[Nerull]

Hey!

I'm modding a CD player, and need some help with a few things.
The schematic is here:
http://www.filefactory.com/?211295

Based on the schematic...

The opamp will be changed to two AD843s with an adapter (socketed of course).
The muting transistors have been removed (no audible distortion occured) and it sounded better (just minor though).

I'll be replacing the clock with a TentLabs clock (XO-2 maybe?)

The output capacitor needs to be removed and replaced as well. In there is a 22uF electrolytic. I want to replace it with a better quality film cap (ICW Clarity Cap from diycable) but I don't know what value. (Would anyone recommend a better coupling capacitor?)
Using the formula

1/(2*pi*Capacitance*Impedance)=Fc

I don't know if I need the input impedance of my amplifier (an M^3) included with the 1k resistor for impedance calculations. I would want the frequency of that equation to be near 2Hz in order to not interfere with above frequencies and not allow DC to pass (or anything else) right?
If I calculate the equation with 22uF and 1k resistance like in the schematic, it comes to roughly 7Hz or so, so it has to take into fact the input impedance of something connected to it, right?

Next, I need to do something about the power supply.
The actual regulator is a PQ05RR12 by Sharp with a built in Reset pin for the master control part of the CD player (to switch everything on like a voltage supervisor). How would I go about replacing that part for better noise control while keeping the reset line? Would I use just a regulator with a voltage supervisor by TI or something to pull up or down the master controller? Or could I possibly hard wire in a voltage regulator tied to the pins of the PQ05RR12 and keep that in so as to keep the reset, while disconnect the pin for power out from the PQ05RR12 and solder a regulator onto the pins of the Sharp regulator itself (not onto the power out of course). It doesn't help that I can't find out whether the PQ05RR12 or PD4900B RESET pin is pulled up or down for reset

.

Any advice on replacing the capacitors and diodes in the power supply? They look awfully cheap, and probably could use more capacitance or higher speed diodes.

The decoupling capacitors for the power supplies may be replaced by the following:
Panasonic 120uF/25V/FC // 100nF/X7R/0612 // 220nF/X7R/0805 // Murata 4.7uF/16V/Y5V/1206
(pulled directly from http://www.diyaudio.com/forums/showt...=&pagenumber=4)
Would this be worth it? I think it'd be important, because it'd filter out the nasty noise and spikes that are probably in that power to begin with.

Lastly, can anyone else see any other mods that can be done further for better performance (besides adding an internal I2S DAC, which I'll be doing a few months from now when it's done)?

Thanks for your time!

~Tom

 

[Garbz]

Get the XO-3. The reclocking circuit will be very useful if you ever get an external DAC.

The high pass filter with the electrolytic cap is with the 22k resistor following it. This gives fc=1/(2pi*22k*22u) < 1Hz. The 1k resistor forms a low pass filter with the 1500p cap at fc=100kHz I think. I would try removing these.

I'd keep the regulator and put some filtering on the output. A CLC filter or something can be rigged up (difficult), or some simple ferrite beads. The B+ supply however looks unregulated so it may benefit from some extra capacitance making the cap swap worthwhile. Not for something better but rather larger.

Do you think diode noise will make it through a large capacitor and the PSRR of the opamps? I'd leave thoes alone.

 

[Nerull]

Quote:

Originally Posted by Garbz Get the XO-3. The reclocking circuit will be very useful if you ever get an external DAC. The high pass filter with the electrolytic cap is with the 22k resistor following it. This gives fc=1/(2pi*22k*22u) < 1Hz. The 1k resistor forms a low pass filter with the 1500p cap at fc=100kHz I think. I would try removing these. I'd keep the regulator and put some filtering on the output. A CLC filter or something can be rigged up (difficult), or some simple ferrite beads. The B+ supply however looks unregulated so it may benefit from some extra capacitance making the cap swap worthwhile. Not for something better but rather larger. Do you think diode noise will make it through a large capacitor and the PSRR of the opamps? I'd leave thoes alone.

So I take out those 1k resistors and 1500pF (jumper the resistor gap). I buy two ICW Clarity Cap (18uF or 27uF, I'm thinking 27uF for better filtering).

Could I replace the 0.01uF capacitors in the power supply section with WIMA MKPs? The electrolytic filter capacitors also need to be replaced, and should all of them be replaced by higher value capacitors? Should I just put the largest capacitors (parallel for better filtering) physically possible? I say this just because the capacitance in the power supply looks meager. Maybe I should make an array of capacitors for high capacitance? (4700uF in parallel for each replaced capacitor)

Would that work out well?

~Tom

 

[Garbz]

Yeah that's a good start.

I wouldn't replace the 0.01uf caps though. There's no real difference between expensive wimas and cheap film caps or even ceramic caps for this application. The important thing here is cap speed and low ESR, which pretty much all caps of that value have already.

I say raise the capacitance but really what the capacitance is doing is simply providing some power filtering. 3300uf is more then enough for this type of application to provide current when demanded (remember the load on the cdplayer is much lower then even trying to drive a headphone). 220uf caps are sufficient for typical headphones, so 3300uf is plenty. The only real technical reason to swap it out would be if you heard 60hz or 120hz hum on the cdplayer's output. Other then that putting a single 4700uf cap would already be overkill

Also do remember that if you put TOO much capacitance after a transformer it can put undue strain on it, also inrush current increases so you could potentially blow a fuse on powerup.

 

[Nerull]

One more thing. If I raised the values of the resistors for the 1/(2*pi*C*R) circuit in order to put in lower value capacitors (cheaper, much cheaper....), would it adversely affect the circuit? I don't see why it would, but would it?

~Tom

 

[Garbz]

The 22k loads the output capacitor. The combination of that capacitor and resistor isn't actually 22k but, ((22k+1k) || load impedance). So assuming you change that to 1Mohm to allow you to plug a 0.1u cap in, if you plug the cdplayer into an amp with a 10k input impedance you end up with a 9.8k (guess) resistor in there and you loose all your bass and midrange too.

So the short answer is no, the resistor isn't the only thing that effects that part of the circuit. What I'd do firstly is check if there is any DC on the output for the cap to block. If not remove the cap and resistor combination completely and make sure your amp has input caps just incase the cdplayer fails in any way. If this wasn't an option I would still do what you sugested as I know that my preamp has an insane input impedance and my cdplayer won't get used with anything else anyway.

 

[Nerull]

I need the capacitor, as there is 90mV offset on each channel. I probably won't use my cd player with anything else, so it's alright.

Since my amp (M^3) has an impedance around 47.6Kohms (according to amb's website) the equation will go like this:

I want a Fc of about 1.5 Hz (x10 and it's 15 Hz, which is so low a frequency I'll be hard pressed to hear it, if at all anyway). Would there be a better frequency for this though? Maybe 2 Hz is best...

The resistance is 1/23k + 1/47.6K and inverting that after adding I get 15507 ohms. Putting that in the equation and solving for capacitance I get...
1/(2*pi*15507*1.5hz) = 7uF

Reasonable enough, but I still want lower. In order to achieve this, I would increase the resistor value (but this in turn increases resistor noise; using voltage spectral density's equation, that is, V/root Hz = (4kTR)^(1/2)). How high do you believe I should go on the Fc, as 10x that amount will give me the top frequency attenuated. Is there a certain point where it won't matter anyway?

What I'm trying to say is, what is the maximum resistor value I should put in the filter circuit without it having too much noise (which is what I'm worried about)?

Thanks!

~Tom

 

[Garbz]

I don't think that spectral density noise or shot noise is much to worry about, but ideally you do want to keep this low.

As it is you can drop in a 4.7uf cap which you can get in the film variety instead of electrolytic which gives you a corner frequency low enough that at 20hz the volume is 0.05db down (you won't notice), and the phase is also low enough (<5degrees between 20 and 100hz).