Sound Technology FM generator question.

I have a trusty old Sound Technology ST1000A FM generator and I have a
question regarding the variable FM output level. The owner's manual states
that the output impedance is 50 ohms, VSWR<1.3, 200Vdc isolation. My
question is if I am using this device to check the sensitivity of an FM
tuner, will the output level be affected if I connect the output terminal to
an RG-58 cable and then use a 75 ohm to 300 ohm matching transformer to make
it compatible with the old FM tuners? It seems to me the 2:1 step up ratio
of the matching transformer should affect the level not even considering
the fact that the 75 ohm input of the matching transformer does not match
the 50 ohm output of the cable.

Thanks for your reply.

 

The fact that the generator 50 ohm output and coax has a 75 ohm load
will give you a voltage of 0.6 of the generator open circuit voltage. If
the load had been 50 ohms, you would get 0.5 of the open circuit
voltage. The increase is 0.6/0.5 = x1.2 (+1.58dB).

75-to-300 ohm transformer should give you a 2-to-1 voltage step-up
(+6dB).

The transformer will have some loss. This should not be more than about
0.5dB.

So, the voltage at the FM tuner 300 ohm input will be the generator
output (into 50 ohms) + 1.58dB + 6dB -0.5dB = Vout + 7.08dB.

This assumes, of course, that the tuner input impedance really IS 300
ohms (which it probably isn't!).

 

I believe the output level dial is calibrated for a 50 ohm load. So using
your numbers if the output dial is set to 10µV then I calculated a 7.08dB
gain to be 22.6 µV. In effect, the sensitivity of the tuner at this point is
7 dB worse than what the dial indicates. Is that right?

Thanks for your reply.

 

The calculation looks OK. However, I suppose it depends on what your
'standard' impedance is.

If you have a 'normal' halfwave dipole at (say) 100MHz, the impedance at
the centre will be about 75 ohms. It would be normal to connect it (via
a 75 ohm feeder) to a tuner with a 75 ohm input impedance. [Note: It
might be more correct to say that the tuner is designed to work best
when fed from a 75 ohms source. In practice, it might not have a very
good 75 ohm input.] Anyway, let us assume that the level of a received
100MHz FM radio signal level (into 75 ohm) is 1μV.

Now, if you replace the 'normal' dipole with a folded dipole, you would
use 300 ohm feeder and connect it to a tuner with a 300 ohm input
impedance. Ignoring distractions like differences in feeder loss, the
tuner input voltage will be 2μV.

However, despite being fed with twice the voltage, the 300 ohm tuner
won't work any better with the folded dipole than the 75 ohm tuner works
with the 75 ohm 'normal' dipole. In both cases, the input power is the
same. Internally, the electronics will be basically the same. The only
difference will be in the matching circuit between the input and the RF
stage.

So, it is reasonable to conclude that, when you specify the sensitivity
of a receiver, you have to specify the impedance. In your test, if your
standard is 300 ohms, then you would say that your tuner was receiving
22.6μV (7.04dB more than indicated on the generator dial). If it was 75
ohms, it would be 11.3μV (6dB less).

Well, I think I'm correct! What do you reckon?

 

I don't think you need to mention the impedance of the receiver when you
mention FM sensitivity. If it is has the correct antenna and the correct
impedance matching circuit, the sensitivity of the circuitry should be
enough. No? That way I can compare one receiver/tuner to another and not
care what the input impedance is. It's kind of like stating that the power
of a speaker is independent of what type of amp it's hooked up to.

I did a little more digging in the owner's manual. There is an intermediate,
almost lossless, antenna matching network (which I don't have) that is
supposed to be hooked up between the tuner and the generator to make things
go together nicely. There is also a simple schematic diagram of how to
construct one yourself if you don't have the Sound Technology matching
device but the drawback is the ~6dB attenuation. So to go from the 50 ohm
unbalanced output of the generator to the 300 ohm balanced input of the
tuner would require three resistors connected as follows:

R1 130 ohms from the center terminal output of the generator to one of the
300 ohm inputs of the tuner.
R2 150 ohms from the shield side of the generator to the other 300 ohm input
of the tuner.
R3 62 ohms which goes directly across the generator output.

In this configuration, the readings on the dial are approximately twice the
actual output. In other words, if the dial is reading 10?V, then there is
only 5?V going to the tuner.

Thanks for your reply.

 

Dave,

Sound Technology offered a 50 to 300 ohm transformer, model 100, that
allowed you to direct read the dial on the 1000A. there was also one
for 50 to 75 ohm conversion, but I've never seen one.

Regards,
Tim Schwartz
Bristol Electronics

 

No. If you have a 75 ohm receiver which requires 'V' volts for a given
SNR, a 300 ohm receiver would require '2V' volts to give the same SNR.
If you were comparing the two, you would need to mention what impedance
applied.

For all you know, the 300 ohm receiver is simply the same model as a 75
ohm version, except that it has an internal 300-to-75 ohm step-down
transformer. For the same performance, the 300 ohm model would
definitely need twice the input voltage.

If you have 75 ohm signal source and a 300 ohm receiver, you can get
'free' gain by simply using a step-up transformer to give you more
source voltage. The limit is when the transformer matches the source to
the load. If you try to us a step-up transformer between a 75 ohm source
and a 75 ohm receiver, you will actually get LESS into the receiver than
you would with direct connection.

Ah, but....
With audio systems, the output impedance of the source is very low
compared with the load impedance. The power fed into the speaker only
really depends on the source voltage and the speaker impedance.

This will probably OK if the receiver input is completely floating, with
no reference to ground (including a centre tap on receiver input coil or
transformer - if it has one).

I reckon that you are better investing in a simple ferrite-cored
75-to-300 ohm balun / matching transformer (or make one - dead easy) and
allow for the small corrections discussed previously. I see that Tim
Schwartz has suggested 50-to-300 ohms but, to me, 75-to-300 ohms is much
easier as it is a simple turns ratio of 2:1. If you really want
50-to-300 ohms, it would be simpler to add a couple of resistors to act
as a resistive 50-to-75 matching pad. But, in over 40 years in cable TV,
I never worried too much about mixing 50 and 75 ohm impedances (but only
where it didn't matter, of course!).

 

If I'm understanding you correctly, this is analogous to the ratings on a
volt meter where you would need to know the internal impedance of the meter
to be able to factor in circuit loading. Yes?

Thanks for your reply.

 

Hopefully the opposite (I think). With luck, the impedance of your
voltmeter will be sufficiently high that has negligible loading effect
on any circuit you are likely to come across.

Just to recap, I think that you don't need to worry too much about the
difference of impedances between the source and the load. Even if it is
balanced, provided that the input of the receiver is floating, you
should be able to feed it directly from an unbalanced generator (with or
without any additional resistors).

But what you do need to take into account is that, for the same
performance, a high impedance input receiver expects a high input
voltage, and a low impedance input receiver expects a low input voltage.
It doesn't really matter how the voltage gets to the receiver input - as
long as you know what it is (whether by measurement or by calculation).

If you are concerned about whether the receiver input is floating, you
can overcome this with a simple balun. This can be a 1:1 ratio
transformer so, effectively, the receiver will still be fed from a 50
ohm source. You simply use a few corrections (as discussed) if the
receiver input is not 50 ohms. However, as you know that you have a 50
ohm generator and a 300 ohm receiver, you might as well use a home-made
or an off-the-shelf matching transformer and balun. You can get these
with an F-connector for the 75 ohm end and a couple of spade connectors
for the 300 ohm end (I have a couple myself). I see that Radio Shack
sell them:

<http://www.radioshack.com/product/index.jsp?productId=2062049>

Although the technical spec is more-or-less non-existent, the insertion
loss should be about half a dB.

 

Tuner sensitivity is measured in dBf -- dB with respect to a femtowatt. 2uV
into 300 ohms is the same power as 1uV into 75 ohms.

 

Ah! That is what I would have expected.

The point is that, in order to measure the sensitivity, you don't have
to feed one from a 75 ohm signal generator, and the other from a 300 ohm
signal generator. But you do need know what the input voltage really is.

Of course, for this sort of equipment, you probably can't rely on the
input impedances being anywhere near the nominal values. What you
calculate and what you get could be quite different. For greatest
accuracy, it would be best to use a resistive and/or transformer
matching circuit.

For matching between 50 and 75 ohms, a cheap cable TV in-line attenuator
(the type with F-connectors) suffices. 6dB is enough, and (if my brain
is working correctly) the attenuation would be 6db - 1.58dB = 4.42dB.
[The 1.58dB is the voltage rise when going from a 50 ohm source to a 75
load.] If you use a transformer to get to 300 ohms, you'll need to
subtract the loss of around 0.5dB, which makes an attenuation of around
4dB. So you're 6dB up because of the 75-to-300 ohm conversion, and 4dB
down because of the attenuator pad and transformer loss - making a
convenient +2dB higher than what the output meter on the signal
generator indicates.