Bitstream and bass

[Rotellian]

I have had 3 Rotel rcd-965bx cd players. I know resolution isnt great etc but the overall sound really is very nice (quite analogue in fact). Ive tweaked and improved all aspects of one before and never really truly improved the bass. What is it that i have been missing? It appears to lack drive and can appear to be a bit slow - is this merely a limitation of the dac (alos used in QED digit, micromega duo/leader? and few others - saa7323). Re-clocking doesnt improve this aspect, improved output stage helps. Would better de-coupling of the dac be a bette route to explore? Or perhaps its merely the filter network after the DAC.

[Dr.H]

Hi,

I tweaked a Marantz CD52 MKII some years ago and to my mind that machine still has the best bass I have ever heard in a CDP. Very tight, deep, resolved.
In fact doing an AB test with a Meridian 20 bit CDP, a buddy and I could not tell the difference between the two!

How did I get the 52 to sing? First, the bass of the 52 was very warm and thick. I discovered that by increasing the size of the cap post the 7805 regulator (which fed the SAA7320 and everything else that needed 5 volts), the bass warmth disappeared to be replaced by a tighter bass. In my case the cap went from 50uF to 200uF. Too much of an increase and the bass seemed to become too thin.
The 52 had a transformer with multipe secondaries. The BIG difference in bass and overall sound came when I desoldered the secondaries which ultimately fed the +-15V lines (for the opamps) and placed an ADDITIONAL transformer into the CDP, thus having a separate transformer (I think it was a 25VA) feeding the opamps.

Post your results here.
Ryan

[Fin]

Hi Rotellian

I have read some of your threads regarding your Rotel CDP and it seems like you've done some interesting things.

I have an old Philips CD624 which uses the same SAA7323 DAC as your Rotel. Today I posted two new threads (I'm new to this forum and to DIY) on my planned mods to this machine.

I'd really like to know what mods you have made in the past and the results. Also, if you get a chance to have a look at my plans, I'd like to hear your comments.

Regards

Fin

[Fin]

Hi again Rotellian

I came across these comments on bass for a Philips CD723. Thought they might be of use to you if you can addapt them to your Rotel. Can't vouch for them as I havn't tried them yet.

Replace the 47 microfarad output capacitors (2511,2512) with 10 microfarad low ESR or bipolar types and bypass these with 0.22 microfarad polyester/polypropylenes.
This speeds up the bass end of the player while still allowing it to reproduce very low frequencies. Bypassing further increases the speed.

Decibel Dungeon http://www.decdun.fsnet.co.uk/project.cd2.html#cdpsu

I also found this in someone else's discussion:

Yes, putting in a separate transformer for the analogue stage will work wonders. Thi is exactly what NIAM do on teh ir CD player upgrades with Hi-Cap powersupplies etc. But something else you can try as well is to add a seperate transformer for the +-10V lines. Just be careful, these lines are not regulated, so your transformer must have secondaries of as close as possible to 6.25V AC. In fact, measure the lines in teh CDP to check that it is +-10V.
The effect of this mod on your CD player will be great: much tigher bass, and lots of separation between instruments. Just be careful again though that you get the value of the secondaries on the new transformer right so that after rectification you end up with +-10V or whatever your CDP uses (should be +-10V).

A final, simpler and less risky mod (the +-10V mod is risky if you're not sure of what youre doing) is to increase the size of the capacitor after the +5V regulator. This is typically 200uF. increase this by about 100uF to get 300uF and listen to the effect on the bass. In my old Marantz, this tightened the bass considerably and could be used to tweak the quality of bass the CDP produced.

I don't trust the 6.25V AC figure!

Good luck

[Rotellian]

Cheers - ive tried different valeus and bypassing (i remember reading somewhere that that might not be a good idea) the output cap. The safest way is to replace with better caps of same value and high enough voltage - perhaps a large film cap. Most obvious are blackgate non-polarised (bg-n bg-nx) My output caps are std polarised blackgates fitted as std. These seem ok to me. I may try replacing them with superior sometime. Im currently investigating my options for a project - maybe another old bitstreamer or a newer cheapo dvd player perhaps. Yes thats Ryans comments on the cd-52. I had nice results by simply putting in a seperate reg on a small board and feeding (cutting and feeding in replacment) the traces that lead off from the main 5v reg for all the servo control etc etc. What i want to do in my next project is put in seperate transformers - i need to read more about how best to do this - i believe you have asked elsewhere about the 0v connection. This is where my knowledge reaches an end - it will work if you connect it to a variety of places - but not all will necessarily be good.

[Fin]

Guido has just answered our 0V or ground question in my other thread "CDP Mods".

One groundplane. All grounds of components, and decoupling capacitorsto the plane like in my article. External supplies, especially the ground, should be connected to the point at the PCB where the circuit is that will be fed by that new supply

All parts on the PCB are connected to the ground (layer), and the ground of all external power supplies should be connected to that ground, closest to the circuits they feed

Merry Christmas!

[Hobbit13]

You'll get a lot faster and more detailed bass by simply replacing the opamps. (BB opa 134 for low-budget ($2), opa 627 ($20) (never heard it, but should be better))

[JohnW]

I no longer have the Data sheet for the SAA7323, but I remember that with the SAA7323 & SAA7350 there are Vref pins that supply the Reference supply for the internal Switch Capacitor network.
The Reference voltage (2.5V) at this pin is used to charge-up the internal Switching Cap, any modulation on this pin will result a modulation of the charge on the cap. The modulation effect is more pronounced at LF (bass) where the effect of the external capacitor decoupling is limited. Increasing the size if the decoupling Cap helps (and Cap Quality - try OS-CON's here), but will increase the “Turn on time”. If you add higher value caps, it’s very important to leave the smaller value (normally 100nF) caps for HF decoupling.

On my SAA7350 designs I used an external low noise 2.5V Reference generator – based upon a 5534 to reduce the impedance at this critical node. Be very careful with this pin, any external noise directly effects the performance of the DAC.

[Fin]

Originally posted by JohnW
I no longer have the Data sheet for the SAA7323, but I remember that with the SAA7323 & SAA7350 there are Vref pins that supply the Reference supply for the internal Switch Capacitor network.
The Reference voltage (2.5V) at this pin is used to charge-up the internal Switching Cap, any modulation on this pin will result a modulation of the charge on the cap. The modulation effect is more pronounced at LF (bass) where the effect of the external capacitor decoupling is limited. Increasing the size if the decoupling Cap helps (and Cap Quality - try OS-CON's here), but will increase the “Turn on time”. If you add higher value caps, it’s very important to leave the smaller value (normally 100nF) caps for HF decoupling.

What size cap is suggested here?
The service manual shows 6.8uF and the 7323 datasheet shows 47uF.

[JohnW]

Fin,

It’s important that the Cap on the Vref decouple pin has low ESR at the internal operating frequency of the switch capacitor network (and its harmonics).

(192 x fs) about 8.46MHz for 44.1 Red Book CD.

Try a large Oscon, say 100uF to 470uF with a working voltage greater then 3V

The pin must also be decoupled with a ceramic cap - say 100nF.