drtechno
Well-known member
most have and Its more of a common practice to change that cord out and ground the chassis with the outlet ground.
Proper wiring to avoid ground loops
- Thread starter mrtnasty
- Start date
Help Support GroupDIY Audio Forum:
In my house the outlet grounds are floating. However plugging 3 wire line cords into the same outlet strips will bond all the chassis grounds together. For better and worse.
====
Here is a "for worse" anecdote. Back years ago before I added redneck DIY ground wires to my kitchen GFCI outlet (bonded to my fuse panel). I noticed that I was getting a small tingle (electrical shock) from my KitchenAid mixer. I tore that mixer apart looking for the source of electrical current leakage with no success. I finally traced the current leakage as coming from a leaky over-voltage protection component inside one of the several cheap outlet strips. The leakage current was small enough that dumping it into a proper safety ground would not even be noticed but I could feel it while cooking barefoot in my kitchen. The current leakage was conducted into the floating chassis grounds of my sundry kitchen appliances. When I clipped out the faulty protection devices the tingling stopped.
JR
mrtnasty
Well-known member
Hi all, thank you all very much for the replies. Turns out the center tap of the input transformer on my preamp was not grounded. The last line in the "Audio input connections" section of the RCA BA-2C manual says "Whenever a balanced source is used, remove the ground lead from the input transformer primary (terminal 1), and ground terminal 7." After grounding terminal 7, the buzz went away.
Which simply means that it was unbalanced. Now, with pin 7 grounded, it is balanced, but not floating. It would be interesting to see what happens with pin 7 not connected. It should be even better.
terry setter
Well-known member
- Joined
- Nov 17, 2013
- Messages
- 177
I remember that I did the test without interconnects. I also remember being confounded by the fact that - in the scientific portion of my brain - I thought should probably have gotten the highest voltage between Piece A and Piece B, then added and tested Piece C, then REVERSED Piece B and re-tested Piece C, etc., etc. for all the components that would end up being interconnected. (Exponentially expanding numbers of plug patterns...). Fortunately, the difference in voltage between positions of the plugs I tried was very, very small, so I put thins kind of thinking out of my mind for the next 60+ years. No one I knew at the time was using isolation transformers in their home systems, but I think that those and grounded, polarized plugs made such grounding experiments unnecessary in almost all circumstances.
CCAUDLE: I remember that the idea was put forth that if the voltage between the chasis was higher, there was better "isolation", and by inference, less possibility to cause noise via interaction. Sorry I don't have the article anymore and am not even sure which of the several HiFi magazines I occasionally read published it.
moamps
Well-known member
Thank you for the clarification.
Yes, I was thinking along those lines too. Considering that it is about very small values of leakage currents (small values of parasitic capacitors if Y capacitors are not used) and consequently high impedances, higher voltage could mean higher impedances. It probably should have been measured with a high-impedance voltmeter (DVM) and the readings probably fluctuated constantly so the reading is quite inconsistent.
Of course, that measurement makes sense only for devices where double insulation is applied and the "audio ground" is connected to the chassis with audio cinch connectors.
This is an interesting post about a problem that seems to haunt all of us from time to time regardless of our expertise, knowledge or experience.
Regarding the earlier variations on whether to connect the shield at the source or the output of a balanced cable or indeed both ends, I've found there is a often a need to mix the scenarios depending on the equipment, the combinations thereof and the environment.
When running a live production business in the 80s we always connected the shield on every balanced cable at both ends no matter what. It seemed bullet proof in live venues where there were many very long cable runs followed by lots of gain to the speaker stacks and there was much EMF due to the proximity of dimmer racks in large lighting rigs.
Once moving into studio land I realised the earthing had become more complex with regard to the noise floor in general, both AC frequencies and RFI.
At first I couldn't fathom why the scenarios would be so different apart from the studio requiring much lower gain paths vs much more sensitive reproduction equipment and listening spaces.
In a lot of cases the only fundamental difference I could find was that most of the time studio systems, unlike live systems, employed studio consoles, patchbays, multitrack tape machines and DAW hardware interfaces that then presented many parallel paths between a small number of chassis
This stands a bit apart from PA systems which often represent a more serial kind of arrangement with a bunch of mic lines arriving at the console then two lines back to the crossover, six lines from the crossover to the power amps and that was about it apart from the mic split to the monitor console.
So, even though I've studied electronics and was given a keen passion from the outset to always think of where the electrons are wanting to go as they try to satisfy their desire to reach earth, my decades of real world experience have left me checking earthing in stages and blocks and accepting that each unique situation seems to so often require it's own unique combination of shielding arrangements.... kind of a 'no right' and 'no wrong' first world scenario.
Regarding the earlier variations on whether to connect the shield at the source or the output of a balanced cable or indeed both ends, I've found there is a often a need to mix the scenarios depending on the equipment, the combinations thereof and the environment.
When running a live production business in the 80s we always connected the shield on every balanced cable at both ends no matter what. It seemed bullet proof in live venues where there were many very long cable runs followed by lots of gain to the speaker stacks and there was much EMF due to the proximity of dimmer racks in large lighting rigs.
Once moving into studio land I realised the earthing had become more complex with regard to the noise floor in general, both AC frequencies and RFI.
At first I couldn't fathom why the scenarios would be so different apart from the studio requiring much lower gain paths vs much more sensitive reproduction equipment and listening spaces.
In a lot of cases the only fundamental difference I could find was that most of the time studio systems, unlike live systems, employed studio consoles, patchbays, multitrack tape machines and DAW hardware interfaces that then presented many parallel paths between a small number of chassis
This stands a bit apart from PA systems which often represent a more serial kind of arrangement with a bunch of mic lines arriving at the console then two lines back to the crossover, six lines from the crossover to the power amps and that was about it apart from the mic split to the monitor console.
So, even though I've studied electronics and was given a keen passion from the outset to always think of where the electrons are wanting to go as they try to satisfy their desire to reach earth, my decades of real world experience have left me checking earthing in stages and blocks and accepting that each unique situation seems to so often require it's own unique combination of shielding arrangements.... kind of a 'no right' and 'no wrong' first world scenario.
mrtnasty
Well-known member
You're saying disconnecting 7 and grounding 1 would make it even quieter? Floating is better than balanced?
drtechno
Well-known member
They don't understand that transformer is not balanced wound. You put the ground lead to one side of the primary for unbalanced connection, the center tap (pin 7) to force balanced operation. If you are using this for a mic pre with standard dynamics you ground the center tap.
There are some input transformers you do this center tap thing. But not often because winding the primary went to a better convention that you didn't have to do that. Pin 7 should be either gnd or a phantom power injection point with these particular type of input transformers.
I'm saying disconnecting pin 7 and using pin 1 as "cold" in a balanced connection.
If the body of the mic is in continuity to earth via Pin 1, it should work.
mrtnasty
Well-known member
Thanks, yea I can see how that would work, but the cable housing (jones connector for the ba2c I built for power, input and output) is also not grounded to the chassis. Before I grounded pin 7, when I put a jumper between the chassis and the cable housing, the buzz was reduced, but now with pin 7 grounded everything seems dead quiet with or without the jumper.
This is a bit confusing. I looked through the early posts and did not see a wiring diagram or a model number referenced until the mention of RCA on the 3 Jan post. These pin numbers are obviously not referring to standard XLR-3 pin numbers.
Is this the correct diagram for the numbers being referenced?
https://www.coutant.org/ba2c/index.html
mrtnasty
Well-known member
Yes, sorry, we were talking about terminal (pin) 7 on the input transformer of the BA-2C, it is the center tap on the primary side.
Is the cable shielded? Where does the shield connect? The wiring diagram I found does not indicate that the connector on the back has a terminal dedicated for chassis.
mrtnasty
Well-known member
I just followed the diagram in the manual, and grounded both xlr shields and the ground wire from the power cable to the pins that connect to the chassis (the PIN numbers are off on this particular jones connector, they are all shifted by 2 from the BA-2C manual)
Is this one of the type of tube designs which uses chassis as part of the circuit reference (aka circuit ground) instead of having separate wiring?
I see pin 10 on the RCA diagram shows as connected to the circuit, it did not jump out to me right away that pin 10 was also the chassis connection.
It is also interesting that the manual has much more discussion of unbalanced connections than balanced; were unbalanced connections so common in the 1940 and 1950 eras? I would have thought that with transformer coupling being so common that proper balanced transformer connections would be the default.
Floating means no connection to the local circuit. The copy of the RCA manual I found has in the input wiring section a note that "Whenever a balanced source is used, remove the ground lead from the input transformer primary (terminal 1 or 2)." No mention of pin 7, so connecting pin 7 of the transformer does not seem to be a common or recommended connection.
With a transformer, "balanced" and "floating" are often the same, the input winding connects with only two connections to the source device, which may also be a transformer winding on an output transformer on the source device.
That may not apply if phantom power is in use, since that could involve a DC connection to the center tap from the phantom power supply (with later designs which supported phantom power). Not always, you could still capacitor couple into the transformer and use two resistors to the phantom power supply like is done on solid state preamp designs.
This design pre-dates phantom power since microphones would all have been either passive or tube in that time frame, so I would not count on the transformer being wound in a way that it could tolerate DC current through the windings, so resistors and coupling capacitors would probably be the recommended method unless you verified that the windings were such that the DC current through each half of the winding canceled and did not push the transformer into saturation.
I must say we would have not wasted so much time over this, had we not taken for granted that a mic pre with a transformer would be, by default balanced.
Kinda weird.
Kinda weird.
drtechno
Well-known member
Who brought that up then? Because there is only one reason to ground a parimary center tap which is if the input transformer you are using does not have a balance wound primary. A lot of people have changed out those transformers but that was western electric mic preamps that didn't have balanced input windings and not RCA. RCA always used balanced wound input transformers.
drtechno
Well-known member
If grounding pin 7 with a source connected reduces noise, then the device you are connecting to has issues going into balanced mode and its impedance is mismatched. If this because you haven't connected anything, that would be normal as it would be an antenna.
