).
-four decoupling caps (Out-Gnd) - I chosed
22uF/50v military-grade Kemet Tants (worked
very good for me in several applications-
-four decoupling caps (Adj-Gnd)-
(100uF Rubycon Za/ZL will be fine here).

3- Third step- prepare The victim:

P3130570.jpg


Cut the Nap rails according to the pic (dremel will help in this case).


4- Handwire all the regulators/resistor/caps: (I know it looks messy, But it all beneath the PCB- you can't see nothing above and the connections are with minimum distance)-

P3150575.jpg


5- Check twice that everything is in place, pay attention that you have connected the 317/1086 to the postive rail and 337 to the (-) rail, check caps polarity (337 Adj and Out to (-)!).

6- Turn on voltage for short time, quickly measure the in/out voltages using DVM. Turn on again and remeasure (scope the in/out to see all clean from oscillations- recommended).

7- Increase the NAP idle current to 45mA, check DC offset at the output (speakers disconected!!).

8- Close the amp, put some music and be amazed :D

P3140573.jpg

(My "NAP"- An 600VA trafo with 1000VA core lies in a separate round case, Hexfred rectifiers and four 10000uF Kendeils- sorry- no space for 22000uF inside NAP140 case! ;) ).

Cheers,
Avi.
 
Avi, what a great post. Can't wait for the rest of the pictures.

Given that the voltage rails are 5v below normal, does this mean that we'll have to turn the volume up further on the preamp for any given volume level?

Nice work!
Carl
 
teddy_pardo said:
Why do you need four 1086 and four 337? I think you only need a couple of each...
Click to expand...

Sorry, of course only two.. I guess I had write It in a moment of bodgophilia ;)

Avi
 
hacker said:
Given that the voltage rails are 5v below normal, does this mean that we'll have to turn the volume up further on the preamp for any given volume level?
Carl
Click to expand...

No, volume will stay the same But the power of the NAP will be slight lower- see Martin's comment in the parallel thread (NCC...)

Avi.
 
I seem to remember LesW mentioning (forgive me if I'm totally off bass here) that a lowering of voltage can result in a softening of the 'Thwack' component of the sound & stuff like that.....

any effects like that apparent or does that sort of this only happen if you regulate the full board :confused:

wonder how it compares to fully regulated?

J
 
An obvious tip: Do everything with the speakers disconnected!!!!

Connect them only after measuring that there's no DC on the outputs..

Teddy
 
JimmyB said:
I seem to remember LesW mentioning (forgive me if I'm totally off bass here) that a lowering of voltage can result in a softening of the 'Thwack' component of the sound & stuff like that.....
:confused:
J
Click to expand...

Hi Jimmy,

My original NAP140 worked with 36v,then I replaced the trafo for 600VA 29v X 4,
and raised the rail voltage to 41v (as in the NAP250),
and now, after regulating the front end, I have 35.5v again, like the original voltage in NAP140, there is now problem with this and it sounds great! ;)

Avi.
 
Nice work, I'm still building my errr... NAP thing with some NAP250 boards, caps & rectifiers from Les so might have a go if I can sort out what goes where but the way I'm going that'll be years away!!
Maybe I can patch up a couple of vero boards with your solution in the meantime.

I remember Les saying (sorry Les - all my info is second hand from better men than I) that the NAP250 modules can take up to about 50V max so might push them hard (45V) for the hell of it.

For my ignorance, is the back end still seeing 41V?

J
 
What sort of current are the lm1086/lm337 likely to be deliverying when listening at very loud levels?

I'm also wondering about resistor power handling; do R1 and R2 on the 1086/337 circuits need to be capable of handling a certain number of Watts? (I suppose this can just be calculated if we know the current delivery from the first question).
 
Fractions of a watt, Carl; the adjustment resistors only pass a constant current, so power rating depends on voltage drop. Lower adjust resistor is the only item to watch - say 5mA x 40v drop, 200mW to disspate - needs to be 0.5w rated.

The front end only draws 25-30mAish (pretty constant) so again - since the 317 is dropping a handful of volts - power rating is no problem.
 
hacker said:
What sort of current are the lm1086/lm337 likely to be deliverying when listening at very loud levels?

I'm also wondering about resistor power handling; do R1 and R2 on the 1086/337 circuits need to be capable of handling a certain number of Watts? (I suppose this can just be calculated if we know the current delivery from the first question).
Click to expand...

The 1086/337 can deliver 1.5A, the current for the front end is very small (I have touched the regs while playing and they were cold (without heatsink),
I'm sure 0.5W resistors will do fine.
 
Damn. I was hoping I'd need special parts, making the job difficult and therefore not too tempting. As it is, I've got all the parts I need at home and am almost certain to dremel the f**k out of my 140 tonight.

Bugger.
 
good luck hacker, and be careful- dont connect the speakers before you have measured DC offset in the NAP output and you are sure everything is OK.

Avi.
 
On a slightly more useful note, most amps have substantially better PSRR on their positive rails due to using current sources there. As a result, the negative rails will benefit much more from regulation. RC decoupling will probably be plenty sufficient on the positive rails ... at least that's what my experiments showed me.
 
One last word of caution: check what happens when you switch on. As in, plug your meter in and switch the amp on and off a few times (or, better, plug into the soundcard and watch the trace a few times). More than a handful of volts for a couple of seconds and you might like to investigate further before committing your speakers.

I know most Naim kit is left on all the time, but if - as a result of PSU mods - you end up with the front-end rails starting-up at different times, the speaker will have to endure a DC offset until the amp sorts itself out. Variable amunt of 'WHOMP...pip' to be expected. Let's just say that steady DC does not endear itself to bass driver voice coils!

That said, in doing what Avi suggests (which is excellent ;) ), I don't see a problem arising - but this may not necessarily be true if applied to other amp topologies.
 
I dont want to talk voice coils :( it a sore topic, however ...

Could one (for temporary protection after such a mod) pop a capacitor in circuit to prove the theory before exposing the speakers to the full amp output, not sure?
 
Yes, but it would be huge - 10000uF would be a good start, just like all those 70s AC-coupled amps...So long as the DC offest disappears after a second or two I wouldn't worry about it.

(read about your burnt tweeter, not nice - but I'd bet it was killed by HF rather than DC. The squeal is the clue)
 
You must log in or register to reply here.


advertisement


Back
Top