Electrolytic upgrade problems - Sounds worse :(

Bought a cheap Cambridge audio AM1 off ebay (£10) to have a play with. I already have an AM1. so have an unmolested one to compare with.

The amp uses the JRC4558D as a pre amp or buffer and the LM4766T chip for the power amp section.

Here's a list of things I did (made the classic mistake of doing it all in one go)

4 x 220uF 50v Nichicon KZ - Power Rail Decoupling
2 x 47uF 50v Nichicon KZ - Feedback?
2 x 100uF 50v Nichicon KZ - Power Rail Decoupling
2 x 22uF 50v Nichicon KZ - Power rail decoupling
1 x 22uF 50v - Just for LED light
2 x 10uF 50v Nichicon KZ - Power Rail Decoupling

6 x 4.7uF 250v JB JFX Polyproplylene Capacitors
2 x 6800uF 63v Panasonic TS-UP 30 Dia x40mm Aluminium Capacitors - Main Smoothers

4 x LQA06T300 6A 300 Q-Speed Diodes - Bridge Rectifier

R25/26 Vishay Dale RN60 33k LM4766T Feedback?
R21/22 Vishay Dale RN60 1k LM4766T Feedback? connected to 47uF to Ground
R19/20 Vishay Dale RN60 1k Signal Path
R1/2 Vishay Dale RN60 150k JRC4558D Opamp Feedback?

RG22Y RC Suppressor (100nF cap + 100R resistor) - Across transformer secondaries.

Emitor resistors and output resistors (5R and 10R) - Mills MRA05.


Pretty sure I got the feedback parts right but I could be wrong (wouldn't be the first time)


The original capacitors are 220uF and 100uF - JWCO -40 105c.
The 47uF (feedback), 4.7uF (coupling) and 10uF are KYS 105c.
Standard resistors look like these but I could be wrong - TE Connectivity LR1F


Anyway, I let all the new parts run in for a few days and sat back for a good listen. Bass was stlll good and there was a more shimmery treble with more detail but the lower mids were sucked out and sounded a lot worse than the standard amp (vocals were thin and lacked authority).

I put the old feedback resistors back in and the old coupling caps. This was much better but still not as good as the standard amp.
I then put the old emitor and output resistors back in and the sound lost a slightly soft touch and was a bit more alive sounding.

It's still not quite as good as the original amp and the rate I'm going, I can see me gradually putting it back to standard. At the moment it sounds a little bit bright in the upper midrange (only on some music) and I'm wondering if it's the Nichicon feedback caps that are causing it? (guess I'll find out when I put them back)

I get the impression that Cambridge Audio spent a lot of time voicing this little amp and improving it is going to be very difficult.
Anyone got any clever ideas?
 
Last edited by a moderator:
Brightness could be ringing from the supplies getting to the output.
and that brightness could make the mid sound recessed/sucked out.
6 x 4.7uF 250v JB JFX Polyproplylene Capacitors
What were these for?
I hope you didn't replace ceramic X7R supply rail decoupling with low esr polypropylene !

Neither 5r, nor 10r, sound right for output resistors, unless one of them is in parallel to a 1uH output inductor.
 
Last edited:

Mooly

Administrator
Paid Member
2007-09-15 8:14 am
This could be as much down to expectation as anything else.

You were all fired up expecting great things when in reality everything you have done has changed nothing. It happens. You are imagining or picking up negatives.

From your description, the single thing I would have done would be to ditch the 4558 in favour of something else.
 
I get the impression that Cambridge Audio spent a lot of time voicing this little amp and improving it is going to be very difficult.
Anyone got any clever ideas?


Your experience mirrors mine with a wide range of equipment. The conclusion is clear:

1. Stick to components you really like and know. Never take component recommendations without extensively testing by yourself.

2. Hard to believe, but sometimes manufacturers do all the hard work properly

3. Do listen after each change if possible. Allow sufficient break in time.


And thank you for posting this. It is far from common to read a sincere account on upgrading :)
 

rensli

Member
2014-07-14 10:00 pm
I suggest to look into feedback related capacitors, problem might be also with input or output coupling if there is any. At least these sections are the most sensitive to capacitor choice on Naim NAP250/140 and Apex AX10.

There is almost always very good results with input capacitors that are very low internal ESR (impedance).
As for feedback capacitor, Nichicon VZ or some equivalent will do better then ultra low ESR on the named amplifiers. ESR about 0.5OHM @ ~100khz

If decoupling is needed, then only with capacitors which have some internal resistance also. Nichicon VS or some unnamed brand will do OK.

For Filter Capacitor banks, i like to use huge capacitors paralleled with smaller one's, if there is room for it, for example 10000uF or 6800uF or 4700uF paralleled with 2200uF capacitors, not less.

I see you used 6800x3 per shoulder totaling 6 of them. I used recently 6x4700uF (cheap SAMWHA :D) for stereo on AX10 amplifier and i can't complain, this amplifier is metal-cased(finished).
Anyway, I let all the new parts run in for a few days and sat back for a good listen

Burn in has never happened to me, it is either good from the first turn on or bad until it is fixed, nothing fixes itself.
 
I know :) but it still doesn't stop the expectation factor kicking in swaying your thoughts. Comparing to the old is only valid done blind.
YaDadada. Take 8 years of aymphonic band orchestra or acoustc piano training to educate your ears. And wear earplugs around most pop&rock performances, guns, motors machinery or power tools, excessive volume damages your ears. Then test with accurate recordings of same acoustic sources. Very subtle variation could fool one , but what bad & good effects I've found from my mods have not been subtle. **** speakers could induce a lot of post amp distortions, use headphones. Top & bottom octave solo piano are particularly hard to get right with a reproduction. Some mods make no difference, I don't have to make up results to encourage myself. Do use a stock test source, I use Young & Warm & Wonderful,. Peter Nero on dynagrove.
OP, do a hf oscillation test on PS rails, speaker output, & op amp output. Fuzz on a scope or significant (>.2 VAC) voltage on a VOM with a a 390 pf ceramic coupler cap in series with the negative probe. You could have a bad solder joint someplace causing oscillation. I've found that 3 times already, my fault each time.
If not oscillation, try new tantalum for the input blocker caps. I've never used hi inductance caps for input. I did use 4.7 uf cpo 50v ceramic caps (on 2 v signal) with great effect, but nobody stocks those anymore.
Metal film resistors over 100k hiss a little less than carbon comp. I don't have an opinion about carbon film, I don't have any except in low fi radios.
4558 hisses a bit at 50x gain in a mag phono circuit, but at lower gains is not that bad.
 
Last edited:
Hey Fat Marley,
If you have a scope and a good signal gen, You can chart different signals through and see and hear the differences Different parts do sound different, try it and see how it changes maybe you get your golden sound!
Best to Ya!
NS

I don't have a separate scope and signal gen but I do have a DATS V2. I tested the ESR of some capacitors with my DATS. Not sure if anyone's interested but I'll upload the screen shots anyway.
 
In order to improve a circuit, first you have to understand it even better than the original designer. Simply 'upgrading' components is a good way to spend money, an unreliable way of stopping the item from working at all, and an almost useless way of genuinely improving it.

rensli said:
For Filter Capacitor banks, i like to use huge capacitors paralleled with smaller one's, if there is room for it, for example 10000uF or 6800uF or 4700uF paralleled with 2200uF capacitors, not less.
A complete waste of time and money. There may sometimes be merit in paralleling with another cap of the same value, but no point in paralleling with a somewhat smaller one. 10000uF in parallel with 2200uF is the same as 10000uF.
 
I suggest to look into feedback related capacitors, problem might be also with input or output coupling if there is any. At least these sections are the most sensitive to capacitor choice on Naim NAP250/140 and Apex AX10.

I spent years playing with the Naim preamp circuit and found that synergy between mods was the most important thing to get right. Put some nice revealing coupling caps in the circuit and it could end up bright and nasty. Put some low noise regulators in, better earthing and some circuit tweaks to lower noise and it could sound a bit dull. But do all those mods together and it would be great.