On 1/24/2014 6:40 AM, wrote:
On Thursday, January 23, 2014 9:45:31 PM UTC-6, Jerry Stuckle wrote:


It is also an issue on HF, especially with less expensive receivers.

You don't see it on the S-Meter because the meter is set such that the

typical receiver noise doesn't show up. But it is there. And it does

affect weak signal intelligibility.

I didn't say that internal receiver noise would show on the
S meter.
I suppose it's possible receiver noise could effect weak signal
reception on HF, but it's going to have to be a really low
level not to override the internal receiver noise on any modern
radio, unless it's some kind of abnormal birdy or whatever.
I just don't see it being a much of a problem on HF, particularly
20m.
If external noise or signals picked up by the antenna don't
override the internal receiver noise on 20m, something is broke
somewhere.
And if the system is that broke, might as well forget anyone hearing
him, particularly running QRP.
He should do OK with a small loop as long as it's built and working
right.



Low antenna efficiency affects not only the received signal, but the
received noise.

And yes, many inexpensive "modern" receivers suffer from poor front
ends. And since transistors (especially bipolar) typically generate
more noise than tubes,
noise can be a worse problem now than in the 60's. You can get low
noise transistors, but these are more expensive.

--
==================
Remove the "x" from my email address
Jerry Stuckle

==================

"Jerry Stuckle" wrote in message
...
And yes, many inexpensive "modern" receivers suffer from poor front
ends. And since transistors (especially bipolar) typically generate more
noise than tubes,
noise can be a worse problem now than in the 60's. You can get low noise
transistors, but these are more expensive.

--

How did ou come up with transistors generate more noise than tubes ?

Most of my research into low noise has been above 50 mhz and the beter tubes
generate much more noise than transistors. The older U310 devices have a
noise figure less than 2 db at 150 mhz and a 6cw4 will have around 3 db.
Cheap gaasfets have noise figuers less than 1 db.

I seem to remember that most tube sets of the old days were stating about .5
uv senstivity on ssb, and many of todays ham trasceivers are way less than
that.



---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com

On 1/24/2014 10:58 AM, Ralph Mowery wrote:
"Jerry Stuckle" wrote in message
...
And yes, many inexpensive "modern" receivers suffer from poor front
ends. And since transistors (especially bipolar) typically generate more
noise than tubes,
noise can be a worse problem now than in the 60's. You can get low noise
transistors, but these are more expensive.

--

How did ou come up with transistors generate more noise than tubes ?


Almost 50 years of experience, including studying both tubes and
transistors in my EE courses back in the 70's. Plus measurements of
both received and transmitted signals, using lab-grade test equipment.

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Back in the 70's, I ran a CAP repeater from my home. Transmit and
receive antennas were separated by about 25' vertically. It was a
surplus Motorola tube rig, running 25W. I was able to run it without
any desense without duplexers. Yes, the channel spacing was 4.25Mhz,
but you can't do that with a transistorized rig.

Most of my research into low noise has been above 50 mhz and the beter tubes
generate much more noise than transistors. The older U310 devices have a
noise figure less than 2 db at 150 mhz and a 6cw4 will have around 3 db.
Cheap gaasfets have noise figuers less than 1 db.


Yes, nowadays, there are transistors with lower noise figures. But they
are relatively expensive, and you won't find them in the less expensive
receivers.

I seem to remember that most tube sets of the old days were stating about .5
uv senstivity on ssb, and many of todays ham trasceivers are way less than
that.


Plus or minus, that is about right. But that wasn't because of the
tubes; they could have done better but it would have required more
amplification and higher cost. Plus with a decent antenna, the
atmospheric noise was higher than that, so there was no need for more
amplification. It would have just been lost in the AGC circuitry.

Even back in the early 70's, commercial tube VHF radios could easily get
..15mv for 20db S+N/N ratio. Not much different than the transistorized
versions today.



---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com




--
==================
Remove the "x" from my email address
Jerry Stuckle

==================

"Jerry Stuckle" wrote in message
...
How did ou come up with transistors generate more noise than tubes ?


Almost 50 years of experience, including studying both tubes and
transistors in my EE courses back in the 70's. Plus measurements of both
received and transmitted signals, using lab-grade test equipment.

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Back in the 70's, I ran a CAP repeater from my home. Transmit and receive
antennas were separated by about 25' vertically. It was a surplus
Motorola tube rig, running 25W. I was able to run it without any desense
without duplexers. Yes, the channel spacing was 4.25Mhz, but you can't do
that with a transistorized rig.


I think you are mixing apples and oranges. For transmiters the tubes
usually have less broad band noise. One reason is not the tube, but the
tuned circuits are much more selective. With the high impedance of the
tubes it is easy to be very selective due to the circuit Q.
For a receiver, it is still all about the noise figuer and having enough
gain (which is not usually a problem) to overcome the noise of the other
parts of the receiver.

Yes, you could run the CAP repeater with seperate antennas with tubes where
you could not with the transistors. As above the circuit selectivity has
alot to do with it. Tube circuits are much more selective when it comes to
broad band noise. Many transistor receivers are broad band in the first few
RF stages. That gives two problems to over come. Broad band noise for the
transmitter (which I am not talking about) and the broad RF stages of the
receiver (Not noise of the transistor/fet but poor selectivity). The old GE
Mastr ll is one of the few that has a fairly narrow front end. I have one
of those on 2 meters.

What do you call relative expensive for a transistor/fet that has a noise
figuer of around 1 db ?

Even in some of the old ARRL repeater handbooks they are putting fet preamps
ahead of the tube receivers.

What tubes are you talking about that has an under 2 db noise figuer at 150
mhz ? I think there was a 417 and maybe a 416 that might make it, but they
were very expensive, especially the 416.



---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com

On 1/24/2014 2:24 PM, Ralph Mowery wrote:

"Jerry Stuckle" wrote in message
...
How did ou come up with transistors generate more noise than tubes ?


Almost 50 years of experience, including studying both tubes and
transistors in my EE courses back in the 70's. Plus measurements of both
received and transmitted signals, using lab-grade test equipment.

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Back in the 70's, I ran a CAP repeater from my home. Transmit and receive
antennas were separated by about 25' vertically. It was a surplus
Motorola tube rig, running 25W. I was able to run it without any desense
without duplexers. Yes, the channel spacing was 4.25Mhz, but you can't do
that with a transistorized rig.


I think you are mixing apples and oranges. For transmiters the tubes
usually have less broad band noise. One reason is not the tube, but the
tuned circuits are much more selective. With the high impedance of the
tubes it is easy to be very selective due to the circuit Q.
For a receiver, it is still all about the noise figuer and having enough
gain (which is not usually a problem) to overcome the noise of the other
parts of the receiver.


No, I am not mixing apples and oranges. Sure, the transmitter tuned
circuits have a higher Q, but that does not affect noise on nearby
frequencies (like 4.25Mhz apart on 2 meters - less than 3% of the
transmitted frequency).

Remember also that receivers also have tuned circuits for input; many of
the older receivers had preselectors to tune the input to the desired
frequency (and these circuits typically had higher Q than transmitter
output circuits).

Yes, you could run the CAP repeater with seperate antennas with tubes where
you could not with the transistors. As above the circuit selectivity has
alot to do with it. Tube circuits are much more selective when it comes to
broad band noise. Many transistor receivers are broad band in the first few
RF stages. That gives two problems to over come. Broad band noise for the
transmitter (which I am not talking about) and the broad RF stages of the
receiver (Not noise of the transistor/fet but poor selectivity). The old GE
Mastr ll is one of the few that has a fairly narrow front end. I have one
of those on 2 meters.


No, tubes themselves generate less noise, especially when running in a
non-linear mode such as Class C. But transistors definitely generate
more noise, as can be identified on a good spectrum analyzer.

What do you call relative expensive for a transistor/fet that has a noise
figuer of around 1 db ?


For manufacturers, anything over a couple of cents per device. But also
GAsFETs are also more susceptible to static charges from the antenna,
requiring additional protective circuitry at the front end.

Even in some of the old ARRL repeater handbooks they are putting fet preamps
ahead of the tube receivers.


Yes, and they also put tube preamps in front of the tube receivers.
FETs were real popular back then, mainly because they weren't tubes -
and didn't have the high power requirements associated. They also were
new, making them ripe for experimentation (quite popular at one time).

What tubes are you talking about that has an under 2 db noise figuer at 150
mhz ? I think there was a 417 and maybe a 416 that might make it, but they
were very expensive, especially the 416.


I don't remember tube numbers any more - that was over 40 years ago, and
I haven't touched a receiver tube circuit in at least 30 years But I
also remember having to design low noise RF circuits - and make them
work. Not easy to do even in the lab; much harder for manufacturers.
And when we were doing solid state amplifiers, it was much harder to get
a great noise figure.

--
==================
Remove the "x" from my email address
Jerry, AI0K

==================

On Fri, 24 Jan 2014 13:04:48 -0500, Jerry Stuckle
wrote:

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Amazing. You might see more hash with a synthesized transistor
xmitter, but for crystal controlled, they noise is quite a bit less
with transistors.

Back in the 70's, I ran a CAP repeater from my home. Transmit and
receive antennas were separated by about 25' vertically. It was a
surplus Motorola tube rig, running 25W. I was able to run it without
any desense without duplexers. Yes, the channel spacing was 4.25Mhz,
but you can't do that with a transistorized rig.

Amazing. These days, 2 meter repeaters do 0.600 MHz spacing using all
transistor equipment, a single antenna, and a notch type duplexer.

Yes, nowadays, there are transistors with lower noise figures. But they
are relatively expensive, and you won't find them in the less expensive
receivers.

Rubbish. pHEMT devices are available with 0.75dB NF at 1GHz for under
$3/ea. For example:
http://www.minicircuits.com/pdfs/PSA4-5043+.pdf
Note that this isn't just a single device but a MMIC amplifier capable
of belching 100 mw (+20dBm) with 1dB gain compression.

I seem to remember that most tube sets of the old days were stating about .5
uv senstivity on ssb, and many of todays ham trasceivers are way less than
that.

That depends on the frequency. The atmospheric noise is so high on
the lower HF bands that improvments in receiver sensitivity simply
results in amplifying both the noise and the signal but the same
amount, resulting in no net improvement in signal to noise ratio
(SNR). Adding more gain also decreases the receiver dynamic range
because at high signal levels, the added gain will cause the receiver
to overload at a lower signal level. Hint: Use only as much gain as
necessary and no more.

Plus or minus, that is about right. But that wasn't because of the
tubes; they could have done better but it would have required more
amplification and higher cost. Plus with a decent antenna, the
atmospheric noise was higher than that, so there was no need for more
amplification. It would have just been lost in the AGC circuitry.

Mostly I agree except for the part about AGC. If your receiver is
already into AGC for weak signal conditions, you're effectively
reducing the receiver sensitivity at the same time. The way AGC is
suppose to work for a SSB receiver is that the AGC starts just above
the level where you can hear an intelligible signal. 12dB SINAD is
about right. If the AGC were perfect (i.e. no slope), then any
increase in signal level above that point will result in no
improvement in SNR because the AGC will do its best to keep the SNR
constant. With a real AGC (dual slope, controlled attack and release
time) the SNR improves somewhat as the input level increases until it
reaches some SNR, where it levels off. I think this is called
"ultimate SNR" or something similar, which is just the SNR of a very
strong receive signal.

Even back in the early 70's, commercial tube VHF radios could easily get
.15mv for 20db S+N/N ratio. Not much different than the transistorized
versions today.

I think you might mean 0.15 uv/12dB SINAD. A 0.15mv receiver would be
considere comatose.

In the 1960's thru about 1983, I was involved in various radio
service, radio manufacture, radio sales, and radio consulting
companies. I had plenty of experience with everything from wideband
GE Pre-Prog thru cellular radios including tubes. I never saw a tube
receiver with 0.15uv sensitivity. Photos of the shop and various
repeaters. Most were UHF.
http://802.11junk.com/jeffl/pics/Old...ers/index.html
The main site was on Santiago Pk:
http://www.trabucooutdoors.com/assets/images-1/odds_ends/santiagopk.jpg
The mess on the far right is the antenna farm.

9 GE Progress Line repeaters with Alpha tone panels (which I helped
design).
http://802.11junk.com/jeffl/pics/Old%20Repeaters/slides/Santiago-01.html
See anything missing? There are no duplexers. There was one receive
antenna and 9 transmit antennas. All tubes. With a typically
1uv/12dB SINAD receiver, the isolation was sufficient. Don't ask
about the tx intermod, which was horrible.

Typically, these tube type repeaters would start out with about
0.5uv/12dB SINAD with new tubes. That's measured directly into the
receiver input with no additional filters. After about 6 months, the
sensitivity would settle down to about 0.75 to 0.90uv and stay there
for about 6 to a year, when it needed retuning. This was using a 6AM4
triode. The only time I saw better sensitivity with tubes was when
someone tweaked the audio freq response, or excessively narrowed the
IF bandwidth.

Somewhat later, in the 1970's, I found myself designing marine radios.
Typical VHF sensitivity was about 0.25uv/12dB SINAD using a dual gate
MOSFET front end such as a 40673 or 3N212. We ocassionally used
JFET's such as a U310 but the sensitivity was about the same.

The problem is that 0.15uv is just too close to the receiver noise
floor to be realizeable. With a 25KHz receive bandwidth:
noise floor = -174dBm/Hz + 10*log(25KHz)
= -174 + 10*4.40 = -174 + 44
= -130 dBm or 0.071 uv
To obtain a 0.15 uv sensitivity, you would need a receiver noise
figure plus a detection SNR of less than:
10 log(0.15/0.071) = 3.3dB
With an analog FM demodulator, that's barely possible and usually
requires a perfect noise-free front end. However, with a 0.25uv/12dB
SINAD sensitivity, there's 5.5dB of margin, which is more than enough
for real receivers.

Mo
http://www.r-390a.net/Receiver-Specifications-Explaned.pdf



--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

"Jeff Liebermann" wrote in message
...
On Fri, 24 Jan 2014 13:04:48 -0500, Jerry Stuckle
wrote:

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Amazing. You might see more hash with a synthesized transistor
xmitter, but for crystal controlled, they noise is quite a bit less
with transistors.

Back in the 70's, I ran a CAP repeater from my home. Transmit and
receive antennas were separated by about 25' vertically. It was a
surplus Motorola tube rig, running 25W. I was able to run it without
any desense without duplexers. Yes, the channel spacing was 4.25Mhz,
but you can't do that with a transistorized rig.

Amazing. These days, 2 meter repeaters do 0.600 MHz spacing using all
transistor equipment, a single antenna, and a notch type duplexer.



What you are saying is more in my line of thinking and limiated expierance.
Around 30 years ago I had an Ameco 2 meter receive converter that used the
6ds4 nuvistors. Probably the best tube that most could afford. Tuned for
the best signal, I could still improve it when adding a u310 preamp. Not
sure how much as I did not have very good test equipment, but noticiable by
ear. I think many of the old sets used a 6ak5 for the rf amp.

Transmitting noise I don't know. All I was looking at was the noise figure
for the receiver as that was the origional topic.

What Jerry was talking about was a tube CAP repeater with seperate antennas.
If the receiver was around .5 uv or worse and the transmitter was cleaner he
could operate with seperate antennas . He said he could do that with the
tubes but not the transistors which I believe. Beter selectivity on the
transmitter and receiver than some transistor repeaters.

I do have a 2 meter repeater on the air with 600 khz seperation. Solid
state and 100 watts. Right now it has a Dow East Microwave phet preamp on
it. Don't recall the exect sensitivy for 12 db sinad but it is under .2 uv
as shown on my hp8924c.
No desense is detected. It does have a 6 cavity bpbr duplexer with the high
selectivity option.



---
This email is free from viruses and malware because avast! Antivirus protection is active.
http://www.avast.com

On Fri, 24 Jan 2014 20:00:52 -0500, "Ralph Mowery"
wrote:

Around 30 years ago I had an Ameco 2 meter receive converter that used the
6ds4 nuvistors. Probably the best tube that most could afford.

I built a few preamps using a 6CW4, 6DS4 or an 8056. The Nuvistors
were about the best commerical low noise receive tubes available at
the time. The 8056 was rated at 4.5dB NF at 200 MHz and 16.4dB of
gain.
http://www.ko4bb.com/Manuals/RCA/RCA_8056_Nuvistor_AN-195.pdf
See the two graphs. A 4.5dB NF was about typical for a good receive
tube.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On 1/24/2014 8:00 PM, Ralph Mowery wrote:

"Jeff Liebermann" wrote in message
...
On Fri, 24 Jan 2014 13:04:48 -0500, Jerry Stuckle
wrote:

The easiest way of seeing it is looking at the output of both tube and
transistorized transmitters on a spectrum analyzer. You will see much
more hash on the transistorized transmitter.

Amazing. You might see more hash with a synthesized transistor
xmitter, but for crystal controlled, they noise is quite a bit less
with transistors.

Back in the 70's, I ran a CAP repeater from my home. Transmit and
receive antennas were separated by about 25' vertically. It was a
surplus Motorola tube rig, running 25W. I was able to run it without
any desense without duplexers. Yes, the channel spacing was 4.25Mhz,
but you can't do that with a transistorized rig.

Amazing. These days, 2 meter repeaters do 0.600 MHz spacing using all
transistor equipment, a single antenna, and a notch type duplexer.



What you are saying is more in my line of thinking and limiated expierance.
Around 30 years ago I had an Ameco 2 meter receive converter that used the
6ds4 nuvistors. Probably the best tube that most could afford. Tuned for
the best signal, I could still improve it when adding a u310 preamp. Not
sure how much as I did not have very good test equipment, but noticiable by
ear. I think many of the old sets used a 6ak5 for the rf amp.

Transmitting noise I don't know. All I was looking at was the noise figure
for the receiver as that was the origional topic.

What Jerry was talking about was a tube CAP repeater with seperate antennas.
If the receiver was around .5 uv or worse and the transmitter was cleaner he
could operate with seperate antennas . He said he could do that with the
tubes but not the transistors which I believe. Beter selectivity on the
transmitter and receiver than some transistor repeaters.

I do have a 2 meter repeater on the air with 600 khz seperation. Solid
state and 100 watts. Right now it has a Dow East Microwave phet preamp on
it. Don't recall the exect sensitivy for 12 db sinad but it is under .2 uv
as shown on my hp8924c.
No desense is detected. It does have a 6 cavity bpbr duplexer with the high
selectivity option.


Receiver sensitivity was 0.2 mv for 20db S+N/N ratio. I don't know
how much better; the surplus signal generator I was using wasn't that
accurate.

And BTW - 'm' can also mean micro, especially when you don't have a
Greek alphabet available. 'u' is not the same as the Greek 'mu' and can
be confusing. Of course, using 'm' for both milli and micro can be
confusing, unless you know the context.

--
==================
Remove the "x" from my email address
Jerry Stuckle

==================

On Fri, 24 Jan 2014 20:29:22 -0500, Jerry Stuckle
wrote:

Receiver sensitivity was 0.2 mv for 20db S+N/N ratio.

I'll assume that you mean 0.2 microvolts, not millivolts.
Also, nobody has uses 20dB S+N/N since about the 1960's except maybe
the ARRL[1]. 12dB SINAD has been around since about the mid 1960's.
http://en.wikipedia.org/wiki/SINAD

I don't know
how much better; the surplus signal generator I was using wasn't that
accurate.

At the time, I used a Gertsch FM-something. It's the box with only
the left handle showing at the extreme right of the pictu
http://802.11junk.com/jeffl/pics/Old%20Repeaters/slides/PMC02.html
It wasn't every accurate, and not really intended to be used as a
signal source. However, I had a precision step attenuator, and since
the Gertsch factory was near my parents house, it wasn't too difficult
to get it calibrated in trade for some grunt work.

And BTW - 'm' can also mean micro, especially when you don't have a
Greek alphabet available. 'u' is not the same as the Greek 'mu' and can
be confusing. Of course, using 'm' for both milli and micro can be
confusing, unless you know the context.

If you follow SI standards, the "m" means milli.
http://physics.nist.gov/cuu/Units/prefixes.html

If you have a PC keyboard, try using the key combination:
alt0181
on the number pad which produces a µ symbol.
http://symbolcodes.tlt.psu.edu/accents/codealt.html


[1] It took me about 10 years to stop using micro-microfarads and
switch to picofarads. I suppose there might be some dinosaurs roaming
the earth still talking about "20 dB quieting".

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558