Old Motifator threads are available in the Archive.
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
Hi All, ok, the setup is fairly simple. computer-usb-MOX-unbalanced-Focusrite Preamp - balanaced- SM57 the problem arises as i have a ground at both the preamp, AND the pc. i have tried an ART DTI in various configurations, however i am still able to record ground loops in A/D recordings does anyone have a solution around this? i have even gone and got a grounded power supply 12V DC regulated for the MOX itself, albeit to no affect.
i have also tried connecting a ground (alligator clips) from the ground pin of a spare power point in the power board, to a metallic spot on the usb cable (the shield) which has reduced the noise considerably.
i am at a loss as to waht to do to further reduce the noise/get rid of it completely. (please note this also occurs if i run the MOX directly into the unbalanced connection of my KRK10s monitor sub. in fact, you can substitute ANY usb interface device, as long as the chain goes: Computer-USB-**interface device**-UNBALANCED-**Grounded device** How does everyone else beat this problem? i could combat the output side of it by putting a mixer in and converting to balanced before the sub, but this doesnt work on the input side, as i only have 1 mixer, and the mox (the recipient) is unbalanced) i would very much appreciate anyone’s thoughts on this issue, as i have been fighting it for weeks :( |
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
Please note that the case where i was observing the pc noises on recordings (ground loop) i did NOT have the speakers plugged in until i did the playback, hence eliminating the possibility of the speakers a factor (beyond the fact that they have a transformer and are grounded, just like the preamp) and leading me to conclude that having 2 grounds is the issue. |
Dily
Total Posts: 258
Joined 11-30-2011 status: Enthusiast |
Try taking your computer off your current setup and see if you will still get GL. |
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
tried that. the psu noises go away, however if i have the speakers and the preamp plugged in at the same time, i still hear a SMALL amount of 50Hz, indicating there is probably a ground loop, however as there is no major source of “garbage” down the signal, ie computer noises, its not so obvious. i should meantion that the 50Hz is present while i have the preamp plugged in, even when the A/D inputs are off, indicating this is an electrical thing rather than simply noise from my preamp. so basically as long as i have 2 grounded connections to the keyboard, i will get noises. when i only have the speakers plugged in, all is well. no ground loop noises at all. introducing either USB (PC) or a grounded input to the A/D inputs, will result in noises. |
meatballfulton
Total Posts: 3022
Joined 01-25-2005 status: Guru |
It’s obviously the computer causing the noise, a common problem. Surprised the DTI didn’t fix it. Is everything plugged into the same outlet/circuit? What is plugged into the computer besides the USB cable to the MOX? |
rogs
Total Posts: 121
Joined 07-16-2011 status: Pro |
I thought you’d found a good solution using optical, as you mentioned in your thread here? : http://www.motifator.com/index.php/forum/viewthread/457630/ Maybe not then?? I still think, as you can read in my comments in that thread, and the other one I linked to, that only transformer or optical isolation will totally cure the problem. (Especially as you have now confirmed that isolating the USB ground doesn’t cure it!) i.e. you need true ‘galvanic isolation’ to do the job properly. Unless, as you have asked, someone has found an alternative, non ‘galvanic’ way....? Not holding my breath on that one..... |
meatballfulton
Total Posts: 3022
Joined 01-25-2005 status: Guru |
He has an ART DTI which is a stereo isolation transformer. But the ground noise is coming from a digital connection...the USB cable between MOX and computer is creating the loop. When he removes the USB cable, the noise disappears. So there’s no place to patch in an isolation transformer. I doubt we will be able to help him diagnose it without being there in person. |
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
Hi Rogs and Meatball, i appreciate you helping me, we’ll have to see how we go! As i have said, i attached a grounding lug to the middle of a USB extension cable, which reduced the noise substantially, AND i blocked the grnd pin of the USB cable which further reduced the noise. It seems to me that the key is in grounding the USB cable.. I am thinking about optical isolation devices that convert USB into optical and then back again, however i have yet to find anything USB 2.0.. I have also considered how i might modify a USB cable to either float it, or ground it… Last night i completely unplugged the speakers, and used headphones so i could completely remove the speakers/mixer/etc as a part of the system. i observed that having the PC-USB-MOX-HEADPHONES (with the preamp plugged in, but the A/D inputs off) had no ground loop. THEN, i turned on the A/D inputs, but have the gain knob on minimum, hence no actual signal of the preamp was sent (hence it was essentially like attaching a ground connection to the A/D inputs). THIS had ground loop. as i raised the gain, i got to the point where the white noise floor drowned out the ground loop noise. as to what else is plugged into the pc, i have LED 2 monitors connected via DVI and a single power point (splitter cable). these are on the same power board as the PC. I have tried unplugging the power from the monitors.. it has had no effect i have an ungrouded, powered USB hub. i have unplugged it from the PC, .. no effect I have a logitech G15 keyboard and wireless G700 mouse both connected directly to the back of the pc. i have an ethernet cable going to a LAN network, however they are already isolated, but for the sake of it, i unplugged it anyway… no effect. the pc and monitors are on 1 power board, and everything else (speakers, mox, yamaha mw10c ungrounded mixer, and focusrite ISA ONE are all on the other power board. this is necessary because i turn the speakers and everything else off at the wall, but leave the pc with power (even when off) so that i dont drain the CMOS battery… I have however tried plugging everything into the one power board, with no effect. i have also tried using a grounded power supply for the keybaord, with (interestingly) no effect. i know (tested with the oscilloscope) that the usb and the MOX input neutral terminals are both tied together, so i would have thought that using a grounded power supply would have grounded the USB at the MOX end and hence got rid of the problem. i think i may have to instead ground the USB at the PC end.... any ideas??? (btw, i checked that the ground on the power supply was tied to the neutral on the power supply, so it was in fact getting used etc. @ rogs: optical allowed me to run a cable from the pc’s optical output to the mixer in paralell with the keyboard without reintroducing a ground loop. i cannot use any optical based audio in this scenario as it would require going both to and from optical which would induce a LOT of noise on the input, hence kind of undermining the whole “get a good preamp” etc. again thanks for your help and ideas, i have been dealing with this for weeks (ever since i started using the MOX USB interface) and i have begun running out of ideas and outside input is MOST appreciated! |
ThrashJazzAssassin
Total Posts: 42
Joined 10-01-2011 status: Regular |
I eliminated a bad case of earth-loop-grunge by carefully stripping back the screen of a short USB cable and cutting the black wire. |
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
@ thrash 3 - which end did u cut it at? (ie was it at the PC end or the MOX end? and did u reconnect the shield? or leave it stripped back?) thanks |
ThrashJazzAssassin
Total Posts: 42
Joined 10-01-2011 status: Regular |
I cut it in the middle and didn’t reconnect the shield |
rogs
Total Posts: 121
Joined 07-16-2011 status: Pro |
There are 5 independent connections on the USB cable, supplied with the MOX.
Connecting the cable to the MOX, at the MOX end, (without the unit switched on) shows that pin 4 and the screen are linked within the MOX, and this point is also common to the gnd (screen) side of the unbalanced output and headphone jacks.
Now because the USB data link is differential,I can’t see there is any need for the ground (pin4) specifically to be used for data transmission, although the two differential data thresholds are referenced to a common ‘gnd.’
What may be worth trying is to disconnect pins 1 and 4 of the USB by cutting the cable connections at one end (either end). When it comes to the screen, it’s probably best to retain the screen connection at the PC end, and cut it away from the metal ‘surround’ at the MOX end. That would give you a differential USB data connection, which was still screened, but importantly not now grounded at the MOX end. Loop broken. Noise gone. BUT, if the 5V part of the USB is used by the MOX, then this would mean the DC ‘return’ for that 5VDC would be via an indeterminate path, through other equipment. Can’t see why the 5V DC part of the USB is required, in this instance. After all, the MOX works fine on it’s own! Might be worth hacking into an old USB cable to try. (Probably not the quality one that came with the MOX)!!
Basically, leave pins 2 and 3 connected both ends, and the screen at the PC end only.
As long as the USB 5V DC is not needed, then you will have found an easy way to break the earth loop.
EDIT: Reading a bit more about USB leads me to believe that the 5V and gnd connections ARE probably required. The differential data needs to be pulled up by the MOX, to overcome the ‘pull down’resistors of the host port. So even if it doesn’t actually use the 5V from the PC (which it probably does?) there still needs to be a DC (gnd) reference between the MOX and PC.
So the USB data, although differential, does not truly ‘float’, and the above idea will probably not work. At least not reliably. Back to the drawing board, I suspect. Sorry about that. Unless someone knows different......??? |
Anton
Total Posts: 169
Joined 08-26-2002 status: Pro |
I had the same noise problems. I disconnected the groundwire of the powersupply from my laptop. It cured the problem but this not a very good idea because it can be very dangerous. |
rogs
Total Posts: 121
Joined 07-16-2011 status: Pro |
Yes - as you say, removing the safety earth from the computer to ‘break’ the loop means that any computer power supply failure might require an earth fault current to flow, through the MOX PCB tracks, before grounding through the ‘surviving’ mixer, preamp, or powered speaker, safety earth.
My suggestion above of breaking the loop at the USB connection might work - and would not compromise safety.
But at least not safety!! |
bassman69
Total Posts: 102
Joined 11-26-2011 status: Pro |
Hi all, I did exactly what thrash3 said (cut the ground and shield) and found it to get rid of 99% of everything regarding noise. There is still a minor high pitched noise if i raise the volume up to maximum (while reducing the gain on the preamp/using the DI without a guitar connected, ie lowering the noise floor so i can hear the high pitched noise without it being drowned out by white noise) however ALL the transient noises are gone (the crackling, the noise when you move the mouse etc). If i now go and plug the keyboard (with the new usb cable) into the UNbalanced input on the sub (leaving out the mixer or the preamp) i get a SMALL bit of noise, akin to the noise i hear when i raise the gain up. if i plug into the BALanced inputs of the sub, using a cable i modified that goes from TS --TRS (where the TS side is connected to the TR side, leaving out the S) the noise completely dissappears. Does anyone have any suggestions for how i could use the unbalanced sub inputs whilst not getting the noise? (i would like to use the unbalanced inputs of the sub for the MOX, and the balanced inputs for an audio interface so i can run them to the speakers in parallel, without requiring a mixer to combine the signal, and introduce noise etc). (NOTE: previously, with the USB cable unmodified, the modified TRS/TS cables had no effect whatsoever in any configuration. this leads me to believe now that the main USB problem has been eradicated, i can use other devices (eg the DTI box and the aforemeantioned cables) to further eliminate the issues. and yes, NEVER cut ground pins. |
rogs
Total Posts: 121
Joined 07-16-2011 status: Pro |
So, good that you have confirmed that the USB data connection doesn’t mind it’s ground reference via this ‘new’ audio ground route. (It’s lost the USB ‘GND’ reference, so it has to find another!)
Looks as if your use of tip and ring, but not sleeve, has the effect of providing the ‘unbalanced’ MOX output with a return, via the negative ‘balanced’ (ring) connections, and not directly to GND.
Slowly getting better all the time, by the sound of things. You might need to need to confirm that the ‘new’ USB GND reference route doesn’t corrupt the data at all?..... |