Full Wave Center Tapped Ground Question

[dbishopbliss]

I suppose this may be fundamental, but I am never quite sure of what the answer is...  all transformers I have used for non-kit builds have required bridge rectifiers.

When I am implementing full wave rectification with a center tapped transformer (see diagram), do I connect all ground wires to 'B'.  By 'all' I mean:

• grounded ends of filter caps
• ground wires of input and output jacks
• potentiometer grounds
• resistor and capacitors in a grounded-cathode amplfier

I'm sure there are others as well, but you get the idea.  

Now for the bigger question... do I connect all of these grounds (including B) to the ground wire that is connected to the chassis which is connected to the third wire of the power cord?

 

[Paul Joppa]

"B" is the negative terminal of the power supply, and it is usually connected to signal ground. Not always; for example the SEX amp creates a ground some 20 voltage above the negative terminal, so that the negative terminal can be used for biasing the power tube section.

The recommended procedure is to connect the power-line safety ground to the chassis, and connect the various transformer cores to chassis ground nearby. The signal (and power supply) ground is connected to the chassis, usually at a distant location so that any power ground currents running through the chassis plate are well away. There are other subtleties but you need a book, not a forum post, to get into those.  :^)

 

[dbishopbliss]

When you say, "... and connect the various transformer cores to chassis ground nearby." are you referring to the wire that goes from the lug connected to the transformer to the chassis ground? 

My latest project is on a wood chassis, so there is no way to connect the signal ground to the chassis at a "distant location".  I was planning on having a star ground for the "negative terminal" of the power supply, signal grounds, etc., then have a single wire to the chassis ground.  Does this seem reasonable?

 

[JC]

OK, now I'm curious:  If your chassis is wooden, what are you intending to connect this single wire to?  Put another way, if your chassis is wooden, what are you referring to as "chassis ground"?

 

[dbishopbliss]

The chassis ground will be a bolt threaded through the wood with a solder lug on it.  A wire will go from the transformer to the solder lug, another wire will go from the ground of the power receptacle to lug, and another wire to the signal ground.  Something like the attached diagram.

 

[JC]

OK, for the sake of discussion, then, perhaps we could call that one the "Safety Ground".  If the transformer is sitting on top of the wooden chassis, then, as you have indicated, it should certainly have a good connection to the Safety Ground.

Will there be any other metal parts that may be touched on the outside of the chassis?  It occurs to me that any of those should also have a good electrical connection to "Safety Ground" as well.  I'm bringing those up now in case one of them might also happen to be an audio or circuit ground as well, just by its nature.

For instance, exposed RCA connectors should probably be connected to "Safety Ground", and they would most likely be at Audio Ground as well.  So, a little strategy might be in order.

 

[dbishopbliss]

The only metal parts that might be touched will be the input and output jacks and the potentiometer.  I'm planning on running a wire from each of those to a star-ground that I have been calling "signal ground".  I was also planning on connecting the negative terminal of the power supply to the signal ground as well.

 

[JC]

Well, it sounds like you've got a good handle on it.  If it were me, I think I would start by connecting Safety Ground And Power Supply Negative to Audio Ground at the input connectors only.  I probably would also try to devise a way to leave my options open so I could try Safety and PS '-' connected to Audio Ground at other points if I wanted to experiment.

Now, I do not know the precise letter of the "code" on Safety Ground, but I would be inclined to say that whatever is Safety Grounding exposed metal parts should probably be of a reasonably heavy gauge, at least relative to what you might use for "Audio Ground".  So, if an exposed metal part is getting its Safety Ground by way of the Audio Ground system, you may want to consider making its Audio Ground connection a little beefier than you otherwise might.

Boy, life was sure simpler when you could just use the metal chassis for all your grounding needs!  Of course, life was also a bit more risky and prone to noises then, too!

Sorry if I dragged you off onto side issues after Paul already answered your question!

 

[dbishopbliss]

Sorry if I dragged you off onto side issues after Paul already answered your question!

No dragging involved.  The only way I really grasp this stuff is to discuss it.  I've always been confused by the different "ground" terms - never sure if they should be connected or not.  Sounds like, ultimately, they are all connected.  However, there are different techniques for how and where they are connected.

 

[Grainger49]

You are right.  I have seen preamps that have signal common that is not connected to the chassis/safety ground (the safety ground should always be tied to the chassis for... safety sake).  I think in most everything I have built and modified the two have been tied together somewhere.  IIRC, the FP 2 has the signal common on T14 and chassis ground on T13.  Bonding the two is a technique that dropped noise in some, many, systems but caused noise in some.  It is associated with the other components and their grounding.

 

[Paul Joppa]

The best reference I know is Ralph Morrison's "Grounding and Shielding Techniques in Instrumentation". It begins with the following quote:

"1 mA flowing in 1 ohm results in a voltage drop of 1 mV"
-Georg Simon Ohm

 

[JC]

And, Ohm could have added "1 mV is a LOT, in an amplifier", if there were amplifiers!

I always make an attempt to be pedantic about naming which ground I am discussing in any given context, and, hopefully, making those names descriptive of the function of the ground at issue.  It is the only way I can think of to keep confusion at bay.

The addition of the third-wire Safety Ground certainly makes the issues more complex, but I can't imagine how we managed to survive as long as we did without it.  A guitar-player buddy of mine from back in the day of no safety grounds received a very instructive lesson one evening on stage:  He was playing his electric guitar plugged into his tube guitar amp, a mid-sixties Fender which of course had no third prong on the AC plug.  It relied on a switch to assign one side of the incoming AC line to chassis through a capacitor.  Thousands of amps were made this way.  The idea was to set the switch for the lesser hum, which usually meant that AC neutral was connected to the chassis through the cap.

Now, the chassis was the all-purpose "ground", of course. EVERYTHING "grounded" to it!  Need a "ground"?  Just solder to the chassis.  So, ultimately, his guitar was "grounded" to the amp through the shield on the unbalanced guitar cord, which meant that the guitar bridge, and therefor the strings, were also "grounded" to the guitar amp.  This also meant that the guitar was assigned to one side of the AC line, by the "grounding" switch on the amp.

All of which posed no particular problem as long as the guitar and amp and player were a little island unto themselves standing on stage.

Then, he stepped up to the mic to sing.  He remembered a blue flash and running into the kick drum, but not a lot else.  Seems the "grounding" switch on the PA amplifier was set to the opposite position of the one on his guitar amp.  Which also would have been fine, if he just hadn't got his lips so close to that "grounded" microphone! 

 

[Doc B]

Presumably the review of the show next day was of "an electrifying performance".

 

[Beefy]

How do ground loop breakers figure in here?

With my PCB solid state gear, I am used to connecting safety/chassis ground to signal ground through a 10ohm, 5W resistor paralleled with a 0.1

 

[Grainger49]

Beefy,

It is a system dependent problem.  I have found that the 10 ohm grounding resistor on Dynaco ST-70s causes noise with Foreplay IIIs and not with FP 2s.  Odd!  The FP III switches the input hot and grounds leaving unused input grounds open (this is most effective to lower noise).  The FP 2 ties all input grounds together.  

And maybe the above factoid about the input grounds is irrelevant.  I don't know.  But I do know that the FP III and ST-70 is a hum prone combination with the 10 ohm resistors there.

 

[dbishopbliss]

All of which posed no particular problem as long as the guitar and amp and player were a little island unto themselves standing on stage.

One time a band mate handed me his guitar to hold while he ran to the bathroom real quick.  He was a little pissed when his SG went flying across the stage.  After that, I learned to tap the mic with one hand while holding the guitar strings with the other.  Then, tap the guitars of other band members in a similar manner.  If you felt a shock, you would switch the ground.

 

[JC]

Well, Doc, I don't think there was a review, but I would imagine there was a considerable amount of medicinal Jack Daniels applied to the injury!

You may recall that the chassis on that vintage of fender was hung into the cabinet with long screws that  went through two chromed steel "straps" on the top of the box which acted as big "washers".  These were also "grounded" by their attachment to the chassis, of course.  Bad idea to be leaning on two different amps at the same time, in other words!

Beefy, I've used a similar approach in the past, even on a Foreplay 2, although I usually find the cap to be unnecessary.  Of course, that is the ONLY point at which the Chassis/Safety ground can be connected to the Audio ground.  So, jacks must be of the insulated type, of course, for instance.

 

[Paul Joppa]

How do ground loop breakers figure in here?

With my PCB solid state gear, I am used to connecting safety/chassis ground to signal ground through a 10ohm, 5W resistor paralleled with a 0.1

 

[dbishopbliss]

So that explains why my 101dB horns hum when hooked up the ice dispenser of my refrigerator. 

 

[Beefy]

Thankyou for the replies, gentlemen!

I have indeed read many different opinions for this, but it is good to know from you all that - at a very minimum - I would not blow anything up if I grounded a tube amp the same way as I do for my SS gear.