Thanks and that was what I had in mind when I had this idea.
BUT I need to know from someone who has actually done this before and I want to to hear from them the actual result, whether good or bad.
Hi,
You can clock two chips with the Flea, one on each output. If you use tightly twisted wire or thin coax, you will have no noise issues.
Just make sure you disable the PCB track that runs on the board from the DAC to the decoder, a lot of people forget that. Remove RD14 and a few wire-jumpers. If you leave the unused trace connected to the clock signal at one side, then you'll have a noise-antenna
If you are using a CD57/67, make sure to connect the clock to the other side of R104, which would be pin 21 (CRIN) of the SAA7372. Marantz used pin 22 (CROUT)
The schematic is correct, but look at the board traces...
Ray
You can clock two chips with the Flea, one on each output. If you use tightly twisted wire or thin coax, you will have no noise issues.
Just make sure you disable the PCB track that runs on the board from the DAC to the decoder, a lot of people forget that. Remove RD14 and a few wire-jumpers. If you leave the unused trace connected to the clock signal at one side, then you'll have a noise-antenna
If you are using a CD57/67, make sure to connect the clock to the other side of R104, which would be pin 21 (CRIN) of the SAA7372. Marantz used pin 22 (CROUT)
The schematic is correct, but look at the board traces...
Ray
Last edited:
Thanks Ray. Great !! I don't realize that your Flea got two outputs, one for DAC and the other for Decoder until you mentioned it and then I looked it up in the manual.
BTW are the two O/Ps just connected together without any buffered separation ?
Today I finished connecting the three Tx onto my CD63 and they work like a charm. I have posted up a few photos to show what I have done so far.
You can see my 4X22000uF smoothing capacitors and their leads are surrounded by copper foils which I have earthed to get rid of interference. The case is also heavily damped. The analogue rails are fitted with Korean built +15V and -15V low noise voltage regulators. Two Burson discrete opamps are fitted. You can also see that most of the capacitors are changed to Rubycon ZLG, Sanyo SEPC and WIMA caps. The sound quality is really nice and the bass is so tight and deep.
When Ray's two low jitter clocks arrive I will feed independent 16.xx and 8.xxMhz signals to the DAC and the Decoder as well as the servo. Afterwards I will build five low noise voltage regulators using LT-1763-5 to provide independent power supply to the servo, DAC and Decoder etc. Then I think the mod is considered complete by that time.
You can see my 4X22000uF smoothing capacitors and their leads are surrounded by copper foils which I have earthed to get rid of interference. The case is also heavily damped. The analogue rails are fitted with Korean built +15V and -15V low noise voltage regulators. Two Burson discrete opamps are fitted. You can also see that most of the capacitors are changed to Rubycon ZLG, Sanyo SEPC and WIMA caps. The sound quality is really nice and the bass is so tight and deep.
When Ray's two low jitter clocks arrive I will feed independent 16.xx and 8.xxMhz signals to the DAC and the Decoder as well as the servo. Afterwards I will build five low noise voltage regulators using LT-1763-5 to provide independent power supply to the servo, DAC and Decoder etc. Then I think the mod is considered complete by that time.
Attachments
Hey guys, just an update on the individual DAC analogue PSU. Definitely subtle gains for the cost and, perhaps more importantly, space. However, after a few days with it, I feel as though it sounds tonally the same but less distorted and more capable, similar to when I doubled up the transformers in my amp (a la Brent). Definitely a worthwhile mod but only after other PSU and clock mods, I'd say.
The outputs are not buffered, but connected to the output of the Tentlabs XO through a 33R resistor.
That sounds rather strange, as this cap provides power to the servo and digital section. I can't imagine that treble becomes totally rolled of by this one cap. What you describe has more to do with something being wrong in the analog section, did you change something there as well?
Regards,
Ray
Regards,
Ray
This is the only change I tried, no changes in analog section. Earlier with Nichicon 4700uf/35v it sounds good, but after change i feel like vocal is to my face. But improvement in bass. After I change I was listening to it for 7hrs. I felt tired. To fit the capacitor through the hole, i extended a thin lead to insert into the hole. Is it ok?
Thanks
Badri
I would be tired too, after 7 hours ;-)
Improved bass sounds logical with a larger cap, and the in-your-face vocals could indicate improved placement and soundstage. The fact that you don't like this (yet) may be caused by the player not being fully modified.
First, I would give it some more time, don't judge too quickly. If you still think the sound is not o.k. after a few days, or maybe a week, there's only one thing to do if nothing else of the electronics was changed: change it back to the Nichicon and see (listen) if the treble is back and decide if you like it better this way.
A bigger cap is not always better, you know. I'm not too fond of replacing a cap by an extremely large one (5x or 10x the value) personally, but the original is 4700uF, so 10.000 should be o.k.
Ray
Improved bass sounds logical with a larger cap, and the in-your-face vocals could indicate improved placement and soundstage. The fact that you don't like this (yet) may be caused by the player not being fully modified.
First, I would give it some more time, don't judge too quickly. If you still think the sound is not o.k. after a few days, or maybe a week, there's only one thing to do if nothing else of the electronics was changed: change it back to the Nichicon and see (listen) if the treble is back and decide if you like it better this way.
A bigger cap is not always better, you know. I'm not too fond of replacing a cap by an extremely large one (5x or 10x the value) personally, but the original is 4700uF, so 10.000 should be o.k.
Ray
Last edited:
I have a 22,000uF of the same cap there and I personally like it a lot. The AG are supposedly similar in construction to Black Gates, which are said to emphasize bass somewhat, and I think you may be correct that the very high treble may be rolled off a tiny bit, but the balance sounds more natural to me and more comparable with vinyl. My speakers are two way monitors and very revealing of harsh treble, so I personally like the effects of the larger caps. I seem to recall that bypassing them with small value X7R or film caps 'undid' some of the effect, which I perceived as reducing bass and re-introducing a harsh top end, but you might like it. However, as Ray says, wait for a while - at least a couple of days.
When I first changed the stock Tx by 3 Txs and changed the 4 smoothing caps to 22000uF on both the servo and the analogue rails I also had the same feeling as yours that the bass seemed to be too much. By that time I had not changed the clock and independent power supplies to the DAC,decoder and servo etc. I think it was because the bass became more tight and extended and therefore we just felt the high frequencies was not right - this is just relative because there was more LF going into your ear drums. By the time I changed the two clocks, the independent power supplies and using burson opamps the music became more balanced with clarity and excellent soundstaging. Therefore don't jump to a conclusion too soon until you have finished with all the mods.
Last edited:
Well, I tend to agree with Bigvoice. I have just changed the original on the board transformer to 3 standalone TXs. I have also changed the smoothing capacitors to 22000uF. On first hearing I also had a strange feeling that the bass was too much. It seemed the sound was not right. I believe my mind had got accustomed to the previous sound quality. When there was a big change in SQ my brain cell just could not cope with it. When I asked my wife who seldom listened to my CD63 for her opinion she said the music was alright and said there was nothing unusual apart from the very tight and extended bass which was not present before the modding.
I have not yet changed the the 2 clocks and independent power supplies to the DAC and decoder etc. I think when I have completed the mod my brain cell will need sometime to get use to the new SQ.
I have not yet changed the the 2 clocks and independent power supplies to the DAC and decoder etc. I think when I have completed the mod my brain cell will need sometime to get use to the new SQ.
I noticed that my two + and - 15V low noise regulators are very hot even though I put in a heat sink on each of them. I think they are at least 80 degree C !! I believe it is due to the high voltage drop from 30V to the regulated voltage of 15V and the excess power is dissipated as HEAT. Luckily there seem to be no problem on long time continuous operation, i.e more than 4-5 continuous running. I wonder if anyone has the same heat issue on their regulators for the analogue section ??
