advertisement


The Jim Current Source Mod.

Hi,

When I originally built this mod, I used (per current source, so you will need 4 or 6 times):
-Green LED
-220R resistor
-22KR resistor

When I calculated the resistor values I used the Vf=2.5V from the datasheet - but FD is right - when you run the LED at 1mA it has a Vf of about 1.9V

I'm still running the 220R resistor - this means I am using the current source configured for approximately 6mA - and I prefer it this way.

If you want to run the current source at 9mA as naim intended, simply change the 220R to 150R.


Finally: Filtering the 22K resistor is good, but I'd suggest just implementing one mod at a time.
Adding the filter will not gain you as much as it did with the old naim current source, since the LED current source has approx 20dB better rejection anyway.


Jim.
 
That type of diode biased constant current circuit could perhaps be improved with filtering between the diode and the transistor base. Cheap and easy and it has the benefit of reducing any diode noise too.
 
Jiiim said:
Hi,

When I originally built this mod, I used (per current source, so you will need 4 or 6 times):
-Green LED
-220R resistor
-22KR resistor

When I calculated the resistor values I used the Vf=2.5V from the datasheet - but FD is right - when you run the LED at 1mA it has a Vf of about 1.9V

I'm still running the 220R resistor - this means I am using the current source configured for approximately 6mA - and I prefer it this way.

If you want to run the current source at 9mA as naim intended, simply change the 220R to 150R.


Finally: Filtering the 22K resistor is good, but I'd suggest just implementing one mod at a time.
Adding the filter will not gain you as much as it did with the old naim current source, since the LED current source has approx 20dB better rejection anyway.


Jim.
Click to expand...

Ok, point taken, one step at a time and all that.. . I wonder if it would be possible to build it on a bit of strip board with flying leads, for easy removal, when experimenting with different resistor values and caps etc?.. . As usual, will report back.. . Cheers guys
 
I did this mod today and i like it, but its not night and day compared with the Avondale mod. Im wondering if its because of the different caps ive used. Ive been living for about a week with the 62k resistor split (4k7 / 47uF mill spec tant / 56k) now ive got 22k followed by a 10uF Oscon sp. When i first did the Avondale mod i used a 47uF 20v Oscon but changed it rather quick when i noticed. . .
 
. . . The voltage was wrong, But the sound had more life with the oscon. If i try splitting the Resistors, would i be able to put a film cap in between (10uF) or would this cause problems? . . please dont beat me up for asking (he says, hiding under the table)
 
If anyone is wondering why my post is split in two, its because i only get 400 bloody letters on this mobile phone.. . if my post makes bugger all sense, im sorry but im having real problems editing it on this shit bastard mobile phone
 
fatmarley said:
If anyone is wondering why my post is split in two, its because i only get 400 bloody letters on this mobile phone.. .
Click to expand...

I'm totally impressed you post via a mobile....I have enough trouble with a full size PC, keyboard and mouse!

Dave.
 
Wilky said:
I'm totally impressed you post via a mobile....I have enough trouble with a full size PC, keyboard and mouse!

Dave.
Click to expand...

I get even less letters when i quote someone.. . Christina aguilera's sounding rather nice/sexy as i write this. . . .
 
Film caps are in (10uF MMK), between the two 11k Resistors. The sound hasnt got as much life as it did with the Oscons, but i think its more natural. I replaced the 22uF tant that was decoupling the input bias Resistors with a 10uF SMR film cap and vocals now sound much more realistic. bass seems to be hitting harder and there seems to be more ambience.
 
Id like to give this a bash- could anyone simplify this mod for thickos who cant do schematics? I cant see much resemblence to my 321 gain boards, and Jiims 2 original schematics seem a tad different to each other to me. They look good in the dark tho I cant make head or tail of which bit goes where! cheers Captain
 
I printed off Jiiims schematic and a standard 321 schematic off the Avondale website. I converted Jiiims mod too the standard schematic, took the lid off the pre amp and found the 68R resistor, checked it was the correct one with a multimeter and replaced with the 220R. The rest is a bit hard to explain, but if you follow the circuit tracks on the board its not that difficult to do.. . .
 
Ooh, Remember you have to remove a transistor and replace it with the led.. . Im tempted to try the Oscons again (one day) because they really rock.. .
 
fatmarley said:
Ooh, Remember you have to remove a transistor and replace it with the led.. . Im tempted to try the Oscons again (one day) because they really rock.. .
Click to expand...

Aye - tis why I like this mod - remove two resistors and a transistor - replace with 2 new resistors and an LED - add a cap accros the LED if you fancy.

Here's some diagrams that may help.

Before:

321_gain_board_schematic.jpg


After: (with some bits missing cos I was lazy when I drew the diagram)

NA321currentsourcemod-1.jpg



Ignore the reference designators - they're not consistent between the two diagrams.
 
Cheers Jiilm.

Hang on a mo- whats being replaced by what.. Fatmarley says the old 68r with a new 220r. Above pic says Q5 with the 220r-?

Also is Q5 (top pic) same as TR3 (acoutica 'new 321')??

This is really doing my brain in.. it has to sleep now.
 
martin clark said:
I was going to suggest a similar thing PD, then looking again, couldn't you simply remove Q5, Q6, R7 and R9, and simply replace with a J5xx (or, cheaper, something like a selected BF245B (c10mA) with drain connected to emitter of Q2, source and gate to 0V). That could entirely eliminate one path for rail noise into the output.

Either way, nice going Jiim, I can see that would be a quieter ccs.
Click to expand...

Would it be possible for one of you clever buggers to draw a diagram of this?
 
Apologies - I flipped the position of the current source on the diagram, since it was simpler to draw it that way round.
The connections are the same. I didnt mean it to be confusing.

Here's the changes:


PHP: 
Before   After
62K       22K
68R       220R
Q5        LED

Q4 on 'before' diagram corresponds to Q6 on 'after' diagram.
Follow the connection lines and all will become clear.
On a schematic, the position of a component is irrelevant - it's the connections between them that matter.

Jim.
 
You must log in or register to reply here.


advertisement


Back
Top