masthead amplifier enigma

A few days ago I was asked a question by a young aerial installer, thus:

"When I'm installing an aerial in an area where low signal strength is
the problem and I measure the BER* of a terrestrial digital signal
direct from the aerial I'll get one result, but if I connect a decent
quality low gain masthead amplifier at the point where the meter was,
then feed the amplifier output to the amplifier, I always get a better
result. It doesn't matter which meter I use; I always get that result.
How come? Surely the signal/noise ratio should be roughly the same in
either case."

"Well lad," I said, very slowly filling my pipe in the hope that a
credible answer would come to me. A minute later, with the pipe
overflowing with Auld Carcinogen, I had no real idea of an answer. "Oh
by the way, how's your mother getting on with the new prosthesis?" I
queried in desperation. "I saw her in Aldi on Wednesday and not many
people were staring. A good sign I thought."

"Oh, she's doing well with it Mr Wright. But what about the masthead amp
thing?"

"Err, what kind of masthead amp do you use?" I asked as a delaying tactic.

"I use Proception ones. The 9dB and 16dB ones. So what do you think?"

I drew in a deep draught of Auld Carcinogen and said, "Err well, I know
that what you've found is true, but a lot of people won't accept it, and
say it can't be so. it's a complicated issue, without a simple clear
answer. Just as you can't see me very clearly right now because the room
is full of acrid smoke, the answer to your question is somewhat
obfuscated. However, I can suggest a few things. There must be a
difference in the noise that the signal encounters when it enters the
amplifier, compared to that which it encounters when it enters your meter."

"It's the same with my analyser, and once I took a receiver on the roof
and it was the same with that. Putting an amplifier in front of it
reduced the drop-out a lot."

"Yes yes, I expect it would be the same, more or less. We can only
conclude that the masthead amp contributes less noise than a meter,
analyser, or receiver. Maybe it's because all those things have
processors and power supplies that generate hash. Other than that I
really don't know. I'll tell you what I'll do though. I have some very
knowledgeable friends. I'll put the question to them, and maybe one of
them will wade in with a better answer."

So come on gentlemen. My credibility is at stake here!

Bill

*Bit Error Rate; a measurement of decoding accuracy, so a practical
indication of signal quality.

Bill Wright wrote:
A few days ago I was asked a question by a young aerial installer, thus:

"When I'm installing an aerial in an area where low signal strength is
the problem and I measure the BER* of a terrestrial digital signal
direct from the aerial I'll get one result, but if I connect a decent
quality low gain masthead amplifier at the point where the meter was,
then feed the amplifier output to the amplifier

A typo of course. This should say 'meter'.

Bill

Bill Wright wrote:


"Yes yes, I expect it would be the same, more or less. We can only
conclude that the masthead amp contributes less noise than a meter,

That would be the obvious explanation. The masthead amplifier has a
lower noise figure than the meter, and enough gain to swamp the meter
and cable noise and the cable losses.

There is probably an advantage in building such meters with front ends
matching those of mediocre TV sets, rather than using state of the art
low noise devices.

At UHF frequencies, the dominant noise source tends to be the receiver
itself. The second source will be the thermal noise from the earth,
assuming no local rogue electronics.

In message , David Woolley
writes
Bill Wright wrote:

"Yes yes, I expect it would be the same, more or less. We can only
conclude that the masthead amp contributes less noise than a meter,

That would be the obvious explanation. The masthead amplifier has a
lower noise figure than the meter, and enough gain to swamp the meter
and cable noise and the cable losses.

There is probably an advantage in building such meters with front ends
matching those of mediocre TV sets, rather than using state of the art
low noise devices.

At UHF frequencies, the dominant noise source tends to be the receiver
itself. The second source will be the thermal noise from the earth,
assuming no local rogue electronics.

Yes, it will be because the amplifier has a lower noise figure than the
meter.

The meter will probably have been designed for high dynamic range
(between noise floor and overload). The need to be able to handle strong
signals (as well as low) will usually mean that the meter itself hasn't
got a fantastically low noise figure.

Some general-purpose 'professional' instruments (such as spectrum
analysers) have built-in preamps of (say) 20dB. These can switched 'in'
- or 'out' - as required.

Such preamps have to be used with discretion. As I said, they generally
reduce the dynamic range of the measuring instrument. For example, with
a spectrum analyser, you might need to switch it 'in' when making
low-level measurements (such was when making measurement near the noise
floor of the SA itself). However, when making high-level measurements,
you would usually have to either switch the preamp 'out', or switch in
some additional attenuation for the incoming signal (often requiring an
external attenuator).
--
Ian

On 27/06/2011 02:09, Bill Wright wrote:

So come on gentlemen. My credibility is at stake here!

Perhaps the amp is rejecting out of band noise?

I've experienced similar with a set back amp. They shouldn't work if you
believe the theory but I connected one to the end of a rooftop that was
showing below 70% quality with bad pixillation and the quality jumped to
80% with no drop-outs when I connected it. I can only assume it's
because the amp is selective as to the noise it amplifies along with the
wanted signal and actually improves S/N.

I don't have all the gismos you Rod Hull impersonators have, so I wasn't
able to test it properly.

The Antiference masthead I'm using here at moment says in the literature
that it specifically rejects Tetra, so it maybe that these amps are
pretty good at rejecting anything outside of the band.

Even if the above is total ********, it sounds plausable. Well, it does
to me.

On Mon, 27 Jun 2011 02:09:18 +0100, Bill Wright
wrote:

the problem and I measure the BER* of a terrestrial digital signal
direct from the aerial I'll get one result

I think you have stated the problem here in the first sentence.
It's wrong to consider that the BER is a function *of the signal*.
It's not. The signal itself has the same BER that it had when it left
the transmitter.
The indicated value is a function of the signal received by the
aerial, the in-band and out-of-band noise received by the aerial, and
the bandwidth, noise factor and linearity of the measuring instrument.
Two different types of measuring instrument will inevitably show
different values.

On Mon, 27 Jun 2011 02:09:18 +0100, Bill Wright wrote:

A few days ago I was asked a question by a young aerial installer, thus:

"When I'm installing an aerial in an area where low signal strength is
the problem and I measure the BER* of a terrestrial digital signal
direct from the aerial I'll get one result, but if I connect a decent
quality low gain masthead amplifier at the point where the meter was,
then feed the amplifier output to the amplifier, I always get a better
result. It doesn't matter which meter I use; I always get that result.
How come? Surely the signal/noise ratio should be roughly the same in
either case."

[snip]

So come on gentlemen. My credibility is at stake here!

Bill

*Bit Error Rate; a measurement of decoding accuracy, so a practical
indication of signal quality.

I've definitely seen the same effect with receivers. Putting a cheap and
nasty amplified splitter just behind the receiver can make all the
difference.

It could be down to noise sources in the receiver but my guess is this:

A receiver has a tuned circuit and perhaps other filters, etc. between
its aerial input and the RF amplifier stage. In most locations where
there is a reasonable signal, this is a good thing. It filters out other
stronger signals so they don't overload the amplifier. In a marginal
signal area it's not so good because of losses in these circuits (several
dB?). The untuned coupling to the first stage of an external amplifier is
better even if the amplifier itself has only a mediocre noise figure
(maybe 1 dB below par?).


--
Steve Hayes, South Wales, UK -- please remove colours from address

Ian Jackson wrote:

Some general-purpose 'professional' instruments (such as spectrum
analysers) have built-in preamps of (say) 20dB. These can switched 'in'
- or 'out' - as required.
There's a switchable preamp on my analyser.

Bill

On Monday, June 2011, at 02:09:18h +0100, Bill Wright recounted:

A few days ago I was asked a question by a young aerial installer, thus:

"When I'm installing an aerial in an area where low signal strength is
the problem and I measure the BER* of a terrestrial digital signal
direct from the aerial I'll get one result, but if I connect a decent
quality low gain masthead amplifier at the point where the meter was,
then feed the amplifier output to the amplifier, I always get a better
result.

Is there any substantial difference between a masthead amplifier and
the device Televes call a "Margin Rising Device" (or is that just a
technical term undergone marketing speak), which has the same
beneficial affect on the signal?

In message , Bill Wright
writes
Ian Jackson wrote:

Some general-purpose 'professional' instruments (such as spectrum
analysers) have built-in preamps of (say) 20dB. These can switched
'in' - or 'out' - as required.
There's a switchable preamp on my analyser.

Luxury!
When I were a lad, I had to mak do wi' a home-made preamp, built in a 4
ounce Gold Flake baccy tin!
--
Ian