I’ve recently grounded the chassis and moved around some of the wiring and I’m getting less fatigue from the DAC. The i2s cable in particular.
Is there a point where this DAC should be grounded to the chassis? So far only my transformer is.
Would it be worth trying to shorten the i2s cable, or should I just be routing it as far away from anything else as possible? Would twisting or braiding the cable help?
Is there a point where this DAC should be grounded to the chassis? So far only my transformer is.
Would it be worth trying to shorten the i2s cable, or should I just be routing it as far away from anything else as possible? Would twisting or braiding the cable help?
@abraxalito Richard interested in purchasing your project kit. Would you PM me so I can send details. Thank you!
I shortened the i2s cable to about 6cm and I perceived better micro dynamics. I'm also pretty exhausted, so I'll need to check in the morning lol. I'm planning to change the wire leading to the RCA connectors with some shielded mogami w2534 I have spare, but I'm not sure how I should wire the shield. Could someone please offer some advice.
Right now I'm thinking R+ and L+ to the centre of the RCA connector as normal using the two of the 4 conductors. R- and L- would need to be shorted with the shield and the other two conductors from the dac end and that would continue to the outside terminals of the RCA connector. I would then short the RCA connectors to the chassis and earth ground.
Right now, the DAC as far as I can tell is not shorted to the earth ground at all. Would this still be correct?
Grounding is a bit of an art. I think try connecting the 0V at the power input to the DAC to chassis, and make sure chassis is earthed. Some recommend a 'ground lift resistor' between 0V and chassis but this gives a safety hazard when exposed metal (the RCA shield terminal) carries a lethal voltage which may result from a fault current overloading the resistor.
I'm not familiar with Mogami W2534, is it a twisted pair internally with 3 conductors in total?
I'm not familiar with Mogami W2534, is it a twisted pair internally with 3 conductors in total?
Thanks for the prompt response! The cable is 4 conductors internally plus the shield.
Chassis is earthed. I’ll take a look at the dac in the morning and take some photos to confirm your instruction. I don’t want to mess up now, thanks!
Connecting DAC 0V to chassis might result in hum if the next piece of kit in your setup (preamp, poweramp) also has 0V to mains earth. Poweramps normally are wired like that.
With 4 conductors inside the Mogami I'd try using two for each of L and R or leave two unconnected. Wiring conductors to 0V will likely only increase capacitance to GND which is undesirable. If you short RCAs to chassis then you'll have an internal 0V loop once the DAC 0V is also tied to chassis. You could though try using a small cap (say around 1nF) between RCA shield and chassis to encourage RF to bypass the DAC.
With 4 conductors inside the Mogami I'd try using two for each of L and R or leave two unconnected. Wiring conductors to 0V will likely only increase capacitance to GND which is undesirable. If you short RCAs to chassis then you'll have an internal 0V loop once the DAC 0V is also tied to chassis. You could though try using a small cap (say around 1nF) between RCA shield and chassis to encourage RF to bypass the DAC.
This is getting a bit complicated. Would you say there is any benefit to grounding the dac instead of just leaving it floating? Mostly I just want to figure out if the fatiguing nature of the dac ATM is just due to the rca cable crossing over some of the power cables.
Could I just run the shield to L- and R- respectively? I could run two separate cables, one for each channel. 2 conductors for L+ and two conductors + shield for L-. I would do the same for the right channel.
Unless of course L- and R- are the same ground then I would just use one cable and short the shield to both L- and R-.
Could I just run the shield to L- and R- respectively? I could run two separate cables, one for each channel. 2 conductors for L+ and two conductors + shield for L-. I would do the same for the right channel.
Unless of course L- and R- are the same ground then I would just use one cable and short the shield to both L- and R-.
Fatigue in my experience hasn't been caused by an RCA cable getting too close to a power cable. What does cause fatigue is having high frequency noise currents (from mains and especially from SMPSUs) running through the shields of unbalanced interconnects.
When the DAC isn't grounded the mains noise (coupled capacitively via the mains transformer) still wants to get back to mains earth so will do so via your interconnects when those terminate in a box where there's a link between 0V and mains earth.
From your pic it seems you have the right kind of mains trafo (EI) which has low capacitance between primary and secondary. You also are connecting the frame of the trafo to mains earth which helps by providing an alternative path for mains noise. So a couple of boxes ticked already.
As I have no experience with this kind of cable, I'd say experiment to see what works best in your context.
Thanks! I’ll test out the RCA cable and report back. Is there anything else I can do to mitigate mains noise outside of a power conditioner?
Here’s a pictures of the new layout, shortened i2s cable and upgraded RCA cabling where shield is going to the chassis, earth ground. L/R are using two conductors each.
The shortened i2s cable definitely improved micro dynamics. I think there’s an improvement of using coax over spdif now which wasn’t there before as well.
I think the shielded RCA cable improved fatigue, but I’m getting mixed results from the wife insisting it’s more fatiguing than before, but I’m out of ideas.
Would it be worth moving the components further away from the trafo? What about using shorter cable for the power cables?
Is there anything else I can do to optimize this dac?
The shortened i2s cable definitely improved micro dynamics. I think there’s an improvement of using coax over spdif now which wasn’t there before as well.
I think the shielded RCA cable improved fatigue, but I’m getting mixed results from the wife insisting it’s more fatiguing than before, but I’m out of ideas.
Would it be worth moving the components further away from the trafo? What about using shorter cable for the power cables?
Is there anything else I can do to optimize this dac?
Last edited:
To get better rejection of mains noise, you could try one or more of these latest design common-mode chokes. They go on the mains side, before the transformer. The highest inductance gives the best noise rejection.
https://www.aliexpress.com/item/1005004385150997.html?spm=a2g0o.productlist.main.3.2190pdMEpdME1O&algo_pvid=9d039f56-64ed-434b-904c-5109a543a0d4&algo_exp_id=9d039f56-64ed-434b-904c-5109a543a0d4-1&pdp_ext_f={"sku_id":"12000028993288656"}&pdp_npi=2@dis!USD!2.98!2.98!!!!!@2102188b16740928272703214d070d!12000028993288656!sea&curPageLogUid=CpCRFu87FqyE
The cable between the lower voltage (purple PCB) regulator and the I2S input board could definitely be shorter if that PCB were moved - is there enough space on the other side of the red PCB regulator? If there's the chance of a connection to chassis closer to the DAC 0V power input, I'd try that too rather than the longish wire you currently have.
You say 'shield is going to chassis' for the RCAs - have you tried without that connection? There will be a large loop with it.
https://www.aliexpress.com/item/1005004385150997.html?spm=a2g0o.productlist.main.3.2190pdMEpdME1O&algo_pvid=9d039f56-64ed-434b-904c-5109a543a0d4&algo_exp_id=9d039f56-64ed-434b-904c-5109a543a0d4-1&pdp_ext_f={"sku_id":"12000028993288656"}&pdp_npi=2@dis!USD!2.98!2.98!!!!!@2102188b16740928272703214d070d!12000028993288656!sea&curPageLogUid=CpCRFu87FqyE
The cable between the lower voltage (purple PCB) regulator and the I2S input board could definitely be shorter if that PCB were moved - is there enough space on the other side of the red PCB regulator? If there's the chance of a connection to chassis closer to the DAC 0V power input, I'd try that too rather than the longish wire you currently have.
You say 'shield is going to chassis' for the RCAs - have you tried without that connection? There will be a large loop with it.
The shield sheathes the wires and is not connected to L- or R-. Would there still be a loop present? 0V is not connected to chassis ground.
I’ll move the PSU of the i2s board closer to it and the dac if you think that will improve performance there is a lot of clearance, thanks!
Do you know which choke I should choose? Would it affect the tonality of the dac?
OK now I'm really confused. What's the role of the black wire terminating near the trafo mounting bolt? - at the trafo end it seems to go to chassis but the Mogami cable obscures where it ends up in the picture.
Shorter wires are normally better but seeing as I don't know yet the reason for the fatigue I'm making suggestions that might influence it. Making loops of cables smaller its good practice EMC-wise.
The choke with the most noise rejection is the best choice, that's the one with the highest inductance, in that link the highest inductance CMC is 28mH, 2A. As for affecting tonality, I don't know - getting rid of fatiguing SQ would be an effect on the DAC's tonality would it not?
A good point on the tonality. Thanks for picking out the model.
The Mogami cables has 4 conductors and an additional shield as pictured below.
I was reading on audio circle that you could ground only one end of the shield to provide some EMI protection. I’m hoping it will do so, but I’m very open to your recommendations.
Are you saying (it seems to me rather obliquely) that the black wire I mentioned in my last post is a connection between chassis and the Mogami shield? It would then follow that the shield is having the reverse effect it is designed for - instead of being a 'sink' for interference, its actually capacitively coupling interference into those 4 inner conductors because there's nothing tying the DAC 0V to the chassis. So the 0V is free to float around unless something in your amp/preamp ties them together.
My suggestion has been to link 0V to chassis with a wire at the 0V power input to the DAC board. Then connect the shield of the Mogami to a centre pin (or both centre pins if you like) of the analog output from the DAC. Connect the shield at the other end to the outer pins of the RCAs.
Its possible this will result in hum due to the grounding arrangements in your amp/pre but I feel we must try one step at a time.
Its possible this will result in hum due to the grounding arrangements in your amp/pre but I feel we must try one step at a time.