A misunderstanding repeated N times is still a misunderstanding, you are wrong, let me explain
You confuse B and H, but this is forgivable, some textbooks also does, so from now on just have B=H
i) Let's suppose that you have a 12Axx connected to 6V3 AC and your heater wires are tight twisted, AC current is 300mA and the external magnetic field is low, can we agree on that?
ii) Now just untwist the wire and make a great loop, and change to 12V6 AC, change the 12Axx connection too, now you have a greater external magnetic field with half the current, can we agree on that?
A reduction in current leads to an increase on magnetic field.
You can also revert the arguments and obtain the opposite result.
Let's suppose now that you have a single loop of wire, a magic one that always conducts the same current (*)
i) Put on it 6V3 AC and measure the magnetic field with a magnetometer.
ii) Put on it 12V6 AC and measure the magnetic field with a magnetometer.
(*) You can also use resistors and/or different wire materials/diameters, so that you can change the current "a piacere"
The result of ii) is greater than the result of i) and results cannot be reverted.
Do you hear me now?
Last edited:
I am not claiming that voltage is magically different, the voltage was a deliberate choice, the magic is in the loop and you can choose the current, so the loop has a given voltage regardless of current.
I must admit that the proposed experiment is a bit twisted, but just for fun.
Previous example with a transformer on post#72 was better, but some people do not want go against his own intuition.
Yes and no, not on that order... Again, from Maxwell equation
∇ x E + (1/c) ∂B/∂t = 0
Integrating over a surface S, applying Stokes theorem, Leibniz theorem for integrals and the definition of EMF
(1/c)∫ (∂B/∂t) . dS = (1/c) d/dt ∫ B . dS = (S/c) dB/dt =
= - ∫ (∇ x E) . dS = - ∮ E . dl = Uac
Then, a time varying voltage produces a time varying magnetic field, or the other way round, a time varying magnetic field induces a time varying voltage.
From Maxwell equation
∇ x H - (1/c) ∂D/∂t = (4 π/c) J
Neglecting displacement current, integrating over a surface S, applying Stokes theorem, etc.
∫ (∇ x H) . dS = ∮ H . dl = H l = (4 π/c) ∫ J . dS = (4 π/c) Iac
Then, a time varying current produces a time varying magnetic field, or the other way round, a time varying magnetic field induces a time varying current.
Last edited:
For those who hate classical electrodynamics and/or its math
Hysteresis Curve
You can see on the scope an AC voltage producing an AC magnetic field B; and an AC current producing an AC magnetic field H.
If this do not convince you, nothing will.
Hysteresis Curve
You can see on the scope an AC voltage producing an AC magnetic field B; and an AC current producing an AC magnetic field H.
If this do not convince you, nothing will.
Sorry but I learned that current flowing in a conductor will generate a magnetic induction B. DC current will generate permanent magnetic induction (also called magnetic field, hence the ambiguity, or flux density), AC current will generate alternating magnetic induction. In a material the magnetic induction will create a magnetic field H.
Wikipedia article on naming convention of B and H
Wikipedia article on naming convention of B and H
Has anybody heard of Ampere Turns?
The more current, the more field. True or Untrue?
How about for 1 Turn?
A loop of wire that has AC current through it sounds to me like the primary of a transformer, even if it is an air core transformer. Any other loop that is close enough and in the right orientation will pick up the AC field (becomes the secondary). True, or Untrue?
A loop of wire that has DC current through it sounds to me like an electro magnet. Any magnetic material that is close enough will be attracted to it. True, or Untrue?
------------------------------------------------------------------------------------------
"You should make things as simple as possible, but no simpler" A. Einstein
"ALL generalizations have exceptions" Me
The more current, the more field. True or Untrue?
How about for 1 Turn?
A loop of wire that has AC current through it sounds to me like the primary of a transformer, even if it is an air core transformer. Any other loop that is close enough and in the right orientation will pick up the AC field (becomes the secondary). True, or Untrue?
A loop of wire that has DC current through it sounds to me like an electro magnet. Any magnetic material that is close enough will be attracted to it. True, or Untrue?
------------------------------------------------------------------------------------------
"You should make things as simple as possible, but no simpler" A. Einstein
"ALL generalizations have exceptions" Me
Like attached modified schematic?
There is an equivalent / substitute for 2SK1388?
What power for schottky depending load?
Attachments
Last edited:
Posting equations and using them out of context does nothing for me. Just because you can plug in voltage and mu into an equation pulled out of a book doesn't mean your point about cause/effect is correct.
You need a moving electron, it's really that simple. Having a bunch of potential does not in and of itself provide meaningful magnetic fields (this is not an RF discussion). Take a loop of wire, and cut it, but leave the shape of the loop intact. Put all the voltage on each end that you like, and you won't be producing a magnetic field. You will have an electric field, which can be shielded by means of non-ferrous copper shielding, etc. Connect that loop up and allow the electrons to flow, and a magnetic field is generated. That same copper shielding will be ineffective against this magnetic field.
I fear the OP has been largely lost in the debate, but this is typical when the equation-writer posts.
You need a moving electron, it's really that simple. Having a bunch of potential does not in and of itself provide meaningful magnetic fields (this is not an RF discussion). Take a loop of wire, and cut it, but leave the shape of the loop intact. Put all the voltage on each end that you like, and you won't be producing a magnetic field. You will have an electric field, which can be shielded by means of non-ferrous copper shielding, etc. Connect that loop up and allow the electrons to flow, and a magnetic field is generated. That same copper shielding will be ineffective against this magnetic field.
I fear the OP has been largely lost in the debate, but this is typical when the equation-writer posts.
Assuming perfectly clean DC. DC, when noisy, has higher dV/dT rates than 60Hz, especially in a switcher.
A cheap switcher is an excellent source of high frequency noise that can easily couple to parts of the amplifier. A 12AX7 amp with 6BQ5s might not have to worry to much about it since miller capacitance in the pre-amp would snub some or most of it but if you start running RF tubes it might sing like a choir at 100KHz+ even if you can't hear it.
True for the field H, but it is also true that the more voltage, the more field B
True
It depends, it may be repelled too.
I must think that you do not understand equations, no offence intended.
In classical mechanics, changes in momentum are caused by forces
d(mv)/dt = F
Rewriting third and fourth Maxwell's equations
∂B/∂t = - c ∇ x E ... (*)
∂E/∂t = (c/εμ) ∇ x B + (4 π/ε) J ... (**)
Following Bohr, the derivatives describe what is changing and the causes of the changes are on the right hand side.
(*) says that the local cause of change in the magnetic field is spatial variation of electric field, in other words, a time varying magnetic field is caused by a time varying voltage.
(**) says that the local causes of change in the electric field are the curl of the magnetic field and the current density.
Those equations together imply that the electric and magnetic fields cause each other.
Happy?
It is not *that* simple.
You are denying about 150 years of physics.
Did you ever use a transformer?
Last edited:
Meanwhile, Faraday wallows in his grave, come on DF, if the displacement current can be neglected, do not be cruel with him, his law is still valid...
DF96 said:
Last edited:
Diodes are rated for peak and average current. If you have 2 diodes (center taped secondary) peak current would be the same, but average one twice lower. The same with bridge rectifier, except voltage rating.
Wow. … just wow.
However, at the risk of poplin's ire, I'll offer this (about the reference to Hysteresis Curve ) regarding the hysteresis curve on the oscilloscope.
The resistor-in-the-primary loop of the small ferrite transformer core is a current-sense device. With the unarguable linearity of E = IR … the R converts I to an E which is good input for the X-channel of the scope.
The amplifier-of-the-secondary winding on the same core is a voltage-sense device. It is not measuring magnetic flux in the core, but the rate of change of magnetic flux in the core. The rate-of-change it measures is exactly the same as the rate-of-change one might measure by amplifying the voltage across the primary all by itself. (There's no need for a secondary, as it turns out.)
If B is essentially ∂H/∂t (which is how I think poplin is wrangling it), then it makes perfect sense that the two are not just 'not the same', but also over time 'out of phase' with each other. Because in the end, we have to remember the "perfect inductor" voltage / current law:
1.
This ideal inductor is not ferromagnetic (or at least if it were, it'd be ideally ferromagnetic, which wouldn't have a bent-and-out-of-phase hysteresis curve). It is exactly lossless, and exactly resistance-less. A storage device for magnetic field energy alone.
But as poplin has also shown elsewhere, even quite well “manicured” garden variety transformer cores (and other ferromagnetic core structures) are anything but exactly linear. Though vexingly difficult to measure accurately, it appears that the inductace of metal core chokes and transformers changes with the total current flowing thru them. Because the hysteresis curve(s) are fairly tight Lissajous, transformer and core magnetoresistive losses aren't very high. High enough, but not outrageous.
Interestingly, this also implies that one might expect the setup referred to at oberlin⋅edu (above) would demonstrate a perfectly circular Lissajous, for a sufficiently large high-turn loop of nearly superconducting wires. Because the phase difference between № 1 equation above and an impinged sinusoid upon an ideal inductor, as dI/dt and I(t).
Just saying,
it doesn't have to be difficult.
GoatGuy
However, at the risk of poplin's ire, I'll offer this (about the reference to Hysteresis Curve ) regarding the hysteresis curve on the oscilloscope.
The resistor-in-the-primary loop of the small ferrite transformer core is a current-sense device. With the unarguable linearity of E = IR … the R converts I to an E which is good input for the X-channel of the scope.
The amplifier-of-the-secondary winding on the same core is a voltage-sense device. It is not measuring magnetic flux in the core, but the rate of change of magnetic flux in the core. The rate-of-change it measures is exactly the same as the rate-of-change one might measure by amplifying the voltage across the primary all by itself. (There's no need for a secondary, as it turns out.)
If B is essentially ∂H/∂t (which is how I think poplin is wrangling it), then it makes perfect sense that the two are not just 'not the same', but also over time 'out of phase' with each other. Because in the end, we have to remember the "perfect inductor" voltage / current law:
1.
I(t) = ∫V(t)/L dt = 1/L∫V(t) dt
This ideal inductor is not ferromagnetic (or at least if it were, it'd be ideally ferromagnetic, which wouldn't have a bent-and-out-of-phase hysteresis curve). It is exactly lossless, and exactly resistance-less. A storage device for magnetic field energy alone.
But as poplin has also shown elsewhere, even quite well “manicured” garden variety transformer cores (and other ferromagnetic core structures) are anything but exactly linear. Though vexingly difficult to measure accurately, it appears that the inductace of metal core chokes and transformers changes with the total current flowing thru them. Because the hysteresis curve(s) are fairly tight Lissajous, transformer and core magnetoresistive losses aren't very high. High enough, but not outrageous.
Interestingly, this also implies that one might expect the setup referred to at oberlin⋅edu (above) would demonstrate a perfectly circular Lissajous, for a sufficiently large high-turn loop of nearly superconducting wires. Because the phase difference between № 1 equation above and an impinged sinusoid upon an ideal inductor, as dI/dt and I(t).
Just saying,
it doesn't have to be difficult.
GoatGuy
I try to avoid noisy solutions, I am designing a shunt regulator for heaters, still having problems with a soft-start due to MOSFETs 3V or so is too much for a 6V3 heater. I am thinking in a BJT shunt transistor, but a lot of fakes here.
Valves are very expensive here, and if I must use garlic I will.
You are one of the few who thought of a transformer, the best example of B and H in action.
I tend to think more in voltage than in current because the expression is easier, and as I said before both solutions are consistent.
Working with DC on heaters allows me to not deal with very complicated equations for a double helix, however I am building a magnetometer to compare coax cable with twisted wire, and yes, I will think on current then.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.