|[AMRadio] Re: T3 Mod Iron Specs|
k4kyv at charter.net
Thu Feb 21 11:25:26 EST 2008
>> Anyone know the approximate "designed" frequency range for a T-368
>> modulation transformer?
>> Brian / wa5am
> The frequency range of the standard T-368 mod-transformer is 300 - 3500
> CPS. The mod iron is rated for 250 watts MAX., with a peak working
> voltage of 5000, and max altitude operating range of 10,000 ft. Maximum
> rated carrier power out of a standard T-368 400 Watts AM voice, and AM
> and 450 watts CW/FSK. NOTE: that is two different ratings for FSK mode.
> Jim> WB2FCN
We all know that a modified T-368 can do much better than that. It is not
unusual for a transformer to vastly exceed the nominal frequency response
given in the specs. 300-3500~ is the minimum guaranteed response, +/- a
given number of dB's. If the actual transformer exceeds that, it still
meets those contracted specs.
I once used an audio driver tranformer, made by Chicago Transformer Co, with
those same exact specs stamped on the case, but the thing was actually
flat from well below 100 Hz to beyond 7 or 8 kHz. So unless you plan to run
"communications grade" audio, those specs are meaningless, except for giving
you an idea of how poor the response might be. In many cases, I suspect
the communications-grade transformer comes from the same stock as similar
transfromers rated for broadcast or other higher audio quality applications,
and the company simply stamped the minimum ratings spelt out in the
contract, instead of going to the expense and trouble of modifying an
existing design to degrade its frequency response down to communications
grade, unless the customer specifically demanded it. After all, what would
be objectionable about shipping out a lower cost product whose quality is
actually superior to what was called for?
To determine the real response, you need to test the transformer with a
signal generator. Load the secondary with the specified resistive load.
Put maybe half the specified primary impedance in series with the primary,
and connect to the generator. Sweep the frequency response using a scope or
good audio level meter. Many VOM's have a calibrated dB scale. You will
need a generator capable of delivering enough voltage to get a good reading
on the scope or meter. You may be surprised how good the response on that
transformer is. IIRC, the BC-610 audio is specified at 100-5000~, and I
suspect that transformer will turn out at least as good.
>> Let's say your transmitter is limited for 200 to 3000 cycles in the
>> audio section, either by coupling caps, cathode bypass, cathode
>> resistors, plate loading, and any transformers... what is the result
>> of using an external EQ and audio chain that pumps audio from the mic
>> that is beyond either end of the audio range the transmitter is
>> allowed by design to pass?
>> I think this is a common mistake that a lot of us make that can cause
>> some serious harmonics and other artifacts.
According to the engineering data included in one of my UTC catalogues, they
recommend that the flat frequency response of an audio transformer should
extend one octave above and one octave below the intended frequency
response expected from the amplifier, to assure that the audio waveform
coming out of the transformer is close to identical to what is fed into the
transformer. Low frequency phase shift may alter the assymetry of the voice
waveform, and, with a pushpull amplifier, high frequency phase shift may
result in objectionable distortion caused by one side of the pushpull
circuit not being exactly 180 degrees out of phase with the other.
You can test the phase shift response of a balanced transformer with an
oscilloscope. Feed audio into the transformer with a wide-range audio
signal generator. On each balanced primary or secondary winding, connect
the ground side of the scope to the midtap of the transformer winding, and
connect one end of the winding to the horizontal input to the scope, and the
other end to the vertical input. Adjust both horizontal and vertical gain
for equal deflection on the scope tube. If there is no phase shift, you
should see a straight diagonal line. If there is phase shift, the diagonal
line will expand into an elongated oval shape. In the worst case, with 90
degrees of phase shift, the pattern will become a perfect circle.
If it is a pushpull grids to pushpull plates transformer, run the test with
both windings, by reversing primary and secordary and re-running the test.
If there is a balanced 500/600 ohm output with a midtap, run the test with
that winding as well. While you may not even use the midtap, the test will
nevertheless give you an idea of the phase linearity of the transformer.
Throughout the useful frequency range of the transformer, you should see the
straight diagonal. At either end of the response, it will begin to fatten
out into an oval. If there is any nonlinearity, which may occur at the
extreme ends of the response curve, the diagonal line may show curvature or
zig-zag, or the oval may change into an irregular shape.
I have accumulated a fair number of UTC Linear Standard series audio
transformers, whose rated specs are typically 20-20,000 Hz, with some better
than that and some not quite rated that good. In the vast majority of
cases, I have detected phase shift above about 11,000 Hz, as well as dips
and peaks in the response beyond that frequency. I have some older ones
that begin to drop off and show erratic response above 6000~. I have found
the same thing with broadcast quality Thordarson and Chicago transformers.
But I have seen some that actually did meet their nominal specs.
The tests I mentioned above are all without any DC through the secondary of
the transformer. If the T-368 transformer shows disappointing low frequency
response or phase shift, you might be able to improve its operation in a
transmitter by using an adequate modulation reactor and coupling capacitor,
to take the DC off the secondary. OTOH, if the tests give good results,
the characteristics my degrade considerably when the transformer is run in a
transmitter with full DC through the secondary.
More information about the AMRadio mailing list
This page last updated 22 Oct 2017.