Hi @abraxalito
I have a strange problem that I can't seem to fix. I am trying to put my system together so I can switch back and forth between the Phidac and the Celibidache. However, I am getting a ground loop issue (at least I assume it is that). Here are the main components and how they are connected.
Setup 1. SMPS 28v. I had it hooked up with all three wires: main, neutral and ground. I used this to power the whole system. It went like this:
28v -> TPS7a4 regulator to 24v to Celibidache.
28v -> TSP7a4 19v -> LT3045 12v to Phidac.
28v -> BPM15-150-Q12-C to +15v 0v -15v for headphone amp.
There is a switch that turns on either the Phidac or the Celibidache.
There is a switch that selects the output from either the Phi or the Celi to go to the amp.
This works as it should.
Setup 2. I have a recycled amplifier case I wanted to put the SMPS into but it only has main and neutral. No Ground Wire. All components are hooked up the same but now there is noise in the system. The fact that the I2S are all connected did not make a difference since I tested it with disconnecting one Dac and the other. No difference.
The BPM15-150-Q12-C gets warmer and sometimes when I touch it, the noise goes away. If I power the BPM15-150-Q12-C from a separate power supply, the noise goes away.
Is this a problem because the BPM15-150-Q12-C is Isolated? I tested the center tap from the BPM15-150-Q12-C to it's ground on the input side and measured continuity.
Having multiple PSUs is not ideal, so if there is a suggestion to fix either this circuit or, alternatively, is it safe to run another power cable (that includes a ground wire) to the box and connect the ground to the chassis and then the SMPS to the same point on the chassis?
I have a strange problem that I can't seem to fix. I am trying to put my system together so I can switch back and forth between the Phidac and the Celibidache. However, I am getting a ground loop issue (at least I assume it is that). Here are the main components and how they are connected.
Setup 1. SMPS 28v. I had it hooked up with all three wires: main, neutral and ground. I used this to power the whole system. It went like this:
28v -> TPS7a4 regulator to 24v to Celibidache.
28v -> TSP7a4 19v -> LT3045 12v to Phidac.
28v -> BPM15-150-Q12-C to +15v 0v -15v for headphone amp.
There is a switch that turns on either the Phidac or the Celibidache.
There is a switch that selects the output from either the Phi or the Celi to go to the amp.
This works as it should.
Setup 2. I have a recycled amplifier case I wanted to put the SMPS into but it only has main and neutral. No Ground Wire. All components are hooked up the same but now there is noise in the system. The fact that the I2S are all connected did not make a difference since I tested it with disconnecting one Dac and the other. No difference.
The BPM15-150-Q12-C gets warmer and sometimes when I touch it, the noise goes away. If I power the BPM15-150-Q12-C from a separate power supply, the noise goes away.
Is this a problem because the BPM15-150-Q12-C is Isolated? I tested the center tap from the BPM15-150-Q12-C to it's ground on the input side and measured continuity.
Having multiple PSUs is not ideal, so if there is a suggestion to fix either this circuit or, alternatively, is it safe to run another power cable (that includes a ground wire) to the box and connect the ground to the chassis and then the SMPS to the same point on the chassis?
@abraxalito Since the thread for the Phiamp is closed and I am running it from the Celibidache, I will post my question here 
Based on the picture you attached, I do have some questions, (only because it is not working for me). The version I have is the r2 (in the picture is r1) and the left and right channel have more resistors/capacitors but unlike in this picture they are not equally populated. The right channel has more resistors and the left has some blank spots.
I tried replicating this setup; A) both boards powered with 12v from the same source.
B) Ground and left signal to one board and ground and right signal to the other board input.
C) Based on the jumper from the input cap to the second channel and two wires going from the board output to the headphone adapter, both channels are running in parallel and get hooked up to the headphone plug together. (If I am interpreting this picture correctly)
D)The headphone jack is not grounded to the boards but goes to the DAC ground.
Very first test I tried one board, and it produces sound in one channel only, as it should (since these are supposed to be two boards paralleled) but it was grounded from the board to the headphone amp. However, when I hook up both boards according to this image, I do not get full stereo, just a very faint volume in one side only. So my question is: Since version r2 is different (more resistors/capacitors) and one channel is not populated like the other, what is the correct way to run them in parallel? Do I still jumper from the input cap to the other channel? Do I only attach one output instead of bridging the left and right outputs per channel?
Based on the picture you attached, I do have some questions, (only because it is not working for me). The version I have is the r2 (in the picture is r1) and the left and right channel have more resistors/capacitors but unlike in this picture they are not equally populated. The right channel has more resistors and the left has some blank spots.
I tried replicating this setup; A) both boards powered with 12v from the same source.
B) Ground and left signal to one board and ground and right signal to the other board input.
C) Based on the jumper from the input cap to the second channel and two wires going from the board output to the headphone adapter, both channels are running in parallel and get hooked up to the headphone plug together. (If I am interpreting this picture correctly)
D)The headphone jack is not grounded to the boards but goes to the DAC ground.
Very first test I tried one board, and it produces sound in one channel only, as it should (since these are supposed to be two boards paralleled) but it was grounded from the board to the headphone amp. However, when I hook up both boards according to this image, I do not get full stereo, just a very faint volume in one side only. So my question is: Since version r2 is different (more resistors/capacitors) and one channel is not populated like the other, what is the correct way to run them in parallel? Do I still jumper from the input cap to the other channel? Do I only attach one output instead of bridging the left and right outputs per channel?
So it seems one board works and the other does not. The diode lights up on one and not on the other. I have taken some voltage measurements while powered on of the 4 capacitors as arranged in the photo going clock wise. Input voltage is 11.5v
Cap 1 on the good board is 10.02v. On the bad board is 3.25v
Cap 2 (the one in the middle) 11.48v on the good board. 11.48 on the bad board.
Cap 3 on the good board is 10.01v and 4.07 on the bad board.
Cap 4 on the good board is 5.57v and on the bad board is 2.4v.
Cap 1 on the good board is 10.02v. On the bad board is 3.25v
Cap 2 (the one in the middle) 11.48v on the good board. 11.48 on the bad board.
Cap 3 on the good board is 10.01v and 4.07 on the bad board.
Cap 4 on the good board is 5.57v and on the bad board is 2.4v.
The LED feeds the reference voltage which is used to generate the supply rails for the amps. With the LED off the reference voltage is not going to be biassed. Cap4 (I'm taking it that this is the bottom left on in the pic) is across the reference. The power to the amps is multiplied up from the Vref by 1.8X. Caps 1 & 3 are across those rails. So the investigation of what's gone wrong should start by looking into why the LED isn't lighting up. The LED is fed by current source Q1,2 (next to power inlet) and a check of the voltage across the current source resistor (R1 I believe) should reveal whether the CCS is working.
Just a heads up that grounding the power board did indeed solve my issue. Would you happen to know what kind of fuse I should install on this sucker at my inlet? I’d like to add one.
I’m currently using the Celibdache in a budget desktop set up with a Fosi V3 class d amp and some thrifted Elac Debut B6 and a motherboard toslink out. Sounds pretty great. Detailed and unfatiguing.
Sparkos op amps are waiting to be installed on the v3 as IMO the combo is currently a little on the polite and uninvolving side.
I’m currently using the Celibdache in a budget desktop set up with a Fosi V3 class d amp and some thrifted Elac Debut B6 and a motherboard toslink out. Sounds pretty great. Detailed and unfatiguing.
Sparkos op amps are waiting to be installed on the v3 as IMO the combo is currently a little on the polite and uninvolving side.
@abraxalito Thank you for pointing me towards wondering why the LED wasn't working. It was missing.. which explains a lot. Well, I had a spare and so it is all good with the Phi Amp.
I am still struggling with the Celibidache though. I have rewired my setup a few different ways and am using different power supplies etc. and have not been able to cure my problem.
The Celi makes a shhhchhhrrrooochhhsweeeeshhhh noise. It is in both channels. I can play music from both channels, so that part works. The noise goes back and forth between them, sometimes louder in one side and then it moves to the other. It does not matter what PSU I use. So now I am considering that a component on the DAC is bad. When I touch the capacitors I can hear it, primarily in the left channel. The same goes for the inductors on the filter. Compared to the Phidac where the two diodes light up quickly, almost immediately, the Celi takes a few seconds for the LED to light up and it is a more gradual lighting up, rather than just ON. Can you help me figure out how to test for a bad component?
I am still struggling with the Celibidache though. I have rewired my setup a few different ways and am using different power supplies etc. and have not been able to cure my problem.
The Celi makes a shhhchhhrrrooochhhsweeeeshhhh noise. It is in both channels. I can play music from both channels, so that part works. The noise goes back and forth between them, sometimes louder in one side and then it moves to the other. It does not matter what PSU I use. So now I am considering that a component on the DAC is bad. When I touch the capacitors I can hear it, primarily in the left channel. The same goes for the inductors on the filter. Compared to the Phidac where the two diodes light up quickly, almost immediately, the Celi takes a few seconds for the LED to light up and it is a more gradual lighting up, rather than just ON. Can you help me figure out how to test for a bad component?
@pelopidas - I would first suspect the PSU(s) you have been using. Could they be switchers which are outputting switching noise? I'm unclear also when the noise appears - you say 'when I touch the capacitors I can hear it'. So without touching the capacitors, you don't hear it? When you don't touch the inductors on the filter, you also don't hear it?
Its normal that the yellow LED on Celi lights up slowly, there are some largeish caps which have to charge up before it can light.
Its normal that the yellow LED on Celi lights up slowly, there are some largeish caps which have to charge up before it can light.
Hi @abraxalito, so I got a 24v TPS7A4701 based regulator to test and see if it was the psu making the noise. The same issue persists. There is noise when the Celibidache is turned on. It does play music from both channels but the noise is always there. I can switch back and forth between it and the phidac and there is no noise on the Phidac. Can you tell me what measurements I should take to see if there is a component(s) that is damaged?
Sorry to keep bugging you, but is there a choke you would recommend on digikey?
Is this what I'd want to hook up in series to the transformer? https://www.digikey.ca/en/products/detail/hammond-manufacturing/159ZB/3869271
@ Abraxalito, I took some measurements:
TP1 19.2
TP2 10.1
TP3 20.5
TP4 2.5
TP5 4.9
TP6 1.7
TP7 10.1
TP8 10.1
TP9 10.1
TP10 10.1
TP11 10.1
TP12 10.1
based on previous questions that some others had, I think these are all within spec. I do measure almost 10v on the output capacitor relative to ground. I suppose that is normal too. One thing that is weird is that if I measure the top of the capacitor (the metal that shows through the sleeve) I am getting over 2v. I don't recall having had anything quite that high before. The symptom is still a constant shhhhhhchhhhh sound. When I touch the top of the output capacitors it reduces that noise.
Those voltages all look fine, I agree. 10V on the output cap relative to 0V is also fine. Measuring exposed metal on electrolytic caps I'm not sure tells us anything much. However the fact of touching the cap's top reducing the noise is a very interesting datapoint. It makes me wonder if the noise results from some kind of unwanted oscillation in the I/V stage. Touching the output cap wouldn't normally have any effect as the output impedance is low.
I think try powering the DAC up but without the filter module plugged in. This might give an indication whether the noise is coming from the output (I/V) stage. Having no filter disconnects the DAC chips from the output so if the noise persists (there won't be any music, irrespective of input) then it must come from the I/V stage.
I scrolled through a dozen or more pages of CM chokes on Digikey and didn't find any using the latest flat wire construction, Digikey are overdue for updating their offerings in this department. The Digikey offerings being last century tech won't be as performant as those on Ali but you might try one like this : https://www.digikey.com/en/products/detail/epcos-tdk-electronics/B82731M2401A033/1243528. The Hammond choke you linked isn't a suitable one.
I removed the filter and powered it up and the noise is there. I hear it on both channels.