Using a toroidal transformer as a mod transformer Part II

[w4bfs]

Ok Stu ... I'm going thru this quickly so please check my math ... calc of primary to secondary impedance ratio: 6400/8 = 800 and the square root of this gives turns ratio so (800)1/2 power gives a turns ratio of 28.28:1 so retaing the 1143 turn secondary this give 1143/28.28 = 40.42 or a 40 turn primary .... what do you think ....leave the other windings alone, add some extra insulation and wind a 40 turn primary of at least .... I = (P/R)1/2 = (400/8)1/2 = (50)1/2 = 7.07 Amp @ 400w level ... 16ga should be ok, teflon insulated if possible ... can you get that much more thru the core?

freq resp testing should be checked ... 40 turns may not establish sufficient flux for low freq transfer ... 73...John

[W1DAN]

Stu:

Sounded real smooth on the air today...

Good job!

73,
dan
W1DAN

[AB2EZ]

Dan

Thanks! 

John

I think your proposal is interesting.

I need to use both "800 VAC" windings in series... because I want to keep the B-field (and, by implication, the H-field) within the original design intent under "worst case" conditions.

I.e. V~n dB/dt. If dB/dt is a sine wave, then: B(peak) ~ V/(2pi x f x n), where f is the frequency, and n is the number of turns. For this application, I would define "worst case", in terms of the potential to saturate the core, as: V= 1600 VAC rms (2262 volts peak), and f=50 Hz sine wave modulation.

My target turns ratio (now obtained with two transformers in tandem) is 26:1, in order to match 8 ohms into ~5333 ohms (1600 volts B+ and 300 mA => 5333 ohms).

Using your suggestion, I would need to wind a new primary on top of the existing windings with 1143 x 2 / 26 turns ~88 turns. [Basically, this is the same as what you calculated, once you take into account the need to use the two 800 VAC windings in series, and my target modulation resistance of 5333 ohms]

If I did that, I would only need one transformer. 

Sometime when its raining and/or cold outside, I may try it. [~5.5A (rms) when testing with a sine wave]

At one point, I asked the fellow who owns Antek to make up a special version of the transformer (one unit, of course) with the 120 VAC primary windings cut in half ... which would accomplish roughly the same thing. However, I think he was (appropriately) focused on other things... like filling large orders.... and I also think he might have been a little nervous about what I was going to do with a transformer like that:  

"Hi, I'm an amateur radio operator who wants to purchase a custom 800 VA transformer from you... rated for 60 VAC input and 1600 VAC output. Since I'm not acting on behalf of a corporation, I can't really indemnify you for the consequences of what I do with it. In fact, if I kill myself, my wife will probably hire a lawyer to sue you for selling a dangerous, non-standard product to someone who is just an "amateur". I would like only one... I'll never buy any more... and I'm only willing to pay slightly more than the price you charge for your unmodified stock units".

As an aside, it looks like each 171 turn primary, which corresponds to a "120 VAC" winding, also corresponds to a "120 VAC rms x 1.414= 170 volts peak" winding. So, maybe these things are intentionally designed so that 1 turn => 1 volt peak. On the other hand, maybe its a coincidence.

Best regards
Stu

  

[w4bfs]

I guess my last concern is about peak voltages between the 800V windings ... If they are also bifilar wound, I believe the peak voltage will appear near each of the ends ... could be the show stopper ... I want to commend you on your excellent work and look forward to a qso ...Stu is the man ! ...73 ...John

[Steve - WB3HUZ]

Heh. Excellent idea. I wonder if you could make it work with a good variac and use the existing winding?

Stu,
Most low frequency toroid cores are tape cores. From my early days of switchers I remember the thinner the tape the better the high frequency response. So this could cause a run on variac cores. Imagine winding a homebrew transformer or choke  on a 240 volt 30 amp variac core....or a 20 amp 115 VAC core.....bring out your dead.

[WA1GFZ]

Steve I have run audio through variacs but never measured the performance. I know a 60 Hz variac works great at 400 Hz. There is no reason why the existing winding couldn't serve as the primary. Then just add insulation and the secondary of choice over it. The only limit is the size of the hole to pass the wire shuttle thriugh. A 20 amp 115 VAC variac isn't very large.

[w4bfs]

yes ... been thinking about this too ... I wonder if Antec (sp?) would sell blank cores ... seems like the secondary might need to be several layers with good insulation between each ... assuming a 600V limit for Formvar this would point to say 5 layers ... I havn't done much work with toroids, how does this sound ?

[WA1GFZ]

Yup, it should work. I would also not bifilar wind it. I would put each winding on 180 degrees of the core with some space between them. This way each winding is the same distnce from the core and have the same resistance.
You will need to build a shuttle to hold the wire as you pass it through the core.....or have a big back yard.
I bet a used variac will be cheaper than a bare core. The full winding on a variac is usually good for 135 volts at 60 hz
"I tested my rig under full modulation, just plugged the mod transformer into a wall socket".   

Now Gary the transformer man can cob up a toroid winder. Not that would be an interesting device

[w4bfs]

do toroids have a problem with leakage reactance like EI cores ?  if not then I assume no need to interleave pri/sec windings  ..iszat so ? .... at least doing multiple windings eases up on how much wire you have to carry in the shuttle ...sounds like fun ...73...John

[WA1GFZ]

leakage reactance is an issue with all inductors. High ratio transformers are not easy.

[AB2EZ]

Leakage inductance is a lot less of an issue if the permeability of the core material is very high (as in the case of ferrite cores, provided they are not saturated by the application of an unbalanced DC current to the windings)

Leakage inductance ~ [the H field in the air near the surface of the core of the toroid x the cross sectional area of the portion of the winding that is not occupied by the core] / [the H field in core x the cross sectional area of the core x the permeability of the core material].

The H field in the core of the toroid is approximately equal to the H field in the air near the surface of the core of the toroid (Somewhat higher on the inside of the toroid, and a somewhat lower on the outside of the toroid...  because the line integral of H around the toroid is the same on any closed path that lies within the winding)

If the permeability of the core is 1000, then even if the winding is loosely would around the core (lots of space for insulation) the leakage inductance will be a lot lower than it would be for a tightly wound winding on a core whose permeability is a lot lower.

[You can really "see" this effect in the toroidal rf transformers that many of us use to build Class E transmitters. Talk about loosely wound    ... in a 1:N transformer, one of the "windings" just passes through the core. Most of the cross sectional area of that "winding" is occupied by the air space enclosed by the two leads going from that "winding" back to whatever it is connected to. Nevertheless, it works like a nearly ideal transformer. If you were to open the secondary, it would have a lot of inductance (I use these to make safety chokes)... but with the secondary connected... there is essentially no apparent leakage inductance.]

For example, if the cross sectional area of the portion of the winding that lies outside the core is 10% of the cross sectional area of the core... then the ratio given above is approximately 0.1 / permeability of the core material

Stu

 

[w4bfs]

[You can really "see" this effect in the toroidal rf transformers that many of us use to build Class E transmitters. Talk about loosely wound    ... in a 1:N transformer, one of the "windings" just passes through the core. Most of the cross sectional area of that "winding" is occupied by the air space enclosed by the two leads going from that "winding" back to whatever it is connected to. Nevertheless, it works like a nearly ideal transformer. If you were to open the secondary, it would have a lot of inductance (I use these to make safety chokes)... but with the secondary connected... there is essentially no apparent leakage inductance.]

For example, if the cross sectional area of the portion of the winding that lies outside the core is 10% of the cross sectional area of the core... then the ratio given above is approximately 0.1 / permeability of the core material

Stu

 
[/quote]

Ok ...I'm still trying to figure out quote feature  ... please 'bear' with me ... sounds like you have just described a current transformer ... use a lot of those for power plant stuff ... some people have been hurt when open circuiting the secondary of a ct in an energized ckt ... seems the open ckt voltage will climb towards infinity ... have never really ynderstood why ...73 ...John

[AB2EZ]

John

"Current transformers" are optimized for their specific application... but for the purpose of answering your question....

Like any properly-functioning transformer with current passing through both windings:

The primary current x the number of turns in the primary ~ the secondary current x the number of turns in the secondary (the difference from equality having to do with the magnetizing inductance of the transformer)

If the primary has 1 turn (just a wire passing through it) and the secondary has 10 turns, then current in the primary will be 10x the current in the secondary (again: provided both windings have current passing through them)

However, if you open up the secondary while the primary has current passing through it, then the transformer's secondary will have a voltage across it of: 10 (secondary turns) x the rate of change in the magnetic flux in the core. (This is Faraday's law of induction).

After an initial transient, and assuming the core material doesn't saturate...the rate of change of the flux in the core is equal to the cross sectional area of the core x the permeability of the core material x the rate of change of the current in the 1-turn primary / the circumference around the core. (This is Ampere's law)

The rate of change in the current in the 1-turn primary is 2 x pi x the amplitude of the current x the frequency of the current (e.g. 60 Hz). (This follows from basic calculus)

Putting it all together... if you have a conductor carrying 1000 amps at 60 Hz, and you clamp a 1:1000 turn current transformer on to it, with the secondary properly terminated in a 1 ohm resistor... then the current in the secondary will be 1 amp (i.e. 1000 amps / 1000 turns) and the voltage across the resistor will be 1 amp x 1 ohm = 1V

If you open the secondary, then (after an initial transient) the voltage across the secondary will be:

(1000 amps x 2 x pi x 60 Hz) x 1000 turns x the cross sectional area of the core (in the plane of the windings) x the permeability of the core / the circumference around the core (in a path that is perpendicular to the windings).

Suppose the permeability of the core material is 800 x (4 x pi x 10**-7) Henrys per meter (i.e., the relative permeability of the core material is 800, and 4 x pi x 10**-7 H/m is the permeability of free space); and suppose you have a toroidal-shaped core whose cross sectional area is 10mm x 25mm, and where the average circumference around the core is 2 x pi x 50 mm. [I.e. the core is 25 mm long, 110 mm in outside diameter and 90 mm in inside diameter]

Then the voltage across the open secondary will be:

(1000 amps x 2 x pi x 60Hz) x 1000 turns x (.01 meter x .025 meter) x 800 x (4 x pi x 10**-7) Henrys per meter / (2 x pi x .050 meters) = 302 volts

The voltage drop of the 1-turn primary passing through the core will be 0.302 volts

Stu

[w4bfs]

Ok , one last question ...In the toroid, is a one turn winding (wire enters the annulus-wraps around the core and exits thru the annulus) the equivalent of just a conductor entering and exiting the annulus (straight thru the center) ? Intuitively, I think not.  There is certainly a difference in how much of the surface of the core is exposed to the wires B field .... I am sorry to say my calculus is beyond rusty and is more like corroded javascript:void(0);

[AB2EZ]

John

The winding you described would correspond to two turns.

A wire with a current flowing through it has a magnetic field curled around it. Since the core has such a high relative permeability, for all intents and purposes... only the portion of the wire that passes through the hole in the toroid creates the effect of a "turn".... in the sense of producing an H-field inside the core... that goes around the core in the plane perpendicular to the wire.

See the .jpg attached below

For the winding you described, the winding passes through the hole in the toroid two times... thus it produces the effects of two turns.

Even when there is only one pass through the hole in the toroid... both ends of that wire go somewhere... and they form a "turn" in that respect.

Stu

[af6im]

Great post Stu. Homebrewing and mythbusting, a GREAT combo!

73,
AF6IM
Mark

[w4bfs]

yes Stu and all the other fellows who take their personal time to help clear things up KUDOS

[KC2ZFA]

I'm necro'ing this thread to ask whether Stewart, or anyone else, thinks
these amps are well-suited for use in this modulation scheme:

http://tinyurl.com/3lmylrn

73 de Peter

[AB2EZ]

Peter

These amplfiers look okay... but there are two things to consider

1. If the maximum output power capability of the amplifier is much higher than it has to be to modulate the rig, then there is a risk that you will produce a high enough peak voltage at the output of the step up (modulation) transformer to blow something up... e.g. the Heising capacitor, the Heising choke, the step up transformer, the r.f. tube's bypass or blocking capacitor, the r.f. tube

As an example:

In my homebrew 375W rig, the B+ on the RF tubes is 1700 Volts. The corresponding average plate current is 300 mA. [510 Watts of input power to the Class B RF deck].

The modulation resistance (not the RF load impedance) is 1700V/0.3A = 5670 Ohms.  

I am using an audio amplifier rated at 800 Watts maximum audio output into 8 Ohms (bridged mono). My step up transformer has a turns ratio of approximately 26:1. The 26:1 step up transformer converts this to 5670 / (26 x 26) Ohms = 8.4 Ohms... which is an excellent match for the audio amplifier.

I need 1700V  x 1.25 of audio voltage (peak, not RMS) to obtain a 125% modulation peak. To get 1700V x 1.25 at the output of the step up transformer, I need 1700V x 1.25/26 = 82V of peak (not rms) audio from the audio amplfier. This corresponds to a peak audio output power of 82V x 82V / 8.4 Ohms = 800W.

Assuming the the maximum audio output power rating of the audio amplifier corresponds to rms output power, I have a comfortable 2:1 margin between the 800W rms output power rating of the amplifier (into an 8 Ohm load) and the 800W peak audio power needed for 125% modulation.

If the audio amplifier was delivering its maximum output power of 800W rms into an 8 Ohm load... this would correspond to 113V peak output voltage, and therefore 113V x 26 = 2941V peak at the output of the modulation transformer.[Remember, these audio amplifiers behave as ideal voltage sources... so 113V is equal to the peak output voltage of the amplifier... regardless of the load on the amplfier.]

Therefore, I have to make sure that the stuff connected to the output of the modulation transformer can handle 2941V of peak audio... because that's how much the audio amplfier will produce if an input audio spike occurs.

If you use an audio amplfier rated for 2000W maximum into 8 Ohms... (much more power than you need to modulate the rig)... then you will have to consider that will happen when that amplifier puts out its peak voltage.

It is okay to use a lower step up ratio for the modulation transformer... to limit the peak voltage at the output of the step up/modulation transformer. This will imply that the audio amplifier is looking into a higher load impedance than the load impedance that results in its maximum output power (which is fine for these amplifiers).

Bottom line: these amplifiers should be okay... provided you select the turns ratio of the step up transformer to be just enough to produce a peak audio output voltage (in response to an audio input transient) that the step up transformer and the components connected to the output of the step up transformer can handle.

2. Some audio amplifiers are not well shielded against r.f. and/or external magnetic fields. If I recall correctly, Tom (K1JJ) had to put a lot of space between his audio amplfier and his RF deck when he used this approach drive the last audio modulator stage of his 4-1000A rig.

Stu

[flintstone mop]

Right Stu
I remember K1JJ's grief from transients. Fortunately the amplifier didn't mind the inconveniences and jolts. And Tom was able to tame everything down to prevent any damage.

The ease and low cost for such high power amplifiers was just a dream about 5 yrs ago.
So, it seems that there would be more engineering needed to use these high power amps and the 'mod transformer' to interface to your typical plate modulated P.A.

Fred

[w4bfs]

you might consider a string of varistors (high tech) across the secondary or a spark gap (low tech) for peak voltage protection

In Electric Radio mag the spark gap spacing (in inches) between 'horns' of the spark gag = (V - 350) x 10 -5  

example: want a spark gap to fire at 2500V so the spacing would = (2500 - 350) x 10-5  = 2150 x 10-5  = .0215 inches

can set using feeler guages

By the way, I got around to dismantling one of the Antek low voltage toroids ... seems like all winding sets are bifilar wound with extra insulation ... makes sense when you consider how the hipot test is specified .... if only 2 layers of Formvar are the insulation between the bifilar twins then the question that gets begged is how do the high voltage windings survive when hooked up as full wave center tap ? (in series)  so I'm still wondering about survivability as a high voltage mod xfmr ...73 ... John