I think without any DC blockers there could be minor inbalances in the opamps and thus a few mV DC present in the output of the CD63. For me the RCA has about 20-30mV DC bearing in mind that I haven't used any DC blockers. If a blocker is to be used as you said I will use at least 4.7uF audiophile quality caps.
Ah! Millivolts!!!!! That's ok then. Unless you have an expensive meter it will be reading out anyway. You could try swapping the opamps round to see if you can lower the 33mV reading. Your amp will have DC blocking on the input anyway (most likely).
I'm sure I started with 2.2uF and went up to 4.7uF when I changed to supremes. I think your amp will play a role in what's going on. If cost and size was not a factor, I'd go bigger!!
My PSU didn't work, as I kinda expected. It's measuring the correct DC against it's own ground but has no potential against the player's. I feel like I have to bring the grounds together somehow. My first thought was to wire from the negative terminal of the smoothing cap to the regulator ground pin, which would connect it to the PCB ground, but I don't want to break anything through experimentation. Any tips from someone who's done this already? Ian?....
That is the correct way. Don't worry the way you connect the grounds together will not fry your kits .
Yep, that worked like a charm - thanks higlander. I now have my left and right channel DAC analogue pins running off a separate supply each. Wow, the charge on these caps lasts forever after powering down. It's too early to say how it's affected the sound. I listened to a couple of albums from cold and it started out not seeming much different but then seemed to sound more clear and capable as time went on. I'm excited to do the others now!
My PSU didn't work, as I kinda expected. It's measuring the correct DC against it's own ground but has no potential against the player's. I feel like I have to bring the grounds together somehow. My first thought was to wire from the negative terminal of the smoothing cap to the regulator ground pin, which would connect it to the PCB ground, but I don't want to break anything through experimentation. Any tips from someone who's done this already? Ian?....
Come on Ben!!! You've got a bit more experience with all this to be asking that!!! Lol!!!
The DC grounds need to be tied together always!!!!! As in example, If you were to tie the pos of the psu to player gnd, you'd create a negative rail (assuming the ac side of the bridge is floating).
Hi Ben. Keep on with the good work and keep us posted !! Most of my components have come but I am moving a bit slow as I don't have the time to carry out the mod. I can only carry out the work bit by bitHa, not that much experience really. I'm still only on my first player, miraculously (touch wood!). I guess I did pretty much know what to do but was being coy 'cos I feel like I'm well overdue breaking it again - I could almost smell the smoke
That's the best way to do it though, I think. Take your time and enjoy the journey.
It looks as though you're really taking the time to do it properly. It's not the mess of wires and electrical tape that mine is right now
What regs are you using, if you don't mind me asking?
Good question !! Initially I thought I would use Brent's sPower unfortunately the price drove me off. Then I resort to Tentlabs shunts but I found that the current output for a 15V shunt is insufficient to drive the Burson opamps that I have just ordered. I use Burson because they occupy less space than Ray's DOS. I recalled that sometime ago a member has mentioned in this forum about low noise regulators from Korea like this ULN-HC79A Ultra Low Noise Negative Regulator -5V~-30V Output High Current Type | eBay
I have just bought a pair (+ and -15V) and they are on their way to me. I think the cost is reasonable and the specification looks GOOD. I have no idea actually how good they are but from the photo the workmanship seem to be alright, with fingers crossed
One more thing, presently I am in contact with Ray to order 2 low jitter clocks with Xtal (16.xx and 8.xx Mhz) and one power supply. Ray has quoted me the cost and I will place my order when he gives me his account details.
Last edited:
As to the other regulators I may consider trying this:Good question !! Initially I thought I would use Brent's sPower unfortunately the price drove me off. Then I resort to Tentlabs shunts but I found that the current output for a 15V shunt is insufficient to drive the Burson opamps that I have just ordered. I use Burson because they occupy less space than Ray's DOS. I recalled that sometime ago a member has mentioned in this forum about low noise regulators from Korea like this ULN-HC79A Ultra Low Noise Negative Regulator -5V~-30V Output High Current Type | eBay
I have just bought a pair (+ and -15V) and they are on their way to me. I think the cost is reasonable and the specification looks GOOD. I have no idea actually how good they are but from the photo the workmanship seem to be alright, with fingers crossed
One more thing, presently I am in contact with Ray to order 2 low jitter clocks with Xtal (16.xx and 8.xx Mhz) and one power supply. Ray has quoted me the cost and I will place my order when he gives me his account details.
http://www.diyaudio.com/forums/digital-source/54009-marantz-cd63-cd67-mods-list-1846.html
BigVoice used LT1763-5 which occupied very little space. I have checked the data sheet and the noise level is quite low.
I'd be interested to find out what semiconductors are being used on those regs. I suspect that one of the chips is a voltage ref and the other is the Opamp used error correction and controlling the output device. A popular choice for the pos reg is D44h11 as used in lots of discrete regs. Looks like the dropout required is in the region of 3.4v which sounds about right for that output device.
You could probably further improve performance of the reg by swapping out components!!!!!
You could probably further improve performance of the reg by swapping out components!!!!!
Ok when the regulators arrive I will find out what the chips are on the regulators and let you know.I'd be interested to find out what semiconductors are being used on those regs. I suspect that one of the chips is a voltage ref and the other is the Opamp used error correction and controlling the output device. A popular choice for the pos reg is D44h11 as used in lots of discrete regs. Looks like the dropout required is in the region of 3.4v which sounds about right for that output device.
You could probably further improve performance of the reg by swapping out components!!!!!
Re. the idea of swapping out components from these regulators, ah I don't think my eyes and hands are good enough to carry out such delicate work
The two DIY regs that I've been considering to upgrade my 'off-piste' LM317s are ALWSR and SuperTeddyReg. Both are available either as bare PCBs or pre-built and tested. They're both a little on the larger side.
The other pre-built commercial regs that have really caught my eye are Paul Hynes series and shunts, which are small TCO replacements like SPower. They seem well regarded and are very competatively priced.
The one thing that I haven't seen anywhere is any kind of objective 3rd-party comparison. We need a 'Which?' report for low noise regulators.
Edit: I also previously considered getting Fleas in a big way, because Ray offers complete kits. I've already built two. I've blown that option by getting 9V transfromers though. Their drop-out is 13V!
The other pre-built commercial regs that have really caught my eye are Paul Hynes series and shunts, which are small TCO replacements like SPower. They seem well regarded and are very competatively priced.
The one thing that I haven't seen anywhere is any kind of objective 3rd-party comparison. We need a 'Which?' report for low noise regulators.
Edit: I also previously considered getting Fleas in a big way, because Ray offers complete kits. I've already built two. I've blown that option by getting 9V transfromers though. Their drop-out is 13V!
Last edited:
@UV101: I know you have experience with clocking the decoder directly. I am sure the SQ of CD63 will be different if the decoder is clocked independently. My question is can I just split the 16.xxMhz clock output from the external low jitter clock and connect one cable to the DAC's pin 28 (XT1) and connect a wire from the same signal signal source to pin 13 of the decoder. (I know for MK I I need to isolate R14 and feed the signal directly to pin 13 of the decoder, while for MKII I need to cut the leg of c521 near the DAC side and connect a clock signal onto the floating leg of this cap.) The reason I ask you is because I will use Ray's flea kit and it has only one 16.xx Mhz output. Unfortunately in the flea there is no second 16.xxMhz buffered output similar to that of Brent's C2 clock. That is why I will use the same signal source to connect both the DAC and decoder but I am not sure if it will work or not. Hope you have the answer for me as presently I do not have Ray's flea to test it out.
Last edited: