How to measure impedance ? [Electronics]

[Externet]

Hi. This could be on some other forum, but, here it goes.

 

If any electronics guru is around, I would like to know how to measure the (antenna) input impedance of a television.

From ~50MHz to ~800MHz. in whatever steps. 1 MHz, 5MHz, 10 MHz...

Would you have the proper instrument ? What is it ?

[Enthalpy]

Rich people use a "network analyzer". Comfortable, overkill for one antenna. Find a radiocomm laboratory that has one, or buy a used one on eCreek for instance. Takes some time to learn for someone used to VHF and UHF.

 

Amateur radio used a measurement bridge years ago, this must still exist. Once you have a generator, the bridge is dirt-cheap to tinker.

 

You can also tinker a meter for reflected waves (a directional coupler). Needs a generator, and won't detail the real and imaginary part of the impedance, but the result gives directly the quality of the matching.

 

All methods involve feeding the antenna with a signal, which get radiated, and this is not desired...

- Do it in an anechoic chamber if you can

- Or do it quickly with very little power in an isolated valley and don't be caught...

- A noise source is better than a single frequency for that.

[John Cuthber]

Entalpy,

in this case all the techniques involve dissipating power in the telly. That's a different problem from transmitting it.

[Externet]

All methods involve feeding the antenna with a signal, which get radiated, and this is not desired...

- Do it in an anechoic chamber if you can

- Or do it quickly with very little power in an isolated valley and don't be caught...

- A noise source is better than a single frequency for that.

Thanks.

I do have a antenna analyzer up to 170 MHz. Even if was the range I want, it is not the case.

 

Reworded ----> Am looking to measure the impedance a television tuner antenna input presents to over-the-air signals. Not a tv antenna impedance.

And if you think it is 75 ohms; well, cannot be at more than a few points in the range. Unable to find credible plotted data on the net. TV tuner data sheets do not reveal the truth.

 

A typical antenna (not tuner !) plot looks like :

----> http://s588.photobucket.com/user/Innernet/media/Screenshot-TVantennapdf.png.html?sort=6&o=40

[Enthalpy]

Oops.

 

Well, the methods are the same. Do not overload the preamp; odBm is already an important signal for it.

 

The hidden question is: a well-made preamplifier is NOT optimum when its input impedance matches the cable, because apart from getting the maximum input power, it must also create the minimum noise, which does not happen at impedance matching. The minimum noise is when the driving resistance is the ratio of the preamplifier's noise voltage and noise current, and the susceptances of the soure and preamp compensate.

 

- The optimum impedance is between the source matching and the minimum noise. The impedance mismatch is easily 2:1.

- This optimum impedance is impossible to reach over such a wide band, and difficult even over 470-860MHz.

- 50 to 800MHz covers two VHF TV bands and one UHF (split in two usually). The TV's input probably has a 2 or 3-way filter feeding as many preamplifiers, so a random impedance between the bands would be absolutely normal, and even desireable to attenuate outband noise.

- You get the optimum generator impedance (hence the optimum preamp impedance seen by the generator) in the datasheet of the preamplifier transistor (...or tube, I've seen the last ones). The manufacturer uses to give a noise figure and an optimum impedance in VHF and UHF+, rather than a voltage and current noise in LF.

 

Still feeling a need to measure it?

[Enthalpy]

Why TV tuners don't tell the input impedance? All the ones I've seen are meant for 75 ohm, the cable impedance that gives the minmum loss. So I imagine TV producers prefer not to worry the customers with that. In addition, matching a 75 ohm cable or source with a 75 ohm load makes very little difference in the power transfer (slightly more in the noise power, still negligible).

 

Imagine a 1Vrms 50ohm source: a matched 50ohm load gets 5mW, a 75 ohm load 4.8mW, that is 4% less or 0.2dB.

 

So just feed with 75 ohm. I've never seen a TV with a symmetric antenna input; such an input would be 300 ohm (seen 240 ohm on German FM sets, it makes no difference neither).

 

If willing to measure a preamp's input impedance, -30dBm is about the maximum, -40dBm better.

Edited August 25, 2014 by Enthalpy

[John Cuthber]

Impedance matching doesn't just affect power transfer.

If the antenna, cable, and preamp don't all have reasonably matched impedances, then you would get reflection and, potentially, ghosting on the image.

 

I have seen antenna inputs where the first thing the signal meets is a 50 ohm resistor to define the input impedance, followed by a fast, high Z, buffer amp.

If you can open up the TV and check to see if there's a 75 ohm resistor, then you don't need to make the measurement.

[Enthalpy]

A matching resistor worsens the noise, and normally you shouldn't find any in a good preamplifier. Well, everything exists. This is fully acceptable if a preamplifier is near the antenna, and the television receives a strong signal, possibly at its mixer.

 

Antennas use to be poorly matched because covering a wide band, like 470-860MHz, demands trade-offs.

[Externet]

Thanks, gentlemen.

Knowledge, knowledge... the key for everything. Having the test equipment shown is not the whole story.

 

Scratching my head ----> http://s588.photobucket.com/user/Innernet/media/P1010489_zpsde41b47e.jpg.html?sort=6&o=77

(Text below pictures)

 

Edited/added : ----> http://s588.photobucket.com/user/Innernet/media/P1010491_zpsc2b6321c.jpg.html?sort=6&o=80

 

and ----> http://s588.photobucket.com/user/Innernet/media/P1010493_zpsf30db7fa.jpg.html?sort=6&o=79

 

How much am I goofing ?

Edited August 25, 2014 by Externet

[John Cuthber]

Just out of Idle curiosity, why on earth do you want to know the input impedance of a TV?

[Enthalpy]

Yes, why?

 

In addition, the measure setup isn't very good... It will give some notion of the impedance seen at the coaxial T, that is, through a length of coax. Passing by the preamp with a T right there, then going to the meter, would have fewer drawbacks; it would tell by how much the preamp's input shunts a 25ohm resistance.

 

The generator probably shows an output impedance near to 50ohm resistive. The meter is more difficult to make resistive; it uses to be more accurate if the input is an attenuator with a mechanical switch.

 

Anyway, this setup will not detail the resistance and reactance, which demand a vector meassure (0° and 90° components), as made by a network analyzer. But a measure of how good the matching is can be done with a reflectometer (a directional coupler)

http://en.wikipedia.org/wiki/Power_dividers_and_directional_couplers

without demanding a vector measure.

[Externet]

Thanks, dear very respected gentlemen.

Most modern TV tuners use mosfets/fets devices at front end. Being high impedance gate devices, I believe they do not respond properly to circuits ahead of them to force near matching to 75 ohm antennas, wasting signal capture. Am just trying to confirm that is a cause for marginal reception/pixelation in ATSC. And attempt overcoming the limitation.

So far, using a 280 ohm antenna on a TV instead of its original configuration, got rid of a plague of marginal/multipath/loss reception problems. Suspect the industry would do much better breaking ties with 75 ohm pseudo standards that correspond to other eras and not to modern tuner design. But instruments are based fixed on mostly 50 ohm standards and their response to other impedances equipment is simply wrong. And manufacturing/design uses such 50 ohm intruments. There are no others.

I believe modern tuners are forced to near match 75 ohm because nobody builds antennas in other impedances that could outperform the currently available ones.

Will, at some point, modify a few tuners to route the antenna signal to the gate without much loading and evaluate results. All within my limited skills as you have noticed.

Now I changed the configuration of my test equipment for evaluation.

The analyzer in duplex mode, 0 MHz shift, instead of tracking generator. Means both RF generation and reception analysis simoultaneously working. Switchin-in the tuner in circuit is revealing sharp deep valleys loading the generator at many spots in the span. Well, there are compartmentalized tuning sections for each band, some of that is justifiable.

Yes, I have a Celwave circulator model CC460-S sort of directional coupler, but of limited frequency span.

Spanking welcome.

[Enthalpy]

MOS are excellent UHF preamplifiers (and mixers). They bring a low noise and a better linearity (against intermodulation and transmodulation) than bipolars. Using them is absolutely right.

 

Impedance matching is done rather easily and well in VHF and UHF. Traditionally we used transformers, presently wideband LC circuits are more common. This permits the MOS to see the high source impedance it needs to be silent (its noise voltage is rather strong, but its noise current very small, so a transformer that multiplies the signal voltage makes a good signal-to-noise ratio). It also permits the source to see an impedance that matches itself more or less: not the several kohm of a dual-gate MOS input, but something like 40 ohm for a 75 ohm input.

 

This permits good MOS preamplifiers to have a noise figure like 1.0dB or even less, which does indicate a proper matching.

 

It also permits a MOS to amplify. Without impedance matching, 25mS transconductance with 50ohm input and output would result in nothing. Only the voltage gain at the input transformer, and the kohm load impedance at the drain, permit a MOS to amplify.

 

The strong impedance transformation needed by a MOS is not obvious to obtain on a wide band like 470-860MHz. Software helps now.

 

----------

 

75 ohm is the preferred impedance of a coaxial line because it minimizes the losses. A higher impedance demands a narrower central conductor whose resistance increases, a lower one reduces the wave impedance but this impedance reduces the losses when bigger.

 

At low frequency (when the telephone transmitted at few kHz) magnetic material can reduce the losses by increasing the wave impedance. Impractical in UHF.

 

A smaller dielectric constant also reduces the losses by increasing the wave impedance. Such cables have mainly air as an insulator, and little plastic to hold the central conductor. Their wave impedance can be (not always) 93ohm to keep the optimum diameter ratio.

 

A higher impedance, like 300ohm, is impossible in a coaxial line. It exists for flat ribbons, which are by nature symmetric and unshielded.They can't replace coax cables as is, and unshielded cables are bad for TV, among others because echoes due to reflections on the landscape create image ghosts, which the antennas try hard to remove through directivity, and the cable should not pick up.

 

In other words: excellent reasons made the 75 ohm standard, which is here to stay. Until you put the whole tuner right at the antenna, and then the cable can be an optics fibre if you wish.