Nap 140 rebuild project and grounding questions

[Bigb3n]

Skip further down if you're not interested in the back-story.

About 10 years ago, as a poor student, listening to my uncle's nac 82 + nap 250 amp set up got me scouring ebay to get myself some naim amps as cheaply as possible.
I ended up with the remenants of somebody else's projects picked up pretty cheaply:
- A NAC42 without the sleeve or volume pot and with a dodgy source selector
- The insides of a NAP140:


I got a case built to original specs off ebay and fixed up the NAC42 enough for it to be functional if ugly (and it still is).
I soldered some connectors onto the NAP140 guts and it played music through some rubbish test speakers. Then I got a job, got busy, bought a nap 160 and got on with listening to music.

I'm trying to finish off half done projects at the moment and the NAP 140's guts have been in a box nagging me to do something with them for ages. I have the sled half of a case from the NAC42 rebuild which will get me going but I'll need to try to procure a sleeve or a suitable alternative at some point. I've drilled most of the required holes with a terrible quality corded drill and press. I'm in the middle of replacing the bulk caps.
The boards have some damaged traces where the previous owner replaced some capacitors and made what looks like some minor modifications.


Here's where I could use a little help:
(I tried embedding or linking to images on imjur but my post was rejected due to inappropriate or spammy content!)

I'm having some doubts about whether the boards are originally from the same amp and have some questions on ground wiring.

The power supply board is a NAPS140/3 and the amp boards are NAPA5/7. From browsing google images, it looks like some NAPA boards used in the NAP140 have three connections on the "front" end while some earlier boards only have two. It looks like the two connections are always used for the signal and its shield and the additional third connector is wired to the supply board where it is connected to ground through a cap on the boards which have it.
My supply board has the traces and pads for these extra ground wires but the amp boards don't have the third connectors.


What is the purpose of the third connection? Do I need it if the amp boards are lacking the connector?
Where should the signal shield be connected to? One of the DIN connector pins? The case?
The previous owner has moved the ground wires to remove the caps from the circuit. Why do this?


Is there anything else I should be worried about with these boards given they have a questionable history?
Any components I should definitely change or tests I should perform?

Thanks for reading if you made it this far!

 

[MJS]

What is the purpose of the third connection? To keep the amp stable and prevent power supply noise injection

Do I need it if the amp boards are lacking the connector? If they're using NA001 transistors then yes

Where should the signal shield be connected to? The 0v connection on the amp front end, and the DIN plug 0v connection
One of the DIN connector pins? Yes The case? Absolutely not

The previous owner has moved the ground wires to remove the caps from the circuit. Why do this? Some misguided attempt at improving something that Naim had to bodge in order to keep the amp stable and noise-free perhaps?

Look, the 140, and 180, when using the later 250 boards had to have a few fixes to prevent ground loops and injecting the capacitor charging spikes into the front end of the amp. They did this by adding a capacitor from the front end 0v connection to the star earth on the power supply. The 140 used physical caps, the 180 used a pair of wires tied in close proximity to the 0v loom.

Putting the guts of the 140 or 180 in a new case without understanding the rather strange earthing topology that Naim had to use to make it compatible with the power supply board and DIN input is fraught with danger. At best you'll have a humming, noisy amp, at worst it will be unstable and fry itself. In order to test it you'll need an AC millivolt meter and scope to look at the noise waveform at the output of the amp. It should be a sawtooth due to the rather limited PSRR of the amp circuit. if that's good then you'll need to check the amp under varying levels and loads with a distortion meter to look for ultrasonic oscillation - these amps run on the knife edge of stability. Sorry if this isn't just a "put in the box and make these connections" type post because Naim did that, then they had to modify it to keep it happy.

 

[Bigb3n]

Thanks for the quick reply! That confirmed most of my expectations. I was mostly unsure about the third connection/mismatch between the amp and supply boards.

So the return from the pre-amp supply and both the signal channels all use the same DIN pin? Seems like a bit of a compromise.

What is the purpose of the third connection? To keep the amp stable and prevent power supply noise injection
Do I need it if the amp boards are lacking the connector? If they're using NA001 transistors then yes

My amp boards have the Sanken 2SC2922 transistors so I guess that implies that it isn't needed.

Where should the signal shield be connected to? The 0v connection on the amp front end, and the DIN plug 0v connection
One of the DIN connector pins? Yes The case? Absolutely not

That's what I thought, the extra wire just confused me.

The previous owner has moved the ground wires to remove the caps from the circuit. Why do this? Some misguided attempt at improving something that Naim had to bodge in order to keep the amp stable and noise-free perhaps?

Some kind of experiment seems likely, yep.

Look, the 140, and 180, when using the later 250 boards had to have a few fixes to prevent ground loops and injecting the capacitor charging spikes into the front end of the amp. They did this by adding a capacitor from the front end 0v connection to the star earth on the power supply. The 140 used physical caps, the 180 used a pair of wires tied in close proximity to the 0v loom.

Thanks for the explanation.

I'm a control systems/software engineer with some power electronics experience (from working on automotive traction motors). My analog circuit knowledge has diminished a lot through lack of use but I'm keen to learn. I understand that I might need to do some work to get this in working order. That's part of the appeal!

I have access to a couple of reasonable scopes at the local hackspace (not quite the TDS4054 I used to work with). I'll have a hunt for resources regarding the stability. Sure there's already some info out there.