YAMAMOTO YDA01 Kit - My new DAC build

[ravi_d]

After a long gap I could do some progress on the dac build. Assembled all the 9 power supply boards (tested separately, using a normal UNITY multi-meter, 4 boards are showing a difference of .1 volt difference from the required volt and are fluctuating randomly (all others are showing exactly 3.3v and 5v continuously), Is it safe and what could be the reason) and assembled them to the main board using connector legs and assembled the main board parts as well. Also made progress in the SS o/p board and connected them using 6N pure copper wires. So the assembling is almost complete and fixed them on a temporary ply board (same size as the final enclosure).

Nice to know the progress on the DAC.

As per the PCM1794 spec sheet the voltage going for the Analog power supply (VCC1, VCC2L, VCC2R) can be in between 0.3 V to 6.5 V and the Digital power supply (VDD) 0.3 V to 4 V. Also its given the voltage difference the Analog power supplies (VCC1, VCC2L, VCC2R) should not be more than 0.1 V. If you maintain the difference with in 0.1 V that should be fine.

On the CS8416 the Analog Power Supply (VA) is important and to be fed with clean power.

> I have used 6N pure copper wires for connecting the dac main board and the o/p stage. Do we need to use the 6N copper for the power supply also for both main board (7 sets) and o/p stage (2 sets)? Is the purity of the copper in anyway critical for the power supply (from RCores to the PS pcb)? How good is the CAT5 cables compared to the 6N copper? I am planning to use 3 wires braided together for each board (+ - +). Please advice.

Specs:
- 2 RCores x [2 x 15v w/ .354A each] 10vA
- 2 RCores x [2 x 7v w/ .7A each) 10vA

>Is it recommended to fix all the 4 RCores on a single plane (to prevent the mutual interference if any)?

Make sure you use separate transformers for powering Digital and Anlalog circuits.
BTW what voltage the I/V stage (Output board) requires?
Where did you plan to use those two transformers with 2 x 15V output voltage each?

 

[cmsajith]

Nice to know the progress on the DAC.

As per the PCM1794 spec sheet the voltage going for the Analog power supply (VCC1, VCC2L, VCC2R) can be in between 0.3 V to 6.5 V and the Digital power supply (VDD) 0.3 V to 4 V. Also its given the voltage difference the Analog power supplies (VCC1, VCC2L, VCC2R) should not be more than 0.1 V. If you maintain the difference with in 0.1 V that should be fine.

On the CS8416 the Analog Power Supply (VA) is important and to be fed with clean power.



Make sure you use separate transformers for powering Digital and Anlalog circuits.
BTW what voltage the I/V stage (Output board) requires?
Where did you plan to use those two transformers with 2 x 15V output voltage each?

I am using 4 Rcores (2 x 7v (Main board) and 2 x 15v (Output Stage)).

I am not sure whether the fluctuation of .1v is actual as I am using normal multimeter. Anyways its always +/-.1v.

 

[reignofchaos]

It is probably multimeter error. From my engineering days, I do remember that even the fluke multimeters would oscillate 0.1V all the time. Cheapo ones will only be worse.

Only way to be 100% sure is to connect it to a scope.

 

[cmsajith]

So I will go ahead as all the PS boards are having same parts from the same lot.

The only problem with those parts is that some of them are out of production especially the SANYO OSCON caps and are rare to find and hence very expensive. So experimenting with the parts is almost impossible.

A novice question, is there any chance for this difference because of any soldering issues, I dont see any other reason apart from the multimeter fault. I use CARDAS Quad Eutectic Silver solder.

 

[reignofchaos]

So I will go ahead as all the PS boards are having same parts from the same lot.

The only problem with those parts is that some of them are out of production especially the SANYO OSCON caps and are rare to find and hence very expensive. So experimenting with the parts is almost impossible.

A novice question, is there any chance for this difference because of any soldering issues, I dont see any other reason apart from the multimeter fault. I use CARDAS Quad Eutectic Silver solder.

I highly doubt that. If continuity test between the soldered wire points beeps correctly, then soldering is not the issue at all.

 

[ravi_d]

I am using 4 Rcores (2 x 7v (Main board) and 2 x 15v (Output Stage)).

I am not sure whether the fluctuation of .1v is actual as I am using normal multimeter. Anyways its always +/-1.

The real test is those voltages shouldn't fluctuate much and should stay with in their limits with the load connected.

 

[kaushik]

costliest DIY dac here at HFV:
1.8Lakh DAC kit only ! ... i need to shift my job many times it seems

 

[omishra]

for both main board (7 sets) and o/p stage (2 sets)? Is the purity of the copper in anyway critical for the power supply (from RCores to the PS pcb)? How good is the CAT5 cables compared to the 6N copper? I am planning to use 3 wires braided together for each board (+ - +). Please advice.

Cat5 cable has good copper and 24SWG can carry up to 500mA Max current. Only caveat it is teflon coated jacket and while soldering it can melt and do short-circuit in grouped wires. Care is required at stripping and soldering stage only.

>Is it recommended to fix all the 4 RCores on a single plane (to prevent the mutual interference if any)?

Magnetic fields of r-core or toroids are internal and very very less leakage. Any type of mounting should not create any problem. But separating with perforated iron sheet can have added benefit of EMI protection. (ref my Pass B1 box )

I am not sure whether the fluctuation of .1v is actual as I am using normal multimeter. Anyways its always +/-.1v.

This is problems of DMM (most of the times). It never settles at small values. Don't worry about these reading fluctuation, assume average reading is present there. Let it give 5 sec settling time before reading.

 

[cmsajith]

Cat5 cable has good copper and 24SWG can carry up to 500mA Max current. Only caveat it is teflon coated jacket and while soldering it can melt and do short-circuit in grouped wires. Care is required at stripping and soldering stage only.

Another dummy question, The recommended power rating for the 7v trans for main board is 500mA but the RCore is having 700mA. Is there any issue in using the Cat5 in this case as there wont be any requirement than 500mA? Also If I braid 2 / 3 strands (along with insulation) of Cat5 cable and combine them at the 2 ends, will it be sufficient to carry 700mA?

The commercial version is using custom made torroid transformer and the parts used includes vaccum treated wood also to avoid any vibrations caused by the torroids. How about the RCores, planning to mount them on rubber feets.

Where can I find shielded hookup wires (for the i/p and o/p signals)?

Magnetic fields of r-core or toroids are internal and very very less leakage. Any type of mounting should not create any problem. But separating with perforated iron sheet can have added benefit of EMI protection. (ref my Pass B1 box )

Thanks, Already planned for either perforated / full sheet of aluminium separation or in a separate external box.

 

[gobble]

The real test is those voltages shouldn't fluctuate much and should stay with in their limits with the load connected.

I think if the multimeter probe does not make firm contact with the electrical point of measurement, the readings will fluctuate...

Shaky hands may be the reason. A reminder to quit smoking, alcohol and whatever else that needs shaking...

--G0bble

 

[omishra]

Another dummy question, The recommended power rating for the 7v trans for main board is 500mA but the RCore is having 700mA. Is there any issue in using the Cat5 in this case as there wont be any requirement than 500mA? Also If I braid 2 / 3 strands (along with insulation) of Cat5 cable and combine them at the 2 ends, will it be sufficient to carry 700mA?

I can give an example -
Consider this, you have glass filled with 700ml, your stomach has max capacity of 500ml, your food pipe(throat) can niggle 500ml at a time but your drink only 50ml at certain instance. glass will be kept filled always. So will that burden that food pipe? No.

Similarly, you DAC may not need more than 200-300mA constant current from one supply line(it has 4 IMO), wire is well capable of 500mA, also supply is 700mA full capacity. It will be replenished 50 times in a second. DAC will take what it needs, no problem.

The commercial version is using custom made torroid transformer and the parts used includes vaccum treated wood also to avoid any vibrations caused by the torroids. How about the RCores, planning to mount them on rubber feets.

These EMD r-core -> cores are very nice, tight and I find absolutely no vibration anywhere in transformer. They are wound over bobbin using computer controlled winding machine. Only part missing is label. While ordering did not give label choice to cut cost as quantity was measly. I think you can mount them directly w/o any issue.

Where can I find shielded hookup wires (for the i/p and o/p signals)?

You can look into thick USB slave cable or DVI cable of computer monitor which contained good quality 4 pairs of shielded stranded cable with flexible ground wire running along with aluminum foil. Also they are twisted for good results. Make sure to ground that foil/ground wire.

So far I did not cut HDMI cable for that purpose. My interest is upto using ferrite bids from it only.

Thanks, Already planned for either perforated / full sheet of aluminium separation or in a separate external box.

Better you stick to iron partition, IMO that will do better EMI shielding than Aluminum. Its ferromagnetic v/s paramagnetic material debate.

 

[sachu888]

The commercial version is using custom made torroid transformer and the parts used includes vaccum treated wood also to avoid any vibrations caused by the torroids. How about the RCores, planning to mount them on rubber feets.

Where can I find shielded hookup wires (for the i/p and o/p signals)?

I have mounted R_core directly on chassis for my B1.There is no vibration, humming or current leakage.
You can use USB cable for in/output leads,but its thick and can lift PCB traces.I have used it on my CNC before settling down to CAT5 and I have lifted few traces on PCB.CAT5 is a way to go IMO.

Regards,
Sachin

 

[cmsajith]

Planning to make hookup cables myself out of pure silver wire (single core 28awg) with teflon tube and a shield from the usb cable / RG6 cable. I had made ICs using them and are pretty good. I would have bought the shield from SP road but I have to buy 10m min (if nothing works out, this will be the last option)

 

[jls001]

Planning to make hookup cables myself out of pure silver wire (single core 28awg) with teflon tube and a shield from the usb cable / RG6 cable. I would have bought the shield from SP road but I have to buy 10m min (if nothing works out, this will be the last option)

May be you can braid the silver wires in Litz formation to get higher effective gauge (and current carrying capacity), and use teflon to shield. Where do you source silver wire? I have seen it on ebay.in but I am doubtful about the purity. Also, is the teflon soft and flexible, or hard and tough type. I have only found the hard and tough type. I was thinking of doing a DIY IC using silver cores and teflon shielding.

 

[cmsajith]

May be you can braid the silver wires in Litz formation to get higher effective gauge (and current carrying capacity), and use teflon to shield. Where do you source silver wire? I have seen it on ebay.in but I am doubtful about the purity. Also, is the teflon soft and flexible, or hard and tough type. I have only found the hard and tough type. I was thinking of doing a DIY IC using silver cores and teflon shielding.

I am sourcing it from
AWG28 0.3MM Solid Core UPOCC SILVER Audio BARE WIRE 20M | eBay

Seems to be a reliable seller and sourcing it from Wan Lung, Taiwan.
I have used oversized transparent teflon tubes and are very flexible. I have used it for preventing the +/- from shorting, the insulation is basically the AIR.

 

[sachu888]

Peter denials of Audiosector also recommends 26 awg solid copper/CAT5 wire for input/output.
"As I mentioned previously, I never found the need for shielded wire and what I'm normally using is 26ga gold plated solid core copper. I only put teflon sleeving where the wire might come in contact with other wires, or chassis parts.I got it from a friend and he didn't disclose his source (most likely in UK)

Other alternatives can ba CAT 5 wire or Cardas uninsulated silver wire and Vampire Cast Copper Wire (available from Percy).

Kimber 19ga TCSS is also good for signal as well as DH Labs Revelation Series 23ga Solid Core Silver hook-up wire available from Partsconnexion."
BTW a good read of building gainclone by Peter Denials here Commercial Gainclone kit- building instructions - diyAudio

Regards,
Sachin

 

[cmsajith]

Peter denials of Audiosector also recommends 26 awg solid copper/CAT5 wire for input/output.
"As I mentioned previously, I never found the need for shielded wire and what I'm normally using is 26ga gold plated solid core copper. I only put teflon sleeving where the wire might come in contact with other wires, or chassis parts.I got it from a friend and he didn't disclose his source (most likely in UK)

Other alternatives can ba CAT 5 wire or Cardas uninsulated silver wire and Vampire Cast Copper Wire (available from Percy).

Kimber 19ga TCSS is also good for signal as well as DH Labs Revelation Series 23ga Solid Core Silver hook-up wire available from Partsconnexion."
BTW a good read of building gainclone by Peter Denials here Commercial Gainclone kit- building instructions - diyAudio

Regards,
Sachin

Yes, he is correct. Unless we have some power/supply wire or transformers next to these wires, it is similar to an IC soldered to the pcb. Since I have good enough shielding from the trans and the power supply through another shield (similar to trans shield), I will try using bare silver wire or 6N copper wire with teflon tube if required.

Can we use the same silver wire / copper wire for digital coax i/p?

 

[gvenu]

Sajith,

Any further update on this DAC build?

Regards,
Venu