Negative Voltage

[hevans1944]

So I guess at this point my brain is still looking for clarity on how 0V chassis ground of the amplifier is defined. According to the battery analogy, if we make some reference to less than 0V, we must have arbitrarily defined 0V by fixing it to some concrete potential.

Absolutely correct! It looks like you understand potential difference after all, positive as well as negative. For the Fender amp, the center-tap on the secondary of the power transformer is the 0V reference for the rectified and filtered DC. You should note on the schematic that this is connected to "chassis ground" so every voltage produced by the power supply... the full-wave rectified B+ or plate supply voltage as well as the half-wave rectified negative bias voltage... is measured with respect to the center tap on the secondary of the power transformer. The center-tap is electrically connected to the chassis for several reasons that I won't describe here, but the most common reason is it simplifies the wiring to do so.

Note the 6.3 VAC filament winding is also center-tapped, and the center-tap is connected to chassis ground. This is a hum-balancing connection, left over from the days of directly-heated filament cathodes. Not so important with vacuum tubes using indirectly heated cathodes, which all the tubes in this amp are, but it does no harm and may do some good at no extra cost. If you really want to reduce "hum" you can operate the tube filaments from rectified and filtered DC, which IIRC some "high end" amps did. Probably need a pair of Golden Ears to hear the difference.

 

[(*steve*)]

And if ground is 0V then I'm back to square one wondering how on earth (no pun intended) a voltage with respect to ground can be negative

Imagine you are standing at ground level in front of a building that has 11 storeys above ground and three storeys of parking below it.

This building is not in the US, so you expect to enter at the ground floor and go up one level to the first floor, another level to the second floor, and so on. So ground is 0, fist floor is +1, second floor is +2, and so on. The floors below ground are labelled B12, B2 and B3. We can think of them as -1, -2, and -3.

Given that we are on ground level (0), the floor number tells us how far up or down we need to go to get there.

However, when we walk in the door, we find that we are on the second floor! (Perhaps the building is built into a hill and some other entrance opens to the ground floor?)

Here we have an example of the local ground being different from the building ground. From our outside perspective, the building ground is 2 floors below is. To us it is B2, or -2.

Similarly, if we are on the top floor, that floor is zero, and the floor labeled as B1 is 11 floors below us.

If we were to use a relative standard when arranging a meeting, we may never agree what floor it's on. Thankfully the building management have labeled all of the floors for us, so we can agree (by looking at a standard set of floor numbers) how far up or down we need to go.

All this is fine if we're inside the building, but it helps less if we are outside it (we are trying to find a fault in the floor levels in the building).

Let's say we know that a particular room is supposed to be on the 7 to floor. If a person is waving a red flag through the window, we can count windows up to that floor. However, we already know that if we are standing outside the building, out reference for ground might not be the same as that used by the building. We might come up with 5, 6, 7, or maybe even 8 (if we're standing in a driveway directly entering the basement level 1). What we need is another person washing a black flag from a known ground floor window. With that, no matter where we stand, we can measure the floor number as determined by the building's designer. If we now find that the red flag is ok the 6 to floor, we know that the building has a fault; the room is not at the level we are told it is.

So the ground reference for a circuit (typically synonymous with 0V) may not be the same as some other ground reference. It is a reference used by the designer to which other voltages are measured. From this point, negative voltages are at points that are simply less positive (or more negative) than the reference.

As an interesting example, I have a circuit which uses both thermionic valves (tubes) and semiconductor op-amps. When looking at the op-amp part of the circuit it's convenient to see the power supply as +/- 15V, but when looking at the valve part, it's more convenient to consider one end of the directly heated cathode to be ground, and the HT to be +30V. In this case, the reference point in the circuit differs depending on what voltage you're measuring. Some points in the circuit are labelled with two voltages because one measurement makes more sense if you're looking toward the tube or away from it.

An example of this is the anode voltage. If I am measuring it to set the DC operating point of the valve, it is 15V with respect to the cathode. However, this point also feeds into an op-amp, and in this case the voltage is close to 0V with respect to the op-amp's 0V supply rail.

 

[PeteVee]

Imagine you are standing at ground level in front of a building that has 11 storeys above ground and three storeys of parking below it.

This building is not in the US, so you expect to enter at the ground floor and go up one level to the first floor, another level to the second floor, and so on. So ground is 0, fist floor is +1, second floor is +2, and so on. The floors below ground are labelled B12, B2 and B3. We can think of them as -1, -2, and -3.

Given that we are on ground level (0), the floor number tells us how far up or down we need to go to get there.

However, when we walk in the door, we find that we are on the second floor! (Perhaps the building is built into a hill and some other entrance opens to the ground floor?)

Here we have an example of the local ground being different from the building ground. From our outside perspective, the building ground is 2 floors below is. To us it is B2, or -2.

Similarly, if we are on the top floor, that floor is zero, and the floor labeled as B1 is 11 floors below us.

If we were to use a relative standard when arranging a meeting, we may never agree what floor it's on. Thankfully the building management have labeled all of the floors for us, so we can agree (by looking at a standard set of floor numbers) how far up or down we need to go.

All this is fine if we're inside the building, but it helps less if we are outside it (we are trying to find a fault in the floor levels in the building).

Let's say we know that a particular room is supposed to be on the 7 to floor. If a person is waving a red flag through the window, we can count windows up to that floor. However, we already know that if we are standing outside the building, out reference for ground might not be the same as that used by the building. We might come up with 5, 6, 7, or maybe even 8 (if we're standing in a driveway directly entering the basement level 1). What we need is another person washing a black flag from a known ground floor window. With that, no matter where we stand, we can measure the floor number as determined by the building's designer. If we now find that the red flag is ok the 6 to floor, we know that the building has a fault; the room is not at the level we are told it is.

So the ground reference for a circuit (typically synonymous with 0V) may not be the same as some other ground reference. It is a reference used by the designer to which other voltages are measured. From this point, negative voltages are at points that are simply less positive (or more negative) than the reference.

As an interesting example, I have a circuit which uses both thermionic valves (tubes) and semiconductor op-amps. When looking at the op-amp part of the circuit it's convenient to see the power supply as +/- 15V, but when looking at the valve part, it's more convenient to consider one end of the directly heated cathode to be ground, and the HT to be +30V. In this case, the reference point in the circuit differs depending on what voltage you're measuring. Some points in the circuit are labelled with two voltages because one measurement makes more sense if you're looking toward the tube or away from it.

An example of this is the anode voltage. If I am measuring it to set the DC operating point of the valve, it is 15V with respect to the cathode. However, this point also feeds into an op-amp, and in this case the voltage is close to 0V with respect to the op-amp's 0V supply rail.

Thanks Steve, but I think we've moved past that and established that I get that part of it. It seems my confusion arose from thinking that 0V as defined on the amp schematic was truly 0V, an absence of electrical potential as opposed to 0V as actually a reference measured against a concrete electrical potential.

 

[PeteVee]

Absolutely correct! It looks like you understand potential difference after all, positive as well as negative. For the Fender amp, the center-tap on the secondary of the power transformer is the 0V reference for the rectified and filtered DC. You should note on the schematic that this is connected to "chassis ground" so every voltage produced by the power supply... the full-wave rectified B+ or plate supply voltage as well as the half-wave rectified negative bias voltage... is measured with respect to the center tap on the secondary of the power transformer. The center-tap is electrically connected to the chassis for several reasons that I won't describe here, but the most common reason is it simplifies the wiring to do so.

Note the 6.3 VAC filament winding is also center-tapped, and the center-tap is connected to chassis ground. This is a hum-balancing connection, left over from the days of directly-heated filament cathodes. Not so important with vacuum tubes using indirectly heated cathodes, which all the tubes in this amp are, but it does no harm and may do some good at no extra cost. If you really want to reduce "hum" you can operate the tube filaments from rectified and filtered DC, which IIRC some "high end" amps did. Probably need a pair of Golden Ears to hear the difference.

Thank you. Detailed explanations like this shed a lot of light on the subject for me and show me where I was making incorrect assumptions.

 

[(*steve*)]

as opposed to 0V as actually a reference measured against a concrete electrical potential

It may be 0V as measured against some reference point, or it may be the reference point itself.

I'm not really enthused about the word "concrete". If you mean stable, or well known, or something like that, then I'd be happier.

0V on a supply rail suggests it is the voltage reference. 0V elsewhere may mean something like a virtual ground (i.e. it has ground potential but is not connected to ground). And here I am using ground as a synonym for the reference voltage.

 

[PeteVee]

It may be 0V as measured against some reference point, or it may be the reference point itself.

I'm not really enthused about the word "concrete". If you mean stable, or well known, or something like that, then I'd be happier.

0V on a supply rail suggests it is the voltage reference. 0V elsewhere may mean something like a virtual ground (i.e. it has ground potential but is not connected to ground). And here I am using ground as a synonym for the reference voltage.

Yeah, my grasp on the correct terminology needs improvement. What if I had said "an absence of electrical potential as opposed to 0V as a reference measured against a quantifiable electrical potential."

 

[davenn]

Yeah, my grasp on the correct terminology needs improvement. What if I had said "an absence of electrical potential as opposed to 0V as a reference measured against a quantifiable electrical potential."

several posts you have made even since the detailed post I made with the diagrams, still show you don't understand
the term potential (potential difference)

here it is again, .... you CANNOT say ( as you have been) that a particular point has an electric potential.
( full stop .. end the statement there)
it MUST be relative ( referenced) to somewhere else, else it doesn't make sense as Steve re-commented on a couple of posts ago

going back to my single 9V battery, it has a potential difference of 9V between the terminals.
Whether it is a positive or negative potential depends on which terminal of the battery you state is the 0V reference point as I had shown. If you make the + terminal the 0V reference, then the - terminal will be measured at a -9V potential difference relative to the + terminal 0V reference point. or conversely the + (0V) terminal is also at a + 9V potential referenced to the - terminal

I took all that further with the multiple batteries as a practical example and Steve used the floors (levels) as an analogy of my multiple batteries.



Dave

 

[PeteVee]

several posts you have made even since the detailed post I made with the diagrams, still show you don't understand
the term potential (potential difference)

here it is again, .... you CANNOT say ( as you have been) that a particular point has an electric potential.
( full stop .. end the statement there)
it MUST be relative ( referenced) to somewhere else, else it doesn't make sense as Steve re-commented on a couple of posts ago

going back to my single 9V battery, it has a potential difference of 9V between the terminals.
Whether it is a positive or negative potential depends on which terminal of the battery you state is the 0V reference point as I had shown. If you make the + terminal the 0V reference, then the - terminal will be measured at a -9V potential difference relative to the + terminal 0V reference point. or conversely the + (0V) terminal is also at a + 9V potential referenced to the - terminal

I took all that further with the multiple batteries as a practical example and Steve used the floors (levels) as an analogy of my multiple batteries.



Dave

OK, Dave. The clarification is appreciated, as always.

 

[duke37]

A valve amplifier will need the grid negative to the cathode which is used as the reference to determine the valve current.
Radio output valves usually have a resistor in the cathode to ground connection to raise the cathode above the grid which is referenced to ground. This has the disadvantage that a capacitor is needed across the resistor and a subsantial part of the HT energy is lost.
High power amplifiers have the cathodes connected to ground and a separate low power negative supply is used to supply the grids. All HT power goes to the valve and the grid voltage can be made stable.

 

[hevans1944]

Radio output valves usually have a resistor in the cathode to ground connection to raise the cathode above the grid which is referenced to ground. This has the disadvantage that a capacitor is needed across the resistor and a subsantial part of the HT energy is lost.
High power amplifiers have the cathodes connected to ground and a separate low power negative supply is used to supply the grids. All HT power goes to the valve and the grid voltage can be made stable.

Yes, I remember from my Novice year as a newly-minted amateur radio operator, building a crystal-controlled oscillator driving the ubiquitous RCA 6146B beam power pentode, that I absolutely needed to connect the cathode to chassis ground (which was, hopefully, also RF ground). I was running the power tube biased for Class C operation, meaning the grid was biased way far into cut-off, except at the positive peaks of the RF oscillator signal, which was turned on and off by my telegraph key to transmit Morse Code in the Novice part of the 80m amateur radio band.

Clearly a cathode bias resistor, by-passed or not with a capacitor, would never work for Class C. So I had to jinn up a separate negative bias supply for the grid, and couple the RF oscillator signal to it with a capacitor. The plate circuit of the 6146B was fed about +400 V through an RF choke and coupled with a mica capacitor to a pi-network. At the time I was pretty much over my head in calculating the component values required to resonate the plate circuit for Class C operation and match (using the pi-network) the plate impedance to my 50 Ω coaxial cable feed going up three floors (inside a rain drain pipe) to attach directly to a half-wave 80m dipole on the roof. No balun (whut was that?), so the shield on the coax probably radiated as much as the dipole... but, what the hey, it tuned to resonance and loaded up just fine. I made lots of "contacts" during my Novice year, even exchanging QSL cards with some of them.

So I guess I lucked out on choosing the two variable capacitors and air-core inductor for the pi-network. There were a lot of examples in the ARRL Radio Amateurs Handbook, 1965 edition, to choose from. I also got the idea (and a circuit) to implement full-break-in keying or QSK for my transmitter from the ARRL Handbook too. It seemed like a real cool idea to be able to receive and hear other radios in between dits and dahs of my own transmissions. Back in those days, almost all communications were half-duplex, only one transmitter being active (keyed) at any given instant. Usually this was performed by a toggle switch that alternated the antenna connection between a separate receiver and the transmitter. I suppose if you had a fast enough relay, operated by your telegraph key, you could use that to switch back and forth between transmit and receive, but most Hams just flipped a toggle switch at the beginning and end of each transmission to alternate between transmit and receive.

So this newbie built a nifty circuit that muted the receiver when the telegraph key was down and the transmitter was active, and allowed reception of the other Ham's signal when the telegraph key was up and the transmitter was NOT pumping out maybe fifty watts of RF. And it did this with no moving parts, as fast as I could operate the telegraph key... perhaps sixty words per minute, although that was faster than my ears could decode Morse. I was extremely disappointed after I got this rig on the air to discover not a single Ham that I communicated with had ever heard of QSK operation. They all began and ended their transmissions by flipping a toggle switch!

 

[BobK]

we must have arbitrarily defined 0V by fixing it to some concrete potential.

Yes, that is the concept. I think you have been confusing potential and voltage. A voltage is always the difference between two potentials. A single potential can be arbitrarily given any number you want. But the difference between any two potentials remains the same whatever number you pick for one of the potentials.

Bob

 

[PeteVee]

Yes, that is the concept. I think you have been confusing potential and voltage.

Absolutely, you're right to surmise that I don't have it 100% straightened out as evidenced by misusing some of the terminology (among other things).

I think you have been confusing potential and voltage. A voltage is always the difference between two potentials.

So a potential is equivalent to a charge (i.e. an imbalance of electrons) and a potential difference (voltage) is the difference in charge between two points, correct?

A single potential can be arbitrarily given any number you want. But the difference between any two potentials remains the same whatever number you pick for one of the potentials.

Sure, that is super straightforward. I can think of analogies. Music theory for example. You can play any composition with any one of the twelve tones as the key center by shifting the pitch up or down as long as you maintain the mathematical relationships of the intervals.

Up until very recently I've been hanging on pretty tightly to the idea that voltage simply referred to the amount of work a given amount of current/electrons could do as dictated by the amount of resistance in the circuit. So I guess I sort of thought of voltage as a modified or qualified expression of current. But I'm starting to see that maybe that doesn't take into account the concept of voltage being the difference in charge between two points in a circuit. Maybe you can clarify for me. Is one or both of my conceptualizations lacking or are they two different ways of looking at the exact same thing?

 

[BobK]

You are getting there. Most of what you said in the last post is true, but...

No, potential is not equivalent to charge and voltage is not equivalent to a difference in charge.

Here is an analogy I like to use for voltage and charge.

Think of charge as analogous to mass.
Think of the earth's gravitational field as analogous to the electric field.

Now, voltage is equivalent to height!

Given a mass, raising it's height raises its potential energy. It gains the same amount of potential energy when raised from 0 feet to 3 feet or from 10 feet to 13 feet or from -1 feet (in a hole) to 2 feet. You must do work on the mass to raise it in height. Lowering the height of a mass lowers it's potential energy and it does work as it falls (think of an old pendulum clock powered by weights)

Charge behaves exactly the same in an electric field. It takes work to raise it it's potential. It does work when you lower it's potential. The only difference is that charge comes in positive and negative varieties and mass does not. The negative variety of charge simply works the opposite of a positive charge. So a positive charge gains energy when going to a higher potential , but a negative charge gives up energy (does work) when going form a lower to a higher potential.

Bob

 

[PeteVee]

If I can indulge you further, I'd really like to straighten out what the definition of potential is since the point that has been driven home is that the term voltage refers to a difference in potential between two points. I did a lot of reading today including reading through this thread and even a lot of apparently very smart people had a hard time reaching a concensus.

https://www.physicsforums.com/threads/potential-vs-voltage.92456/

Here's a quote from that thread:
"Electrically speaking, potential and voltage are the same. The only difference is the way each "term" is used to discribe a given scenerio. When you talk about voltage, you're generally talking about doing work with it. When you talk about potential, you're generally talking about relating one reference to another."

This kind of sums up the two perspectives I mentioned in my last post. Do you agree with the gist of the quote?

...voltage is not equivalent to a difference in charge.

When I made the proposition "a potential difference (voltage) is the difference in charge between two points" in my post yesterday it was off the top of my head, but interesting I consulted my copy of "Electronics for Dummies" (an apt title, I might add ) today and came across this: "Voltage is...the force that results from a difference in charge from one point to another"

Is the book definition incorrect?

 

[BobK]

Yes, it is incorrect. If you take a single charge and place it at the center of an otherwise empty universe, every point in that universe will have a potential. Where is the difference in charge in that scenario? This is what is called, in Physics a scalar field. A property of a scalar field, is that if you move a charge from one point to another the work done is the same no matter what path you take in doing so. Notice that placing the second charge at different places will have it at a different voltage. So no, voltage is not a difference in charge, it is a difference in potential. Gravity is also a scalar field which obeys exactly the same laws, hence my analogy.

An interesting aside. Many gravitationally based "perpetual motion machines" look like they could work by moving masses in the gravitational field through different paths. They don't actually work because of the property of scalar fields. Eventually, when all the masses are back at the same point, you have done exactly 0 work since their motion has not caused any change in potential.

Bob

 

[PeteVee]

Wow, that's cosmic stuff I think I'll just have to be content with a practical understanding. Thanks to all for your input.

PV