[mekanoplastik]

hi all

this is my home cd player amp output section, there are 3 "cheap" dual opamps (philips) and i am thinking about replacing them with something more exotic....any suggestions?

i will probably end up replacing them with 8620's or 627's as they seem to be the standard but i also would like to know if it would be possible to combine them 2-1 to to get differnet sounds

or if there are is an opamp known to behave specially well in cd players.....825's 797's ???

(any other idea for possible modifications are welcome...specially about the stuff at the output)



m.

 

[puppyslugg]

The opa627/37's are single channel opamp. You'll need an adaptor, like BrownDog's.

Don't forget the ad843.

 

[jeffreyj]

Quote:

Originally posted by mekanoplastik hi all this is my home cd player amp output section, there are 3 "cheap" dual opamps (philips) and i am thinking about replacing them with something more exotic....any suggestions? i will probably end up replacing them with 8620's or 627's as they seem to be the standard but i also would like to know if it would be possible to combine them 2-1 to to get differnet sounds or if there are is an opamp known to behave specially well in cd players.....825's 797's ??? (any other idea for possible modifications are welcome...specially about the stuff at the output) m.

Tweaks really are the most fun you can have, I think, because it's so easy to try different things out and see whether it helps, hurts, or, as is probably the case more often than people generally want to admit, does nothing

I am going to disagree with puppyslugg on this one because I think putting an instrumentation amplifier into an output stage that is coupled capacitively with an electrolytic is a big waste of money! The fact that the line output stage is powered by +/- supplies begs the question: is the coupling capacitor even necessary?? Try measuring the DC offsets present at pins 7 and 1 of U408. If either is less than 10mV, I'd think, instead, about ditching C435/6 and replacing the NE5532 with the pin compatible Burr Brown OPA2134PA (at a couple bucks apiece, it won't be so painful if you can't hear any difference!) Check the offset voltage again, and if you feel you must include a DC blocking capacitor, try putting in a polyester film one of approx. 1-4.7uF (or, even better, a polypropylene, but it will be difficult to cram one of sufficient capacitance into the space available, I imagine). The current low frequency cutoff of the output, assuming a nominal 10k impedance (a low impedance for an amplifier input stage, but it's better to be safe than sorry, eh?) is 0.72Hz. Going down in capacitance to 1uF brings that up to ~16Hz. This is, IMO, a tad too high a cutoff, so maybe 2.2uF, then. At any rate, 2.2uF of polyester film is far superior to 22uF of aluminum electrolytic for this application in the aspects of equivalent series resistance and inductance, dissipation factor and stability with temperature/time and voltage. Also, using a polarized capacitor in such an amplifier, as there currently is, is a naughty-no-no as the output is, of course, alternating, and if there is a DC offset present, and it is opposite the polarity of the capacitor's orientation, then the oxide film will be broken down and, eventually, if won't behave much like a capacitor anymore!

Good luck, and let us know how it goes: I love tweak reports!

 

[toiemoi]

Quote:

Tweaks really are the most fun you can have, I think, because it's so easy to try different things out and see whether it helps, hurts, or, as is probably the case more often than people generally want to admit, does nothing

For the first part of the sentence, I completely agree, tweaks are great fun, personnaly rewarding but you have to be cautious, especially when tweaking a low end, full SMD player. I would recommend using DIP8 support to swap for tests. The PCB board will not like multiple soldering/desoldering.

I personnaly have had good results in modifying my player, changing output OP-amp and the capacitors. Non polarizized cap are the way to go, if using caps (the best thing is just a wire !). I personnaly have BG, but I have some SCRs as well, just waiting in case. The best would actually to compare various caps, and combinations, but that is very time consumming.

Toiemoi

 

[jeffreyj]

Might be a little outside the scope of "head-fi", but checking into a lower jitter clock source, better D/A and/or better anti-aliasing filter (esp. if it is analog) might prove to be excellent ways to improve the sound with tweaks.

To toiemoi: yep, fooling around with SMD components is generally not much fun, especially since most manufacturing processes involve gluing them down to the board prior to being reflow soldered! Sure, you can get the solder off with braid, but what do you do about the glue??? "Oops, I just cracked another capacitor in half... now what was that thing's value, again???"

 

[aos]

Actually DC blocking cap is almost always necessary on the output of CD players. Why? Because most DAC chips are powered with +5V and have their output centered around 2.5V. So there'll be 2.5V DC offset at the output of the opamp too. The only exception that comes to mind are chips with current output such as PCM1738 or old PCM-63.

 

[toiemoi]

Quote:

you can get the solder off with braid

Yes... to all beginners, avoid soldering pump that put too much stress on the PCB... or you might regret it. Since I have been using braid, I never messed up a single pcb. You still have to be quick if you do not want to burn ICs.

Quote:

but what do you do about the glue

My technique is plain and simple: DESTRUCTIVE. I cut all the leads of the component I want to remove, get rid of it and unsolder the end legs, clean up everything before putting something else.

Quote:

checking into a lower jitter clock source, better D/A and/or better anti-aliasing filter (esp. if it is analog) might prove to be excellent ways to improve the sound with tweaks

The clock is a hairy problem. I am still reluctant to do it on my player (a DVD player). The DAC, yes, why not but the choice of chips is not that great either. In addition, a more expensive DAC (it actually applies to most electronic components) does not mean a better sound. So you need to experiment... and since I refuse to solder/unsolder smd's more than 3 times, that limits what I can test.

Analog anti-aliasing filter ? Isn't that usually just a cap (low pass filter)?

Other good tweaks can be done on the power supply, better faster caps and diodes and separate power supplies, especially for the mechanics one (good source of pollution. That needs to be isolated).

Toiemoi

 

[aos]

Quote:

My technique is plain and simple: DESTRUCTIVE. I cut all the leads of the component I want to remove, get rid of it and unsolder the end legs, clean up everything before putting something else.

I just use a screwdriver to remove all visible amount of glue (it was red, almost like wax, in my Sony 775). Then I remove solder with the braid. Then I put a screwdriver as a lever to pry the chip apart from the board while at the same time going over pins on both sides with soldering iron like mad. Worked fine for me though it is easy to damage the PCB if it peels of traces together with the chip!! It did happen to one pin in this case, luckily only partially. If you don't want to keep the old chip, you can cut them and then it's easy to remove them. So your technique is a safer bet on commercial equipment, though you do have to forsake the chip you're removing.

Quote:

Analog anti-aliasing filter ? Isn't that usually just a cap (low pass filter)?

No, and the by the way anti-alias filter is the wrong term to use, I think the better one is "reconstruction filter", or just low pass filter. Anti-alias is used when doing A-D conversion, in order to prevent, of course, aliasing. Anyhow, this low pass filter in DACs, it's usually a 2 pole filter done with one opamp and few caps and resistors. The same opamp you're usually replacing. There can be more opamps involved if the DAC chip has differential output (typically higher end chips).

 

[toiemoi]

Quote:

though you do have to forsake the chip you're removing

I usually tweak not too expensive products... so the parts I am removing may be worth $1 at the most (what is the price of a JRC4560 these days?). Not to mention that heating an IC is not recommended.

Yes, reconstructing filter is the appropriate terminology, I was silly of me to use the same terms as Jeffreyj. (anti-aliasing" are low pass-filters in front of ADC to prevent aliasing during conversion).

Are you sure about the order of the reconstructing filter? My previous two players only had 1st order ( a simple cap)... and I remember reading that some players don't use reconstructing filters anymore (especially for medias with higher sampling).

Toiemoi

 

[jeffreyj]

Quote:

Originally posted by toiemoi The clock is a hairy problem. I am still reluctant to do it on my player (a DVD player). The DAC, yes, why not but the choice of chips is not that great either. In addition, a more expensive DAC (it actually applies to most electronic components) does not mean a better sound... Analog anti-aliasing filter ? Isn't that usually just a cap (low pass filter)? Other good tweaks can be done on the power supply, better faster caps and diodes and separate power supplies, especially for the mechanics one (good source of pollution. That needs to be isolated). Toiemoi

Good points, all. It is true that the DACs in most cdp's are dedicated units for the job, so it is not so easy to find something that just plops right in, but there are many different ways of doing digital to analog conversion and with all the variety available, there is surely plenty of temptation to tweak.

I made a mistake on the anti-aliasing thing... those are only to be found on the front end of the ADC, not the DAC. What I meant to refer to is the low-pass filter generally found after the DAC/before the output amp. In some cases this may very well be a simple RC circuit, but an active Bessel filter is best for maintaining excellent phase and step response throughout the passband and well into the stop band.

 

[jeffreyj]

Quote:

Originally posted by aos No, and the by the way anti-alias filter is the wrong term to use, I think the better one is "reconstruction filter", or just low pass filter. Anti-alias is used when doing A-D conversion, in order to prevent, of course, aliasing. Anyhow, this low pass filter in DACs, it's usually a 2 pole filter done with one opamp and few caps and resistors. The same opamp you're usually replacing. There can be more opamps involved if the DAC chip has differential output (typically higher end chips).

Terminology error: mea culpa. Nobody's stupid for using the wrong terms, either, toiemoi, because we have all done so at one point or another. I suppose no one makes any typos around here, either, eh?

That 2 pole filter is a great place to do some tweaks, though. Depending on the topology, it is often simple to change from a Chebyshev or Butterworth to a Bessel, with superior time-domain properties at the expense of fc sharpness (which is hardly a consideration in this application). Matching the R's and C's and/or replacing them with superior types is always worthwhile in filter circuits.

jeffreyj

 

[aos]

All datasheets, and all players I've seen have the analog stage doubling as filter. One active element usually allows 2nd order filters. Filters are not as abrupt as they used to be - I've seen them with 50kHz and 75kHz corner frequencies when used for high resolution DACs (and today they all are), so they shouldn't make any difference in the 20kHz range. It's true you could get away with little or no filtering - I tried that with success. Some versions of my portable DAC had nothing more than a RC filter, and I've also removed it from "SDS labs" DAC I built earlier. I've also seen DACs with only passive RLC filtering.