The AM Forum
March 28, 2024, 08:25:40 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: net dc current in toroidal transformer  (Read 8295 times)
0 Members and 1 Guest are viewing this topic.
w4bfs
W4 Beans For Supper
Contributing
Member
*
Offline Offline

Posts: 1433


more inpoot often yields more outpoot


« on: August 20, 2013, 08:35:21 PM »

http://amfone.net/Amforum/index.php?action=dlattach;topic=30798.0;attach=31360

I am attempting to bring over Mikes class d transmitter using a toroidal modulation transformer

Mike, I'm not sure which Antek toroid you used but I am take a swag at the 800Va AN8440.

If so, how are you not having core saturation problems with the several Amps of DC flowing thru the paralled 115V windings ?

a neat trick ....
Logged

Beefus

O would some power the gift give us
to see ourselves as others see us.
It would from many blunders free us.         Robert Burns
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #1 on: August 21, 2013, 11:14:46 AM »

With that much average current (DC) passing through the parallel 115V windings, there will be considerable core saturation. I.e. the value of the H-field in the core produced by this average current is proportional to the total average current in both parallel windings x the number of turns in 1 winding.

The transformer will produce an audio output voltage waveform that is a distorted version of the amplifier's audio output voltage waveform. This is because, with a saturated core, the audio (time varying) component of the B-field will not be linearly proportional to the audio (time varying) component of the H-field. The details of how the non-linearity of the B v. H curve translates into distortion between of the transformer's audio output voltage waveform (v. the audio amplifier's output voltage waveform) are too mathematically complex to discuss here. In fact, the more I try to think about how one would explain it, the more complex the process seems to get.

Perhaps this distortion is why the author of the article suggests the use of negative feedback as a future enhancement.

Note: Since the output of the audio amplifier behaves almost like an ideal voltage source (i.e. a relatively low source impedance when compared to the load impedance), the distortion due to transformer core saturation, as briefly discussed above, will not be as bad as it would be if the amplifier's source impedance were comparable to that of the load. That is, the voltage across the input winding of the transformer will be linearly proportional to the voltage across the transformer's output winding... even if there is saturation within the transformer's core. If the amplifier behaves like an ideal voltage source, its output voltage will be independent of how much current it is delivering to the transformer. Placing a series resistor between the amplifier's output and the transformer's input will increase the effective source resistance, and will increase the distortion due to transformer saturation. This will happen because: the voltage at the input to the transformer will be reduced by the voltage drop across the series resistor; the voltage drop across the series resistor depends upon how much current is passing through it; and, in the presence of core saturation, the current passing through the primary winding of the transformer (and the series resistor) will not be linearly proportional to the voltage across the primary winding.

Operating the transformer with a saturated core will also produce significantly more core heating, for a given amount of average audio output power. This might not be noticed in voice applications (because, in voice applications, the average audio power is much lower than the peak audio power).

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #2 on: August 21, 2013, 02:23:59 PM »

Perhaps the DC current saturation issue could be resolved by adding an additional winding on the toroidal modulation transformer, and with a DC constant current source supply enough current to that winding (phased correctly) to null out the core flux caused by the DC current feeding the RF amplifier. The idea is to put the core B-H curve back to center. The current source would need to maintain a very high impedance over the audio modulation frequencies, from DC up to at least 5 Khz.

I attach a tube version of this concept as applied to my Gonset G50 6m AM rig. Here the modulator plate current and the RF PA plate current are set to be equal (same number of transformer turns either side of CT), and the result is a near complete null of the high flux bias that causes core saturation in a typical Heising system. The modulators being class A do provide the high impedance of a good constant current source. This is a little different then the subject of this thread, but the concept fits...a little bit.


Jim
WD5JKO


* G50.jpg (37.98 KB, 726x656 - viewed 518 times.)
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #3 on: August 21, 2013, 10:26:01 PM »

Jim

I've thought about this issue on a number of occasions: using a separate winding to inject DC to cancel out the average H-field being produced by other windings.

This can work, and may be a good solution in some applications.

The problem is that, to bring this approach to life, you will need either:

A) A large choke to allow you to inject the required amount of DC efficiently (without wasting power), while providing a high enough impedance at audio frequencies of interest. In terms of the practicalities of implementation, this approach requires a similar inductor as the inductor required in the modified Heising approach. One practical difference between this and the modified Heising approach is that the winding you use to inject the DC could have more (or less) turns than the winding used to carry the modulated B+. So, for example, if you have a choke with lots of inductance, but not rated for as much average current as the transmitter is drawing (e.g. a 20H choke rated at 300ma), you might be able to trade off impedance v. current by using a DC injection winding with more turns.

B) A current source that consumes a lot of power, all of which is wasted in heating of a resistor, or heating of the plate of a vacuum tube, or heating of the regions adjacent a PN junction of a transistor.

C) A current source that consumes a lot of power, a portion of which is wasted in heating, and the remainder of which used for some other purpose (i.e. other than producing wasted heat). E.g. in the class A audio amplifier in the Heising approach, most of the total electrical power supplied to the tube (i.e. the B+ x the average plate current) is wasted in heating the plate of the tube, but some of the electrical power is used to produce the required audio output power.

As long as producing heat is not an issue, using a tetrode to make a current source (or a transistor equivalent) would be a good candidate approach.

See also: the changes/clarifications I made to my first post in this thread.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #4 on: August 22, 2013, 02:07:29 PM »


Stu,

  All good points. So just thinking here, a power supply can be built as a voltage source, or as a current source. Each type could be either a linear type (dissipation), or a switch mode type. So building a switch mode current source that is low loss should overcome one of the stumbling blocks here. I'd have to think about how to do that though while maintaining a hi-z source impedance from DC to >= 5 Khz while the rig was modulated from 0-100% +.

  Another idea is to bias up the core with a permanent magnet. For a toroidal core, a piece of the core would need to be the magnet and then the rest of the core is normal permeable material. I have seen PM add-on for pot cores before.

Here is a thread I found on this subject:
http://forum.allaboutcircuits.com/newsgroups/viewtopic.php?t=1802

Not too many folks say much about that Gonset G-50 circuit. It was dual parallel 6L6 class A before, and still is after the change. In the stock case though, modulation below 500hz at levels above 50% modulation have observable Heising choke saturation effects. After using the P-P transformer in an unusual way to cancel the core B field, the thing modulated clean, and at high level down to < 80 hz....

Jim
WD5JKO
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #5 on: August 22, 2013, 03:07:26 PM »

Jim

In this case, I think you are going to be up against what I call the "law of conservation of aggravation".

I'm pretty sure that you will conclude that any type of current source that behaves as a high impedance at audio frequencies of interest will have an audio voltage waveform across its output that is the same (for equal turns) as the time varying portion of voltage across the modulated transmitter.

For a fixed current, this implies power flowing into the current source part of the time, and power flowing out of the current source for the rest of the time.

In the case of an inductor, the energy associated with power flowing into the inductor is stored in the magnetic field of the inductor. When power is flowing out of the inductor, it is delivered by the magnetic field of the inductor.

Somehow, your current source will need a mechanism to store and release power, without producing heat. Typical current sources will convert the power they are receiving into heat; and they will generate the power they have to deliver via a process that produces heat.

I've been trying unsuccessfully to think of a way of designing a current source that uses a capacitor instead of an inductor to store and release energy without producing heat.

One additional thought:

You could use the two output windings of the transformer to modulate two separate transmitters (with the DC flowing in "opposite" directions through the windings). Unfortunately, the AM modulation would be 180 degrees out of phase between the two transmitters... so you couldn't just add their RF outputs using an RF power combiner. However, you could (in theory, putting aside FCC restrictions) transmit two independent signals on two different center frequencies).

See also: the changes/clarifications I made to my first post in this thread.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #6 on: August 22, 2013, 06:54:19 PM »

Beefus
Jim

Perhaps one way to see how saturation of the B v. H curve causes the output voltage of the transformer to be a distorted version of the output voltage of the amplifier is as follows:

To some degree of approximation, even if there is saturation of the core (within limits), you can model the transformer as an ideal transformer with a fixed ratio of output voltage to input voltage; with a magnetizing inductance across its input.

The magnetizing inductance is constant if there is no transformer core saturation. However, if there is saturation, the magnetizing inductance will decrease as the time varying audio current passing through it increases. Again, this is an approximation of what is going on when the core saturates.

If the audio amplifier has zero output impedance (i.e. if it is behaving like an ideal voltage source), then the magnetizing inductance of the transformer has no effect. I.e., the current flowing through the magnetizing inductance has no effect on the voltage at the output of an ideal voltage source; and, therefore, no effect on the voltage across the input of the transformer. Therefore, if the amplifier is an ideal voltage source, there is no distortion.

However, since the amplifier has some Thevenin equivalent source resistance, the voltage on the output side of the amplifier's source resistance will not be linearly proportional to the voltage on the input side of the amplifier's source resistance if the magnetizing inductor's inductance varies with the current passing through it.

The bigger the amplifier's source resistance, and the bigger the value of any added series resistance between the amplifier and the transformer, the more distortion there will be at the output of the transformer as a result of transformer core saturation.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
w4bfs
W4 Beans For Supper
Contributing
Member
*
Offline Offline

Posts: 1433


more inpoot often yields more outpoot


« Reply #7 on: August 22, 2013, 08:54:52 PM »

Heavy .... much to think about

If I remember correctly the recriprocal of the output impedance of an amplifier is called the damping factor and the higher this number the lower the output Z and the better the amp for this application .... it is similar to a high power subwoofer application where the mass of the woofer cone once set in motion and the driving signal is removed will generate voltage as the cone motion is damped .... servo amps are designed and built with this effect in mind but I dont know if they are now any better or different from other high power audio amps
Logged

Beefus

O would some power the gift give us
to see ourselves as others see us.
It would from many blunders free us.         Robert Burns
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #8 on: August 23, 2013, 10:49:43 AM »


An interesting discussion, and circuit about a switch mode CCS:

http://electronicdesign.com/power/efficiently-source-constant-current-low-resistance-loads

Since Capacitor C1 in the circuit is 10 uf, that would put a 3 ohm load @ 5 KHZ towards the AC on the CCS transformer winding. Since the switch rate is 500 Khz, perhaps C1 could be 1 uf....more doable. Since the modulation would be mostly symmetrical, the CCS should stay on target in the presence, or lack of modulation so long as the critical levels are not exceeded such as CMRR.

Off on a tangent here, I know.  Undecided

Jim
WD5JKO

Logged
w4bfs
W4 Beans For Supper
Contributing
Member
*
Offline Offline

Posts: 1433


more inpoot often yields more outpoot


« Reply #9 on: August 24, 2013, 07:29:28 AM »

this may be a new one .... I may hijack my own thread  Wink

with these revelations, it would be simpler to not use the audio pa and modulation xfmr and instead use series modulation with split dc voltage source .... efficiency may be close and implementation would be simpler ... Occam's razor
Logged

Beefus

O would some power the gift give us
to see ourselves as others see us.
It would from many blunders free us.         Robert Burns
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #10 on: August 29, 2013, 09:10:57 AM »

Jim

One additional thought:

You could use the two output windings of the transformer to modulate two separate transmitters (with the DC flowing in "opposite" directions through the windings). Unfortunately, the AM modulation would be 180 degrees out of phase between the two transmitters... so you couldn't just add their RF outputs using an RF power combiner. However, you could (in theory, putting aside FCC restrictions) transmit two independent signals on two different center frequencies).


Stu

You could recombine the signals by flipping the phase of the output of two transmitter sections. This could be done at the recombination of the two outputs, or earlier at low level depending on what you want the output section to look like. That's just a PP output in its simplest form. In the case of what QIX does, the flip and recombine is done on purpose in order to use the PP nature to drop even harmonics via cancellation.

Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #11 on: August 29, 2013, 11:09:42 AM »

Bear

The phase at issue is the audio phase, not the RF phase. When the envelope of one of the transmitter's RF outputs is at maximum amplitude, the envelope of the other transmitter's RF output will be at minimum amplitude.

An RF combiner, cannot undo this audio phase flip.

In mathematical terms:

One transmitter's RF output will be A x [1 + m(t)] x cos(2 x pi x f x t)
The other transmitter's RF output will be A x [1 - m(t)] x cos(2 x pi x f x t + theta)

Where, A is a positive constant having the units of volts, m(t) is the audio waveform, f is the RF frequency in Hz, and theta is an RF phase shift in radians.

An RF combiner can easily compensate for the RF phase shift, theta, whatever its value is...  but when the audio signal, m(t) is positive, the envelope of the first RF signal [which is proportional to 1 + m(t)] will be larger than A , and the envelope of the second RF signal [which is proportional to 1-m(t)] will be smaller than A.

The output of the RF combiner (including compensation for the RF phase shift, theta) would be:

At the RF combiner's output port 1:

0.707 x A x [1 + m(t) + 1 - m(t)] x cos(2 x pi x f x t) = 0.707 x A x [2] x cos(2 x pi x f x t)

At the RF combiner's output port 2 (usually a dummy load):

0.707 x A x [1 + m(t) - 1 + m(t)] x cos(2 x pi x f x t) = 0.707 x A x [2 x m(t)] x cos(2 x pi x f x t)


That is, the output of the RF combiner at its output port 1 would be an unmodulated signal with twice the carrier power of either input signal. The output of the RF combiner at its output port 2 (usually a dummy load, but could be connected to an antenna) would be a suppressed carrier DSB signal containing all of the modulation power of the individual input signals.



Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #12 on: August 29, 2013, 11:20:06 PM »

Toroids are more sensitive to DC than the old EI stuff but it has the issue as well, if you have ever run 300mA through the secondary of even a KW size BC modulation transformer esp. an old '40's one, the audio quality goes to pot pretty fast.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.061 seconds with 19 queries.