advertisement


Naim CD3.5 mods (not on Acoustica)

Hi John,

Thanks for the offer of hand holding. It will be a slow process for me due to time constraints but at least i have spare boards to mod. I can understand things once i get my head round them. I will study the boards and report back exactly what components i have in all the positions.

I remember hacker years ago saying the current draw of the CD3.5 anologue stage was 93mA but i don't know if that included the original 317 regs or the two alwsr replacements or just the anologue stage without their regulators.

I suppose i will have to work it out for myself. I dont think there are any series resistors between regs and opamps so i may have to insert 1R to measure voltage drop and work out current via ohms law. I can't think of another way to do it.

The CCS mod sounds like the one hacker put on his hackernap boards.

Stu
 
mega lord said:
Finally plucked up the courage to take the the swing out disc tray out of the player and took it apart to see what was lurking inside.

Antistatic measures adhered to including antistatic mat and wristband.

One of the hex screws was a little stubborn to undo but moved in the end after i swore it a little. The assembly came out easy enough and then 8 screws to remove to gain access to the treasures within.

Pulled the pcb out with the transport on and had a good look. First thing to note is there is one lm317 reg in there with resistors i measured in circuit to be 162.9 R and 259R. But the colour codes suggest these are 200R 1% and 600 R 1%. This would suggest a voltage output of 5V.

Also on the pcb are a couple of chips - lm324n and ba6840bfp. Will find out what they are tomorrow, google is your friend. Also on the pcb are a handful of the usual blue 10uf 35V tants, 0.1 uf wima's, a couple of transistors (ztx 384c and ztx214c), a few other caps and 26 resistors.

Pictures to follow when i have found the camera and if anyone is interested. Is anyone interested by the way who maybe has not seen what is inside the pull out tray ?

Stu
Click to expand...

Stu, yes please, a photo would be great... And did you change anything here ?
 
Hi Steve,

I took some rubbish pics last week but have not had the time to transfer them to my pc yet. I promise i will soon. I am such a numpty when it comes to technology.

I am a bit wary of messing about in there, so no i have not done anything. All i could think of would be to do the led acoustica mods to the lm317 and give it a bit of decoupling as it is so far electrically from the main smoothing caps. The track has to go round the edge of the motherboard, past the RAM chip and along to the ribbon cable.

Not sure if the thing would benefit from a flea style gyrator before the lm317, Martin and others assure us there is no benefit from improving supplies to the digital parts. But the thought still niggles away at me! I would hate to muck anything up by playing with something i don't have to. There may be timing issues with other digital devices if the time constant is increased. Who knows, i certainly don't.

Currently trying to get my head round some modding of the two ALWSR powering the anologue stage as generously described by John.

Stu
 
Hello John (if you are watching),

When you mentioned remove C9, did you mean C8 which is across R3 ? (in the tpr)

I also dont know exactly where to cut the track for the opamp RC filtering. Would the resistor need to be on the OP2 side of C5's positive leg (presuming C5 is fitted) ?

Sorry, questions already!

Thanks, Stu
 
Hi Stu,

Just happened to drop by. Yep I misread the drawing, it is in fact C8 to be removed. bad eyesight! I cut the track between the +ve of C3 and the +ve of C5, just at the (only) 45 degree bend in the track. I then squeezed the resistor in on the back of the board between OP2 pin and the +ve of C3.

John
 
john.luckins said:
The ALWSR opamp supply mod will be better the larger the resistance (and cap) used. The AD825 uses typically 6.5 mA so I would suggest a 680 ohm resistance giving you about 19 v total opamp supply. A good value for C3 would be 47-100uF.
Click to expand...

Two things with this - one, such a large dc impedance gets me nervous with opamp supplies. you'll have a large 'pole' in supply impedance falling in the low-LF; in this case it isn't likely to 'motoboat' but it's not great IME. I have had strange audio effect arising in bass presentation which on paper, should be inaudible; a story for a different day.

In fact as little as 3.3R / 3.3uf works well with total stability (actually, audibly improved over the standard reg when I tried this with a tweaky homebrew preamp) There is a reason I suggest this as a minimum - the turnover freq is 15Khz, and thats where the AD825 loop gain starts to roll off. So guess what, with a min 3R3 and 3.3uF supply RC filter, its PSRR stays far more constant out to well above audio ;)

More in the direction John suggests will improve on this, but the AD825 isn't the quietest opamp around* so IMO there's no great urge point to push to extremes.

Oh, if you pick handy value in the region of 10ohms and 10uF + then the secondary benefit is that the opamps RC supply filter also looks exactly like a zobel across the output of this PSU (which is a DC amplifier with extended AC performance) - and help enhances stability into remote and low-ESR capacitors. Win-win ;)



* not that this matters providing the opamps noise isn't correlated with the output signal - which is exactly where the RC supply filter comes in
 
Thanks John for the clarification , i am gearing up to making a start.

Thanks for the input Martin. I will let John comment on your post as he knows what he is talking about :D

Stu
 
Hello john if you are watching,

I think i will start with the tpr mods you mentioned as that seems the easiest to start off with. My alwsr has 499r for both the tpr resistors as the lm1086 is used. They are both dale rn60 resistors (nice). Is 499r too high ? If so i have some 200r welwyn rc55y if you think they will do. I have loads of green led's i got for fleas and 317 meddling.

Next after that i will try the ccs mod and then try the opamp filtering when i have got some confidence built regarding messing with these wonderful regulators.

What do you think about Martin's comment above ?

Thanks, stu
 
Hi Stu,

Sorry for the delay. 499R should be fine for the upper leg resistor on the TPR. It'll give around 2.5mA for the LED and is unlikely to be more than is needed by the load (as it feeds the load directly).

Martin is right (of course) about the pole in the supply impedance at low frequencies but I don't detect a real problem here as the Opamps own PSRR is at its highest here. He is also correct in saying that the opamps weakness in PSRR is at its top end. There is a tradeoff here and I finished up with resistors in the feed of up to but no more than 680 ohms with no obvious bass issues. I may have missed them. They may be subjectively pleasing!? If you are worried then start with 10R and 10uF as Martin suggests and then increase and judge the sound for yourself.

I don't just put an RC filter in there now as I have found a zener based cap multiplier between the ALWSR output and the opamps supply to sound best of all. By happy coincidence this give an output impedance of about 5 ohms at the approx. 5 ma needed by the opamp and gives rejection right down to DC. It seems to me that the circuit is indeed sensitive to correlated noise from the output being fed back to the opamp via its supply. It is a high (OL) gain circuit and I suspect this noise to be disproportionately detrimental as it can be signal/load related.

By the way I use the same cap multiplier to feed the ALWSR voltage reference and that brings further small but noticeable improvements.

John
 
Good points all; I also found that feeding the reference with a bit more separation from the output worthwhile ( I use a ccs).

This is an example of what I meant about things that make no sense on paper - a resistive feed from a tightly regulated source, and a reference with a dynamic impedance of just an ohm should have a native rejection unfiltered better than -80dB ref to output; but a filtered feed to it made for quite unsubtle improvement in clarity and good behaviour.
 
Don't ignore ground injection. CCS massively reduces noise currents injected into earthing lines relative to resistor.
 
Wow, three experts posting in my thread one after the other. Thanks all .

Right, i will leave the 499r dale rn60 and replace the other resistor with a plain jane green led. Oh and remove c8.

Talk of zener based gyrators has me intrigued. Where does the zener go ? Would a ccs be better.

I haven't even done the basic rc filter mod yet and there are already talk of upgrades! I know in time i will have to try them (with your help guys i hope) as Les W said to me on the phone 'we always want more don't we'.

I will try the rc mod first (before zener gyratorz or ccs thingies) as long as cutting the track does not make problems later on.

Exciting times but i need to buy some more desoldering braid and components from farnell. Not another farnell order.

Thanks all, Stu
 
SIDETRACK

Ok, for Steve a few pics of the insides of the 3.5's swing out tray.

DSCF7321.JPG


A bit blurry cos both my hands were shaking. I assure you that is an LM317T.

DSCF7322.JPG


Can just see one of the IC's in the top left hand corner.

DSCF7323.JPG


View of motor, worm gear and sled etc

DSCF7325.JPG


Other side of the pcb with the other IC. Excuse my grubby farmers hands.

DSCF7326.JPG


The carcass.

Hope this is of some help Steve, excuse the crap photography.

One thing that worries me is that this swing out tray is more or less a sealed compartment apart from where the ribbon cable exits. I put my hand on the tray (above where the lm317 is) whilst the thing was playing chemical brothers and it did feel warm. Maybe get the drill out and drill some heat escape holes in a nice pattern of my choice. ;). Oh and maybe some bitumen dampening sheets in there too.


Stu
 
Hi Stu,

Very handy pics, thank you. My guess would be that there is little benefit in playing with the electronics in there, but a judicial amount of damping may be worthwhile. Too much may result in a rather 'thuddy' bottom end however.. But hey, what do I know?! Perhaps our experts will have an opinion.

As for drilling holes - Rather you than me. I doubt the buildup of heat will have much effect once everything is up to temperature. Holes will almost certainly affect the rigidity of the structure and as such will play with resonant frequencies etc. Combined with some damping though, who knows...
 
In a highly professional manner i went around my sytem with a compass to see if i could find any disturbances in the force.

No surprise that the biggesr deflections of the compass happened near the drive units of my speakers. This extended quite a way from the mid bass unit.

Next biggest deflection on the compass happened in the CD3.5. There afe two magnets in the player, one to help clamp the disc and one to hold the swing out tray in. The compass made wild deflections for quite a distance from the magnets, much much more than any transformers.

Well i thought the magnetic field couldn't be doing any good. I removed the magnet holding the swing out tray. I swear i can hear an improvement from doing this especially as my flea powered clock psu is right next to it. I put the magnet back loosely and listened again. I am sure i wasn't hearing things.

An easy mod to try. I will just have to make a wooden block to replace the magnet and stop the drawer being pushed in too far.

Stu
 
mega lord said:
Next biggest deflection on the compass happened in the CD3.5. There afe two magnets in the player, one to help clamp the disc and one to hold the swing out tray in. The compass made wild deflections for quite a distance from the magnets, much much more than any transformers.
Click to expand...

Transformers produce an *alternating* magnetic field (at the mains frequency). Permanent magnets produce a static magnetic field. A magnetic field does not induce an electric signal - unlike a *change* in magnetic field.

Well i thought the magnetic field couldn't be doing any good. I removed the magnet holding the swing out tray. I swear i can hear an improvement from doing this especially as my flea powered clock psu is right next to it. I put the magnet back loosely and listened again. I am sure i wasn't hearing things.
Click to expand...

I am sure you believe you weren't hearing things. The relevant question is "how do you know you weren't?".
 
And then there are interconnects that have large permanent magnets around them (I'm thinking Virtual Dynamics )...and others that purposely have a large static electrical field that surrounds them (Audioquest)...while conventional wisdom is to shield not expose.
 
Julf said:
Transformers produce an *alternating* magnetic field (at the mains frequency). Permanent magnets produce a static magnetic field. A magnetic field does not induce an electric signal - unlike a *change* in magnetic field.



I am sure you believe you weren't hearing things. The relevant question is "how do you know you weren't?".
Click to expand...

Thanks for the info about magnetic fields, something else learned

As for hearing things, i will have to have a serious word with myself !

Stu

PS John Luckins. I have not forgotten about the ALWSR mods, winter is a quieter time for me work wise so time for more invasive mods.
 
You must log in or register to reply here.


advertisement


Back
Top