advertisement


Cambridge Audio 340A SE preout conversion and other mods.

A freind advised this little list, simmlar to what you said Dan:


*C4 C5 on U4 change to BG 16V 220uF (input selector chip)

*C25 C14 on U1 change to Mundorf MKP 2.2uF (signal path)

*U1 change to LM4562 (input buffer) optional.

*C7 C8 on U1 change to BG 16V 220uF (input buffer)

*C3 C37 on U7 U6 change tro Mundorf MKP 2.2uF (signal path)

*C40 C41 C42 C43 use a BG as large as can be fitted 35V+ (power chip local supply rail)

*C45 C46 change to something better again (like BG) 35V+ (20V rails before opamp regs)

*C49 C50 change to 25V BG 470uF or 1000uF (supply to opamps after regs)

*U8 U9 could do with upgrade (opamp regs)
C19 C29 change to Mundorf MKP 2.2uF (signal path)(tone bypass circuit)



I see the main differences are that you suggest bypassing C14, C25, C19, C29 with a wire link, and the above reccomends using Mundorf MKP's? I suppose if its OK to remove them then its better (and cheaper) to link them out?
 
I suppose if its OK to remove them then its better (and cheaper) to link them out?

Yes, that's right. There is no need to have so many DC blocking caps. You might be able to get away without any at all, but it's safe to have just one.

Getting rid of the feedback cap on the power amp will make a big difference too.

You don't need to worry about adding massive Blackgates everywhere either. Doing so may make a marginal difference (it'll make a big difference to your pocket!) , but getting rid of unnecessary coupling caps and changing U1 to a quality opamp will make real differences.
 
Sound advice. Thanks.

I found out that the tube module is self regulated, I just need a 12v 3a transformer. Still not sure if this entirely appropriate as I dont understand amps. The seller suggests using it between a CDP and Amp. Not sure if hes referring to a poweramp ir integrated/pre-power set up.
 
From what I can tell these things are generally used between a CD player and Amp. Im pretty certain I could replace the 'input buffer' in the CA with this module but im not sure If It will handle being fed by a variable output if I were to feed it from the Bal pot. I thought It was simple.



How do I do this preout mod?



I want to build Avondale NCC200's eventually, but first mod the 340 to drive them well. Others on fish have said just build a good preamp from scratch but I want to have remote function and input switching. Modding up the amp seems a decent option as the poweramps are apparently simmilar to Gainclones. Maybe I can build a tube preamp and fit it inplace of the input buffer?



It would help If I understood the circuit in teh 340A, which bit exactly is the preamp? (I assume its the 'input buffer, pots, and associated circuitry?)
 
Ive asked the seller for more info re,: power supply. All I got is that the module needs 12v, 2Amp. I asked for recommonded VA rating of the tx and if it needs dual wining or single, he just came back with the same answer: '12v 3amp'. Maybe his English isnt up to scratch. He was able to tell me that it has on board regulation, just needs a tx.

Ive asked for impeidance specs, as soon as Iget an answer I'll post.

While were on this subject, I wonder if you (or anyone else reading) knows of a simmilar module which would definately be suitable? It might even be possible to build a whole tube preamp into the 340 - driving the power stages and making use of the inbuilt switching and pots? Maybe a long shot but if the power stages are basically Gainclone sureley its worth it?
 
The route you are looking at is certainly feasible, bur you will have to add a PSU. There is another potential problem: getting the earthing right for such a scenario could be hit and miss - earthing problems will cause hum and fixing them can be troublesome even for an experienced constructor. I'm not trying to be negative, just warning you that there is no guarantee of getting it working well. This is one reason to recommend building from scratch, because then you have control over everything.

Either way, you'll need to know how to build a suitable power supply, find your way round a circuit board to cut the right tracks (if you stick with the CA), route your wiring correctly etc.
 
I understand.

Ive had a reply from 'DIY Gene' about the specs of this module. Basically he just sent me a schematic with chinese writing on it.......He obviously isnt too fluent in English. Right now I fell like I can t be bothered, its been crap morning so far though. He,He.

When you first mentioned that I might need to run the preout (post #12)
through a buffer, what eactly did dyou have in mind?
 
A simple unity gain op-amp based buffer should be sufficient, and would run off the existing supplies inside the CA.

Use either a dual op-amp, or two singles. For each section, connect signal feed to non-inverting input. Put a 100K Ohm resistor from the non inverting input to ground. . Connect inverting input to output. Route output to plugs.

BOM
2 x single op-amp (or 1x dual)
2 x 8pin DIP socket, (or 1 socket if using dual)
2 x 100kOhm resistor
wire, scrap of veroboard to build it on, solder etc.

If the point that you plan to use as pre-amp out is DC coupled, you might need a pair of coupling caps. in this case, the signal connection goes via the caps. Use 1uF polypropylene.

Parts should leave change from a tenner.
 
^^ What he said ^^

Here's your current scenario:

Input ---> Source selector ---> Buffer ---> Volume/Balance ---> Power amp

Everything up to Volume/Balance can be considered your preamp. Between the Volume and the Power amp is the place to insert a buffer. You don't actually NEED a buffer, just that over performance will likely be better with one. You could approach it like this:

1) Upgrade the CA, with caps, diodes, etc.

2) Build the power amp.

3) Connect the Volume in the CA to an output.

4) Listen and enjoy!

5) Fit a small buffer after the Volume pot, and evaluate it for any improvement.
 
Ok. Thanks!

I think I'll do that then. PS: decided to go for Avondale NCC200's rather that the NAP140 clones)

A question on power supplies:

Is it beneficial, and simple, to seperate the preamp section from the power. So each has its own supply? Am I right that this would increase the VA headroom available to the poweramps if the pre sections had their own supply?

(Maybe to solve the possible earthing problens you mentioned, I could build an external supply for the preamps?)
 
Yes, as a principle it is beneficial to separate power supplies for different circuits. In the CA the active part of the preamp - the opamps - are powered via a separate winding on the transformer, so that should be fine.

I'm sure you'll get very good results from the NCC200.
 
Its a shame I cant get any data out of the 'diy gene' guy. I really like the idea of tubes, possibly even replacing the input buffer with a tube stage. Or using one to compliment it.

'Tubeshunter has a few compact tube preamp modules and suitable power supllies. Link below if anyone would care to take a look and see if theres anything usable there?

http://stores.ebay.co.uk/tubeshunter_KIT-PCB_W0QQcolZ4QQdirZ1QQfsubZ10826462QQftidZ2QQtZkm

The first amp mentioned from DIY Gene is an ideal size. My guess is it would be best used between the input buffer and the pots. Shame I cant get any specs from him.

I can send a copy of the schematic if anyones interested.
 
This site contains affiliate links for which pink fish media may be compensated.
Here's what I suggest the preamp section of the CA:

1. Bypass capacitors C14, C25, C19, C29 with a wire link. This will take them out of the circuit.

2. Replace opamp U1. With something better e.g. LM4562 if you want an analytical sound or OPA2107 if you want warmth. Better still, fit a socket and try a few 'till you find which you like best. There is no universal best.

3. If you want to reduce gain to give better volume control range, remove R33, R34, R35, R36, C12 and C13. Fit a wire link across R33 and R35.

4. Upgrade D5 - D8 with schottky equivalents.

5. Upgrade C45 and C46 with a low impedance cap e.g. Panasonic FC 2,200uf.

6. Fit better regs at U8 and U9.

7. If you want to upgrade the amp section, replace C3 and C37 with something better such as Blackgates, and link across C22 and C38

Out of these, replacing the opamp will make the biggest difference, in my opinion.

I agree with Robert, I wouldn't assume the Naim clones will be better. The Cambridge uses LM3886s for the power amp section, which are pretty good. There are better circuits than the NAP140 out there.

Dan

Payday is here so Id like to tweak the amp a bit.

Im not planning to spend a lot or get into anything major, just remove most of the DC blocking caps (I dont have a meter so removing all of them would be risky) I want to leave at least 2 as Dan reccomends. A better opamp at U1, OPA2107 for warmth as this is a thin sounding amp.

From what I gather removing the feedback cap in the power amp is a no-brainer. The other caps im not so sure about (Bare in mind I have removed the output dc blocking caps in my CD player) which ones I should upgrade and which to remove?

I also have a pair of BG 220uf caps spare wich I'll be installing on the input selector chip, C4 C5. I'd also like to do the mod above to reduce the gain.

Basically, with a figure of about £50 in mind what would be the best way to spend it?
 
C40 C41 C42 C43 use a BG as large as can be fitted 35V+ (power chip local supply rail)

A freind reccomended the above. I have 4 x 50v 4700 Rubycon PK caps which would fit here. Would it be be worth it and what difference would it make?
 
See above for which caps to keep/bypass.

4700uf is too big for local decoupling. Stick with the same size or a little larger. In truth it's not a big deal here - capacitors "in the signal path" will have more effect than power supply caps.

Don't forget the Schottky diodes too - a small but worthwhile benefit.
 
I had a good look at the Circuit last night and I noticed that the tracks that carry the signal from the Bal pot to the Power stage are really long. Strangely the tracks for each channel are of very different lenghts.... It got me thinking about replacing the track where the audio signal would pass, with something like Silver, Copper, or some kind of plated variant wire. What I would do is solder insulated wire under the board bypassing the different sections of the track, (ensuring any crossed wires are at 45deg to each other) eventually id cut the track so its out of the picture. Has anyone ever done this and if so was it beneficial?
 
(I dont have a meter so removing all of them would be risky)

Mike - do go out and buy a meter. I haven't fully read this thread but it sounds like you've bought some of the NCC200 clones? If so, you really will need a meter for that... I wouldn't remove dc blocking caps without checking the dc that they're blocking. For instance, I had lm4562 opamps in my cdp output stage which gave next to zero dc on their output so I removed the dc blocking caps. Last night I thought I'd do some opamp rolling to look at what dc I get from different ones - lm6172 gave me over a volt, ad826 about 0.5 volts and opa2604 (the original amps I removed) gave exactly zero. Lesson being - measure...

From what I gather removing the feedback cap in the power amp is a no-brainer.

I don't think that can be done without a redesign of the power amp can it? (again havent read the whole thread so apologies if this has been covered.
 
I had a good look at the Circuit last night and I noticed that the tracks that carry the signal from the Bal pot to the Power stage are really long. Strangely the tracks for each channel are of very different lenghts.... It got me thinking about replacing the track where the audio signal would pass, with something like Silver, Copper, or some kind of plated variant wire. What I would do is solder insulated wire under the board bypassing the different sections of the track, (ensuring any crossed wires are at 45deg to each other) eventually id cut the track so its out of the picture. Has anyone ever done this and if so was it beneficial?

I did something like this in my pre-amp recently, and the results have been pretty mixed - I used lengths of shielded cable to replace a path that include some fairly long PCB traces, two IDC connectors and a section of ribbon cable, and in some respects things are worse than the best it could do before.

It might be better, or might not - the tracking should (I hope) have been placed fairly carefully.
 


advertisement


Back
Top