I'm going to take you back to school science lesson.
You have a battery and a bulb and some wires.
The "problem": get the battery to light the bulb.
The "solution":
wire one terminal of the battery to one side of the bulb.
wire the other battery terminal to the other side of the bulb.
It works !!! the bulb lights.
Now break one wire and reconnect it. did the bulb go out?
Now break the other wire and reconnect. did the bulb go out?
You have a circuit. The current flows from the Source (battery) down one wire and RETURNS to the Source (battery) via the other wire.
The circuit STOPS working if you break either wire.
To operate the circuit NEEDS a Flow and Return.
The current that flows out of the source MUST return to the Source.
Now we have left school, we have become adults.
The rules for circuits stay the same.
The amplifier is the Source driving the speaker.
The CDP is the Source driving the Pre-amp.
The volume control is the source driving the Power amplifier.
The INPUT socket on the Power amplifier case is the Source driving the amplifier circuit board.
Every Source that drives a Receiver must have a two wire connection.
All your audio Sources work the same way.
You have a battery and a bulb and some wires.
The "problem": get the battery to light the bulb.
The "solution":
wire one terminal of the battery to one side of the bulb.
wire the other battery terminal to the other side of the bulb.
It works !!! the bulb lights.
Now break one wire and reconnect it. did the bulb go out?
Now break the other wire and reconnect. did the bulb go out?
You have a circuit. The current flows from the Source (battery) down one wire and RETURNS to the Source (battery) via the other wire.
The circuit STOPS working if you break either wire.
To operate the circuit NEEDS a Flow and Return.
The current that flows out of the source MUST return to the Source.
Now we have left school, we have become adults.
The rules for circuits stay the same.
The amplifier is the Source driving the speaker.
The CDP is the Source driving the Pre-amp.
The volume control is the source driving the Power amplifier.
The INPUT socket on the Power amplifier case is the Source driving the amplifier circuit board.
Every Source that drives a Receiver must have a two wire connection.
All your audio Sources work the same way.
Last edited:
Dude, what? I do have a correct, working signal path. I have no idea where you got the idea that I was using the shield as a conductor - I'm not. The braided shield in these cables serves an actual SHIELD here, not a signal return - it's not meant to conduct anything but the currents induced by EMI it picks up. I'm using ONE coax cable for the "hot" signal wire and ONE coax cable for the signal return. So to transmit a "complete" stereo signal (2x signal, 2x ground), that makes 4 coax cables in total (rather inefficient, I know, but that cable was all I had at the point I re-cabled the amp). You can see this very well on the second picture - there is a total of 4 black wires going to each side of the pot, two each are fused together by their isolation and inside each of those black wires, there's an isolated conductor that's covered by the shield. Again, I'm only using the inner conductor as signal/return paths. The shields of all these cables are connected to the preamp's signal ground at the same point - I see nothing intrinsically wrong with this concept. But maybe you know something I don't? For example, I could imagine the parasitic capacitance between the signal "hot" lines and the grounded shield could be causing problems - I really don't know though. I now also have twisted pair shielded cable available, maybe that would be better suited as it would cause the conductors covered by the shield to carry zero current in total?
I just want to use shielded cable of some form (again, using the shield as a "true" EMI shield, not as the return conductor as with a coax cable) for the signal lines I've got running through the whole case to the volume pot at the front because they pass by the toroidal power transformer for the preamp rather closely and I'd like to keep them from picking up any interference.
Edit: You just posted another reply while I was typing up the above, and I have to say that your tone is becoming rather condescending. I study electrical engineering and while I'm not a master of my trade by any means, I do know how electronics and electrical circuits work at a basic level and there's absolutely no reason to "take me back to school science lesson". You're basing your posts on erroneous assumptions and jump to conclusions that are absolutely wrong - if I really had broken my signal return path, I highly doubt I'd even be able to hear any music playing when connecting the amp to a source and speakers.
Last edited:
A coaxial cable carrying signal is a two way conductor.
The shield carries the return current to the source.
The shield carries the return current to the source.
You're a stubborn one, aren't you? I'm not using my coax cables that way! The shield, in my use case, serves as a shield, not a return conductor! See my above post for more details.
My question is: Is this bad practice? If so, why?
Yes it's bad practice.
The signal wires whether at high current, eg. speaker and PSU, or low level, eg. CDP or turntable must be close coupled.
This reduces interference effects.
Splitting a Flow and Return by using TWO coax, one for Flow and one for Return is WRONG.
It does not matter that you have a one end connected shield, the split, rather than close coupled creates the LOOP AREA that introduces interference.
go to two sets of twisted pair and get rid of the 4 coax.
The signal wires whether at high current, eg. speaker and PSU, or low level, eg. CDP or turntable must be close coupled.
This reduces interference effects.
Splitting a Flow and Return by using TWO coax, one for Flow and one for Return is WRONG.
It does not matter that you have a one end connected shield, the split, rather than close coupled creates the LOOP AREA that introduces interference.
go to two sets of twisted pair and get rid of the 4 coax.
interference
Alright, thanks! I'll also change the power wires from pair-twisted (right now the GND connections are just "flying" through the chassis on its own) to "3-braided" (or twisted? Which is better? I prefer the look of tightly braided wires but if twisting them gives better interference performance, I'll go with that) - I don't know why I didn't think of that from the beginning. Ill keep the speaker return wires set up the way they are now, as per your advice. It always seemed to make more sense to have it set up like that, but as I said, Rod explicitly states the speaker return path should go directly to the PSU on his construction and design pages - for what reason, I don't know.
Last edited:
Use twisted.
Braided increases loop area and thus increases inductance and increases interference.
Braided increases loop area and thus increases inductance and increases interference.
Ultimately the Speaker Return does go all the way back to the PSU Zero Volts.
That PSU is the Source and all the current that Flows out of the Source MUST Return to the Source.
But you have to maintain minimum Loop Area.
That means that the actual layout MUST take the speaker return alongside the Speaker flow wire. Then it must run alongside the trace/s that take the power into the PCB output devices. Then it must run alongside the power supply wires to the Main Audio ground. Then it becomes the Main audio ground to PSU Zero Volts wires which is twisted as a triplet with the supply wires.
The Return route MUST mimic as closely as possible the Flow route. Where it can be twisted, do so.
I find this Flow and Return LOOP AREA so simple to implement that I have trouble understanding why so many thousand Forum Members think anything else can do the same job.
Last edited:
I just redid the small-signal connections between preamp and power stage with shielded twisted-pair cable. I left the shield connections as they were before - grounded to the preamp's ground. This resulted in rather strong hum if the cables from the preamp to the pot werent run exactly parallel and close to each other - I imagine this is because the EMI-induced currents in the two channels are different, which "contaminates" the preamp signal ground. Connecting the shields (and with them, the volume pot's body/shaft) to the power stage's ground appears to fix the problem (I honestly don't know why), leaving me with only a very quiet hum that doesn't change with the volume pot's setting - I'm not yet sure what is causing this, but it might be the improperly twisted power lines. I'll try twisting them the right way and will report back with the results. I also still have to isolate the preamp circuit from the chassis - I could imagine that causing some hum as well.
The bottom line so far is that making a really (as in, ear 0.1" away from a driver) quiet integrated amp is a bit more challenging than simply throwing together preamp and power amp circuits in the same chassis.
The bottom line so far is that making a really (as in, ear 0.1" away from a driver) quiet integrated amp is a bit more challenging than simply throwing together preamp and power amp circuits in the same chassis.
Last edited:
Alright, I did all the small-signal cables again (this should be the last time...). It seems I was wrong and connecting the shields to the P101 PSU ground does not fully fix the problem with the hum/buzz picked up by the shields - I was getting the same thing as before, just more quiet. I have now twisted all 4 signal cables (2 to the pot, 2 from the pot) together and that seems to really eliminate this problem. I just think there must be something wrong that I have to twist them together in the first place - what's the point of having shields if you end up having to twist the shielded cables together anyway? Maybe someone who knows more about shielding could point me to where I went wrong with my approach?
I'm now left with an almost inaudible hum that I'm 99% sure comes from the improperly twisted power cables (it doesn't change volume with pot setting, but I can make it more quiet by moving the power cables around) and some very quiet white noise that does change its level with the volume pot. I'm not quite positive on the origin of this noise as it disappears with shorted inputs - however, that also makes me think that I won't be able to fix it by making changes to the amp. Maybe isolating the preamp from the earthed chassis will help. I would have already done that but my local hardware store didn't have nylon screws - they do carry nylon nuts but those aren't very useful by themselves. I've now ordered a bunch of nylon screws on ebay but they'll take a while to get here from China. Perhaps I'll find something locally before they arrive.
What I also thought was noteworthy was that I got what seemed to be a HUGE ground loop (there was hum loud enough to be heard through a closed door) when I connected my phone, which was charging from my PC, to the amp. My gamma2 DAC doesn't give this problem (it does, however, seem to produce some white noise on its own as well as what sounds like it could be SMPS noise - luckily I'm one Mouser order away from finishing my sigma25 to finally give it the power supply it deserves), nor did the phone when it was disconnected from the PC.
Here's a pic of the current wiring setup:
The red/blue cables carry one stereo signal each (red for right, blue for left), and as you can see the shields and the pot are now all grounded to the P101's PSU ground rail.
I'm now left with an almost inaudible hum that I'm 99% sure comes from the improperly twisted power cables (it doesn't change volume with pot setting, but I can make it more quiet by moving the power cables around) and some very quiet white noise that does change its level with the volume pot. I'm not quite positive on the origin of this noise as it disappears with shorted inputs - however, that also makes me think that I won't be able to fix it by making changes to the amp. Maybe isolating the preamp from the earthed chassis will help. I would have already done that but my local hardware store didn't have nylon screws - they do carry nylon nuts but those aren't very useful by themselves. I've now ordered a bunch of nylon screws on ebay but they'll take a while to get here from China. Perhaps I'll find something locally before they arrive.
What I also thought was noteworthy was that I got what seemed to be a HUGE ground loop (there was hum loud enough to be heard through a closed door) when I connected my phone, which was charging from my PC, to the amp. My gamma2 DAC doesn't give this problem (it does, however, seem to produce some white noise on its own as well as what sounds like it could be SMPS noise - luckily I'm one Mouser order away from finishing my sigma25 to finally give it the power supply it deserves), nor did the phone when it was disconnected from the PC.
Here's a pic of the current wiring setup:
The red/blue cables carry one stereo signal each (red for right, blue for left), and as you can see the shields and the pot are now all grounded to the P101's PSU ground rail.
Last edited:
Triple-posting because I noticed a mistake in my last post and the edit timer has run out again: Obviously one red/blue doesn't carry a stereo signal, but rather a mono signal. One red and one blue cable form a stereo pair.
Also, I have now noticed a (very quiet) "hum" component in the noise that changes its volume with the pot setting which I didn't notice before - again, it sounds like the two hums are slightly out of phase. This second hum also doesn't disappear when I short the inputs of the preamp, so I guess something's still a bit fishy. Might be the chassis connection of the preamp?
Also, I have now noticed a (very quiet) "hum" component in the noise that changes its volume with the pot setting which I didn't notice before - again, it sounds like the two hums are slightly out of phase. This second hum also doesn't disappear when I short the inputs of the preamp, so I guess something's still a bit fishy. Might be the chassis connection of the preamp?
Last edited:
Try measuring the Hum + Noise at the output, to enable comparisons of different connections/solutions.
A DMM with a 200.0mVac scale does this job adequately.
0.5mVac of Hum and Noise is a lot.
0.2mVac is just about adequate.
0.1mVac is nearly good.
0.0mVac ( meaning < 0.05mVac) is good.
You need a better measuring system for any lesser amounts of Hum and Noise.
But 0.05mVac in a 50W into 8ohms power amp is 20Vac/0.00005Vac = -112dB S:N ratio for the bandwidth of the DMM.
A DMM with a 200.0mVac scale does this job adequately.
0.5mVac of Hum and Noise is a lot.
0.2mVac is just about adequate.
0.1mVac is nearly good.
0.0mVac ( meaning < 0.05mVac) is good.
You need a better measuring system for any lesser amounts of Hum and Noise.
But 0.05mVac in a 50W into 8ohms power amp is 20Vac/0.00005Vac = -112dB S:N ratio for the bandwidth of the DMM.
My DMM shows 0.000VAC at the output - the lowest setting it has is the 2V scale. I'll take the amp to Zurich one of these days, we have some quite good oscilloscopes at the university that I can use to get a clearer picture. I also found some nylon screws in the electronics store so I'll be isolating the preamp from the chassis tonight.
a 2.000Vac scale is not low enough for measuring and comparing Hum+Noise at the output.
A scope is not good at very low levels either.
A 2mV/div setting allows you to see a ripple of around 1mm high if there is no noise. That 1mm of peak to peak ripple is equivalent to 0.2mVpp
But if there is 1mm, or 2mm, of noise in the signal, it will swamp the ripple. You would need 0.5mVpp of hum to see the ripple rise clearly out of the noise.
A scope is not good at very low levels either.
A 2mV/div setting allows you to see a ripple of around 1mm high if there is no noise. That 1mm of peak to peak ripple is equivalent to 0.2mVpp
But if there is 1mm, or 2mm, of noise in the signal, it will swamp the ripple. You would need 0.5mVpp of hum to see the ripple rise clearly out of the noise.
Last edited:
Unless you have a decent quality 4.5 digit meter with bandwidth better than a few kHz (difficult to find this specified) or a likely very expensive Audio Millivoltmer, this is beyond the "scope" of most DIYs. Few have more than a low cost 3.5 digit DMM.
For those who like buying add-ons for their PC, a USB soundcard is a great option for all types of precise measurements down to the DAC resolution limits and within its input signal voltage limit. Check the discussion of other options here: http://www.diyaudio.com/forums/equipment-tools/259615-any-rms-millivoltmeter-recommendations.html
I quite like this one, for a ready to install system: http://www.diyaudio.com/forums/equipment-tools/231401-quantasylum-qa400.html
For those who like buying add-ons for their PC, a USB soundcard is a great option for all types of precise measurements down to the DAC resolution limits and within its input signal voltage limit. Check the discussion of other options here: http://www.diyaudio.com/forums/equipment-tools/259615-any-rms-millivoltmeter-recommendations.html
I quite like this one, for a ready to install system: http://www.diyaudio.com/forums/equipment-tools/231401-quantasylum-qa400.html
Yeah, as I said, I'll be taking the amp to the university one of these days to see how much the remaining hum has. Also, I believe I've identified another potential source of hum/ground loop: The pot, which makes direct contact with the case (no way around that I think...), is now grounded to the P101 PSU - which means that the PSU is connected to chassis/earth ground via the pot. That doesn't sound too good to me from a ground loop perspective, but I don't know what to do about it - I have to ground the pot somewhere, and it's inevitably going to be making contact with the chassis. Would grounding the pot to the preamp's signal ground be better?
Also, the preamp is now isolated from the chassis - I haven't noticed much change though. There's still a tiny bit of hum coming from somewhere before the power amps (i.e. it changes volume with the pot's setting).
And again, the remaining noise is VERY quiet - I have to go right up to the speaker with my ear to hear it, but I still believe that I can make it disappear completely.
Also, the preamp is now isolated from the chassis - I haven't noticed much change though. There's still a tiny bit of hum coming from somewhere before the power amps (i.e. it changes volume with the pot's setting).
And again, the remaining noise is VERY quiet - I have to go right up to the speaker with my ear to hear it, but I still believe that I can make it disappear completely.
The Vol Pot is the Receiver for the input signal. It needs a two wire connection to that input signal.
The Vol Pot is the Source for the Power amp's signal. It needs a two wire connection to the amplifier.
That's it.
It does not need to be "grounded".
If the pot has a metal casing, or a metal control shaft, then those metal parts should be connected to Chassis.
The Vol Pot is the Source for the Power amp's signal. It needs a two wire connection to the amplifier.
That's it.
It does not need to be "grounded".
If the pot has a metal casing, or a metal control shaft, then those metal parts should be connected to Chassis.
A few things:
-As you guessed correctly, the Alps RK27 pot I'm using has a metal shaft and body. It introduces a rather loud hum into the signal when the body is not grounded.
-My pot is not the receiver for the input signal, it sits between the preamp and the power amp. This creates some ambiguity as to where it should be grounded, as those two grounds are not directly connected.
-I'm 99% sure that I need some kind of connection between my signal ground(s) and chassis/earth ground, as the floating ground from the isolated P101 PSU is at ~88VAC from earth when not connected to it - this means that there would be 88VAC on the heatsinks too, for example, which would obviously interfere with the operation of the amplifier (I'm sure the MOSFETs wouldn't appreciate their heatsinks being at that potential, for example). However, the problem now is that right now, the only connection between chassis/safety earth and P101 ground is via the pot's body - I think this is undesirable (at least it "feels" that way), but I'm not sure how to approach the problem (if it even is one - but I can't imagine any other source of hum before the power amps now that the preamp is totally isolated). I've only had two ideas so far:
a) connect preamp ground to chassis and remove the grounding wire between the P101 ground and the pot's body, or
b) somehow isolate the pot from the chassis, ground it to the P101 PSU ground and install a ground loop breaker between the PSU ground and safety earth so my signal ground is at a well-defined potential in relation to chassis ground.
I do not know whether any of the above make any sense, whether the pot is actually the source of the remaining hum coming from before the power amp or if there are any other (better?) ways to solve this. The only thing I'm absolutely sure of is that my pot indeed does need to be "grounded".
On a completely unrelated note, I've finished my sigma25 PSU for my gamma2 DAC - took me ages to get the placement of everything right and keep the wires from blocking each other's way as I managed to choose impractically large parts for nearly everything there was: A huge (but very nice-looking IMO) power switch, a big IEC socket, oversized and stiff wires and of course a rather big 6VA chassis-mount transformer. Stuffing all that into the B3-080 box enclosure (to match the gamma2's look) was quite the challenge - retrospectively, my motto seems to have been "measure once, drill twice" (below you see the two sets of holes for the sigma25 PCB - the second was drilled when I noticed the IEC power socket was too large for my initial placement of the PCB):
In the end, everything worked out fine though and the ugly holes are not visible from the outside without turning the box upside down - here it is on top of the DAC next to the Millett MAX:
I honestly can't hear any difference in sound, though.
Last edited:
You are wrong.
The vol pot receives the signal from the input socket.
It processes the signal.
The vol pot sends the processed signal to the power amplifier (or what ever stage comes next).
The vol pot acts as BOTH Receiver and as Source.
Understand that and you then stand a chance of getting the wiring right.
A signal NEEDs a two wire connection.
Use TWO wires as a twisted pair from input socket to vol pot.
Use TWO wires as a twisted pair from vol pot to amplifier.
Are you being serious? I explained to you how I set up my amplifier, and you just tell me I'm wrong? The layout of my amp may be stupid and my ideas retarded, but I know for an absolute fact that my pot sits BETWEEN THE PREAMP AND THE POWER AMP BOARDS. Check the pictures I posted. I really do know enough about electronics to be able to tell you the signal path of my amp. And no matter how many times you tell me that I have to use "two wires" to carry a signal, it won't change anything about the fact that I've been doing just that since before starting this thread and have, from the beginning, had the pot wired correctly. Again, verify this on the pictures if you'd like. I don't know where you get the idea that you need to tell me in every one of your posts that I need two wires to transmit a signal. I know that, and I already knew it before you told me. I know how a potentiometer works and I know just as well that the Alps Blue Velvet pot causes buzzing when its body is not grounded. These are facts, just like the fact that I have my pot sitting between the preamp circuit and the power amp circuit, both of which are in the same chassis to form an integrated amplifier (as I've explained more than a few times now). The reason I did this has already been discussed and so far, no one has offered a reason to do it any other way - having the pot right after the input would require a lot of additional wiring and does not seem to offer any real benefits, as the P101's input impedance is high enough for it to be driven by the pot.
Anyway, it seems that just connecting the preamp ground point to the chassis causes an improvement (i.e. decrease) in hum volume - that's also handy because I can now just leave the pot's body alone as it touches the chassis anyhow and is thus grounded. There's still something left though - both "before" (changes volume with pot setting) and "after" the pot (not affected by pot setting). The next step will be (finally) twisting those power wires correctly and running them to the amps on a better path so they don't get the chance to pick up any transformer hum in the first place.
Last edited:
- Status
- Not open for further replies.