Power supplies and cables - observations, considerations & commentary

[genjamon]

Ok, not sure if this is the place, but I need some help troubleshooting. A new HDPlex LPS arrived to my home yesterday, but I'm having issues. One of its four outputs is variable, and I immediately used my multimeter to adjust the output to around 8V - it arrived at 9V. The output is variable from 5-19V.

 

I then plugged it into the microRendu. It powered up fine, and started playing fine. However, I'm getting quite a bit of noise - it's like I'm hearing the computer working - sounds like bursts of data noise, or something like that. This noise is present whether or not music is playing, and is easily audible from listening position.

 

Here's what I've tried:

 

1) Swapped back to the iFi IPower I've been using - noise went away, just like before HDPlex arrived

2) Changed the voltage output of the HDPlex down to 7V. Then back up to 9V. Same noises.

3) Tried with a USB cable that has no 5V connection - noise still there.

4) Removed Ethernet cable from microRendu - noise is still present, but there were fewer bursts, and the noise became a repetitive sound at about 1 second or less intervals.

5) Put a WfS Recovery between microRendu and DAC - powered by the HDPlex, with microRendu using iFi IPower. Noise went away.

 

I'm not sure what's going on here. Any help?

 

One thing I found very interesting was #4 above. The pattern of the bursts of noise changed significantly when the ethernet cable was removed. This is making me think that somehow I'm getting CPU noise from microRendu passing to the USB data outputs or ground. Could this be a grounding issue? I'm just using the stock DC cables that came with the HDPlex.

 

Any ideas?

[genjamon]

It's a ground loop issue. What else are you powering from the HDPLEX and from which DC output?

 

Only powering the microRendu from HDPlex, and from the variable output - tried multiple setpoints from 7-9V.

[genjamon]

Can you verify that the red switch on the back of the power supply is set to the correct voltage for your area?

 

I am in contact with Larry. He has asked me to try a cheater plug for the AC incoming to the HDPlex this evening. The unit did arrive with the red switch saying 220V, and I switched it to 110V before powering up.

[genjamon]

Okay, I was just curious.

 

Thanks for asking. If we find a resolution, I'll post here.

[genjamon]

Larry reported to me this morning that the issue has been resolved.

 

Well, kind of. Yeah, a cheater plug on the HDPlex worked to eliminate the noise. However, this has me using a cheater plug permanently, which is not ideal from my perspective. I'm curious why an older version of the same unit worked fine last weekend in my system, but the one I purchased did not. Furthermore, I've gone to considerable expense to outfit my whole system with high quality power cords featuring the Furutech FI-50 carbon fiber rhodium plated connectors, and their GTX® outlets. Now one of those plug/outlet connections is interrupted by a 50 cent Walmart cheater plug. That outlet and FI-50 connector are together worth more than the HDPlex unit itself. So, I'm not exactly satisfied to leave things in their current condition without at least trying to explore why the issue exists for one power supply but not another, and if anything can be done about it for my new power supply.

 

As to John's request for a diagram, that will take me some time to produce and share, and I have a pretty busy day at work. In the interim, here are the basics:

 

On same power conditioner (20A 3 prong power cord into a single wall outlet):

microRendu (powered by HDPlex using 3 prong power cord, or iFi iPower two prong wallwart) (connected directly to router in another room via Blue Jeans 6a ethernet cable - 50ft)

Lampizator Big7 DAC with volume control (3 prong power cord to power conditioner) (microRendu USB input using hard adapter)

Line Magnetic 518IA amp (3 prong power cord to power conditioner)

(1) GR Research/Rythmik sealed sub (speaker level connection) (3 prong power cord into power conditioner)

 

On different power outlet, same house circuit as rest of system:

(1) GR Research/Rythmik sealed sub (speaker level connection) (3 prong power cord into wall)

 

On different house circuit (another room) - all connected via Blue Jeans 6a ethernet:

TPlink AC router (2 prong wallwart)

Centurylink DSL modem (2 prong wallwart)

MacMini as DLNA server using Minimserver (2 prong power cord, internal power supply)

[genjamon]

Thanks for these thoughts, John. Regarding the question of whether this is microRendu or some other digital equipment, I would refer you to my original post, and #4 in my list of experiments. With the ethernet cable disconnected from microRendu, I was hearing the noise, but it was at regular 1 second or less intervals. A subsequent poster suggested this was all packet noise, and the regular interval noise was the microRendu trying to ping the server without the ethernet cable connected. At any rate, the noise was there with no connection to any other computer, so I think by process of elimination this must mean it's the microRendu.

 

As for running your leakage current experiment, I'm not sure if I have a computer that would work for this. Would the WfS Recovery work? Or how about leaving the microRendu USB output disconnected from the audio system, powering it with the HDPlex, and running the wire from PS negative output to RCA shell of audio system? Would that suffice as a test?

 

First we need to find out if the HDPLEXes have the negative output tied to the GND pin of the power plug. You can test this with your meter. See if both supplies are the same or different in this regard.

 

Next I want to see if we can track down the source of the ground current, it sounds like it is network related, I want to track down whether it is from the microRendu or some other digital equipment (other computer, router etc). A possible way to do this is to put extra work on the router and see if that makes any difference in the noise. For example try having another computer stream an online movie at the same time you are listening (assuming that streaming goes through the same router). If it is coming through the router (and its power supply) the noise should change.

 

Next is to try and change the supply load profile of the microRendu ie make it work harder. I'm not quite sure how to do that, what I want to do is have it do something like resampling or run some other compute intensive process at the same time, but I'm not sure how to do that with the existing interface to the microRendu.

 

What I am leaning towards right now is that new PS is somehow modulating its leakage current to the ground pin by the normal load variations of the microRendu, and the other supplies are not.

 

One way to test this is to power a different computer (not part of the audio chain) off the new PS, (something like a squeezebox, something that can be powered by the new PS, it doesn't have to have anything to do with audio, a smartphone powered off the PS would do fine), and connect a single wire from the negative output of the new PS to an audio system ground (the shell of an RCA jack is good). If it is leakage current through the PS you will make a ground loop with that wire and should hear it. If you do hear it, try it with one of the other supplies.

 

Note that all power supplies have leakage current, but if the assumption above is correct it sounds like the new PS is somehow changinging its leakage current with the fairly small load variations of the microRendu. An interesting cause of this can actually be a mains noise filter, certain designs cut down on noise between hot an neutral by shunting it to the third pin, increasing the leakage current on the ground.

 

John S.

[genjamon]

It appears David at MCRU has been listening and at a better price Regulated Linear Power Supply for UpTone Audio "UltraCapâ„¢ LPS-1 - MCRU

 

At 295 GBP, why not just go with an HDPlex? It's $400, and you get five different output voltages, two of which are variable for great versatility, two of which could independently power LPS-1's (the 12V fixed, and the 5-9V variable), and it's a great linear power supply in its own right, which blows away the iFi level or any SMPS by a mile even powering microRendu and other electronics on its own. Seems like way more of a bargain for similar price than the MCRU...

[genjamon]

I have been borrowing an Ultra from a friend for the past week.  I was able to connect directly to my Lampizator DAC with a USPCB just fine.  I use a Supra Cat8 ethernet cord, and was able to torque it to bend within the space available.  And my power connector is the cheapo screw terminal type, because I use the DIY DC cable recipe recommended by John Swenson using Canare 4S6 star quad wire.  This wire is very flexible, and easily bends with little need for any extra space clearance beyond the power connector itself.  So, all connections fit pretty well and easily in the space allowed by the USPCB.  I suspect if I had a nice commercial Oyaide connector body DC cable design it might be a different story, but am not sure about that.  

 

But yeah, the microRendu format is definitely superior for these kinds of direct connections to the DAC.