Sony STRDH590 antenna

On Thursday, February 20, 2020 at 7:05:26 PM UTC-5, upsid...@downunder.com wrote:
On Thu, 20 Feb 2020 12:19:09 -0800 (PST), Michael Terrell wrote:

On Thursday, February 20, 2020 at 1:18:35 PM UTC-5, upsid...@downunder.com wrote:
On Thu, 20 Feb 2020 07:14:05 -0800 (PST), Michael Terrell wrote:


These days, when the IF gain is
very cheap, there is enough IF gain that the band noise or receiver
front end noise will drive the last IF into saturation. On stronger
signals, the last few IF stages will go into saturation, not just the
last. Thus, effectively, this limiting action works like "AGC"
delivering a few volt constant voltage to the FM detector.

You only want limiting in the final IF amp, not every stage.


The important thing is that all selectivity filtering must be done
prior to the first IF stage that would start to saturate on very
strong signals. This is not much of an issue with ceramic filters.
Just make sure that the RF-amplifier and mixer gain in front of the
ceramic filter has enough gain to mask the filter losses.

The situation was different in the tube era with distributed LC
bandpass IF transformers between each amplifying tube, so the final
selectivity was achieved just at the last IF stage.

Try it, you might even like it! :)

But why would I use AGC for the RF amplifier ? Is the mixer so weak
that it is overloaded by a strong signal amplified by the RF
amplifier? Use a stronger mixer so that full RF gain can be applied
before the mixer. A strong mixer can be a problem with low power
battery powered equipment, but using a weak mixer in a mains powered
receiver is simply incompetence.


AGC in the front end prevents overdriving the mixer in areas with excessive RF, and producing Intermod.

If you use AGC to reduce front end gain due to some strong signals in
some nearby frequency, it will also reduce the gain for a weak signal
that you want to listen, thus the weak signal may be swamped in front
end noise when a strong nearby signal is present. A sharp filter
between RF stage and mixer helps getting rid of some strong signals.
Since the filter needs to be tunable, realistically the bandwidth is
at least 1 MHz (Q=100), so it helps only in reducing strong signals a
few MHz away.


A few MHz? It could be 200KHz away, which a single FM channel in the US.

The adjacent channel filtering is done in the IF filter. A capacitance
diode tuned filter between the RF amplifier and mixer will pass 1-2
MHz to the mixer. The mixer _must_ handle any strong signals 200 kHz
away, which is then handled by the IF filter. If the RF amp/mixer
can't handle that strong signal without gain reduction, the receiver
can't be used to receive a very weak signal only 200 kHz away from a
strong signal, no mater what fancy IF filtering is used.

A gain reduction in the RF amp will degrade the noise figure and it is
no longer capable of masking the noise of subsequent stages. In a
good receiver design, the total receiver noise figure is determined by
the noise figure of the first RF stage _provided_ that the gain is
high enough to mask the noise of the following stages.

Show me any FM tuner without a tuned RF stage.

I said nothing about omitting the RF tuned filters, but they are much
wider than the channel spacing. If some dynamic range claims are
given, look carefully at what frequency separation is used. Most
likely a few MHz is used, i.e. outside the RF filter bandwidth.

Of course, if you have a fixed tuned receiver, you could use a 1/4
wave filter with Q=500, but these are big (75 cm) and hard to tune :)
This would have a 200 kHz RF bandwidth.

I still maintain that applying AGC to the first RF stage is a bad idea
and a receiver design requiring it is a suspect, possibly crap design.

Some equipment might have no AGC, but that would be the $20 portable crap.I have worked with this most of my life, both broadcast and deep Space Telemetry.

In space applications, you usually have a fixed gain LNA at the focal
point to compensate for feeder losses as well as receiver added noise
figure. In such case, reducing the indoor unit RF stage gain is not a
very bad thing.

Television also uses AGC in the RF. All of these use the same AGC in the IF and RF stages.

Analog requires accurate AGC so that the synch pulse levels are
normalized after the diode detector. Feeding AGC into multiple stages
helps keeping the level constant. In a TV receiver, the needed IF gain
is quite low, thus only a few stages are used and you may have to use
AGC on all IF stages as well as on the RF stage.

I serviced CATV Hetrodyne signal processors, including one crap brand that used all PNP RF transistors in the signal path. That isn't a problem other than having to buy the 1000 at a time. Luckily, that company went bankrupt. I serviced hundreds of designs and all used AGC for RF and IF.

That must have been a very old design :), from the old times when
gain and selectivity distribution was not well understood .

You've never built a Telemetry system that had to follow a satellite from the ground, until it was in space, where there wasn't enough land to provide decent physical seperation. One of our turnkey systems for the ESA was more than 40 dB too hot for the LNA. The only solution was a mechanical shutter that opened, as the signal level dropped. Guess how that was controlled. We didn't have to do that to the portable earth station, since it was parked about 20 miles away. These systems were for a contract of just under a half million dollars, not some stereo system. Our company was consider the top in the industry at that time.
 
On Thu, 20 Feb 2020 19:21:57 -0800 (PST), Michael Terrell
<terrell.michael.a@gmail.com> wrote:

I serviced CATV Hetrodyne signal processors, including one crap brand that used all PNP RF transistors in the signal path. That isn't a problem other than having to buy the 1000 at a time. Luckily, that company went bankrupt. I serviced hundreds of designs and all used AGC for RF and IF.

That must have been a very old design :), from the old times when
gain and selectivity distribution was not well understood .


You've never built a Telemetry system that had to follow a satellite from the ground, until it was in space, where there wasn't enough land to provide decent physical seperation.

So this was a full duplex link ?

What was the frequency separation between Tx and Rx ?

On the satellite, what was the physical R/Tx separation at the
antenna(s) ? I guess it is much smaller than the separation between
ground based Tx/Rx antennas.

When you have to use the same antenna for both Rx and Tx, you of
course need a notch filter against the Tx frequency inserted in front
of the Rx chain. In most cases, you also need a notch tuned at the Rx
frequency inserted into the _Tx_ chain to reduce the Tx phase noise
falling on the Rx frequency. A larger Rx/Tx frequency separation also
helps.


> One of our turnkey systems for the ESA was more than 40 dB too hot for the LNA. The only solution was a mechanical shutter that opened, as the signal level dropped. Guess how that was controlled. We didn't have to do that to the portable earth station, since it was parked about 20 miles away. These systems were for a contract of just under a half million dollars, not some stereo system. Our company was consider the top in the industry at that time.
 
On Fri, 21 Feb 2020 02:05:23 +0200, upsidedown@downunder.com wrote:

On Thu, 20 Feb 2020 12:19:09 -0800 (PST), Michael Terrell
terrell.michael.a@gmail.com> wrote:

On Thursday, February 20, 2020 at 1:18:35 PM UTC-5, upsid...@downunder.com wrote:
On Thu, 20 Feb 2020 07:14:05 -0800 (PST), Michael Terrell wrote:


These days, when the IF gain is
very cheap, there is enough IF gain that the band noise or receiver
front end noise will drive the last IF into saturation. On stronger
signals, the last few IF stages will go into saturation, not just the
last. Thus, effectively, this limiting action works like "AGC"
delivering a few volt constant voltage to the FM detector.

The important thing is that all selectivity filtering must be done
prior to the first IF stage that would start to saturate on very
strong signals. This is not much of an issue with ceramic filters.
Just make sure that the RF-amplifier and mixer gain in front of the
ceramic filter has enough gain to mask the filter losses.

The situation was different in the tube era with distributed LC
bandpass IF transformers between each amplifying tube, so the final
selectivity was achieved just at the last IF stage.

Try it, you might even like it! :)

But why would I use AGC for the RF amplifier ? Is the mixer so weak
that it is overloaded by a strong signal amplified by the RF
amplifier? Use a stronger mixer so that full RF gain can be applied
before the mixer. A strong mixer can be a problem with low power
battery powered equipment, but using a weak mixer in a mains powered
receiver is simply incompetence.


AGC in the front end prevents overdriving the mixer in areas with excessive RF, and producing Intermod.

If you use AGC to reduce front end gain due to some strong signals in
some nearby frequency, it will also reduce the gain for a weak signal
that you want to listen, thus the weak signal may be swamped in front
end noise when a strong nearby signal is present. A sharp filter
between RF stage and mixer helps getting rid of some strong signals.
Since the filter needs to be tunable, realistically the bandwidth is
at least 1 MHz (Q=100), so it helps only in reducing strong signals a
few MHz away.


A few MHz?t could be 200KHz away, which a single FM channel in the US.

The adjacent channel filtering is done in the IF filter. A capacitance
diode tuned filter between the RF amplifier and mixer will pass 1-2
MHz to the mixer. The mixer _must_ handle any strong signals 200 kHz
away, which is then handled by the IF filter. If the RF amp/mixer
can't handle that strong signal without gain reduction, the receiver
can't be used to receive a very weak signal only 200 kHz away from a
strong signal, no mater what fancy IF filtering is used.

A gain reduction in the RF amp will degrade the noise figure and it is
no longer capable of masking the noise of subsequent stages. In a
good receiver design, the total receiver noise figure is determined by
the noise figure of the first RF stage _provided_ that the gain is
high enough to mask the noise of the following stages.

Show me any FM tuner without a tuned RF stage.

I said nothing about omitting the RF tuned filters, but they are much
wider than the channel spacing. If some dynamic range claims are
given, look carefully at what frequency separation is used. Most
likely a few MHz is used, i.e. outside the RF filter bandwidth.

This is most easily seen with third order input intercept point (IIP3)
measurements. The 2f1 - f2 intermodulation product is nasty, since it
can fall directly on a weak signal you want to measure. If the
offending signals f1 and f2 are a few MHz away from the desired Rx
frequency, a tuned section in front of the mixer will take care of
them, but if the offending signals are within 1 MHz from the desired
signal, both can pass the RF filter generating intermodulation
products which fall on the IF frequency and thus also pass the IF
filter. Thus the IIP3 figures with a small separation of offending is
much worse compared to greater separation.
 
On Friday, February 21, 2020 at 1:37:48 AM UTC-5, upsid...@downunder.com wrote:
On Thu, 20 Feb 2020 19:21:57 -0800 (PST), Michael Terrell wrote:

I serviced CATV Hetrodyne signal processors, including one crap brand that used all PNP RF transistors in the signal path. That isn't a problem other than having to buy the 1000 at a time. Luckily, that company went bankrupt. I serviced hundreds of designs and all used AGC for RF and IF.

That must have been a very old design :), from the old times when
gain and selectivity distribution was not well understood .


You've never built a Telemetry system that had to follow a satellite from the ground, until it was in space, where there wasn't enough land to provide decent physical seperation.

So this was a full duplex link?

No, it wasn't. It was for Telemetry, not communications. Communications was provided by other equipment that was built in Europe.

What was the frequency separation between Tx and Rx ?

On the satellite, what was the physical R/Tx separation at the
antenna(s) ? I guess it is much smaller than the separation between
ground based Tx/Rx antennas.

When you have to use the same antenna for both Rx and Tx, you of
course need a notch filter against the Tx frequency inserted in front
of the Rx chain. In most cases, you also need a notch tuned at the Rx
frequency inserted into the _Tx_ chain to reduce the Tx phase noise
falling on the Rx frequency. A larger Rx/Tx frequency separation also
helps.

Sigh. My high school's Amateur Radio club built a two meter repeater in the late '60s. That included the required filters. I was a Broadcast engineer where we combined Aural and Visual transmitters to a single antenna. Have you ever worked with one that would handle 200KW at the old Analog Ch55?
 
On Friday, February 21, 2020 at 3:54:50 AM UTC-5, upsid...@downunder.com wrote:
On Fri, 21 Feb 2020 02:05:23 +0200, upsidedown wrote:

On Thu, 20 Feb 2020 12:19:09 -0800 (PST), Michael Terrell wrote:

On Thursday, February 20, 2020 at 1:18:35 PM UTC-5, upsid...@downunder.com wrote:
On Thu, 20 Feb 2020 07:14:05 -0800 (PST), Michael Terrell wrote:


These days, when the IF gain is
very cheap, there is enough IF gain that the band noise or receiver
front end noise will drive the last IF into saturation. On stronger
signals, the last few IF stages will go into saturation, not just the
last. Thus, effectively, this limiting action works like "AGC"
delivering a few volt constant voltage to the FM detector.

The important thing is that all selectivity filtering must be done
prior to the first IF stage that would start to saturate on very
strong signals. This is not much of an issue with ceramic filters.
Just make sure that the RF-amplifier and mixer gain in front of the
ceramic filter has enough gain to mask the filter losses.

The situation was different in the tube era with distributed LC
bandpass IF transformers between each amplifying tube, so the final
selectivity was achieved just at the last IF stage.

Try it, you might even like it! :)

But why would I use AGC for the RF amplifier ? Is the mixer so weak
that it is overloaded by a strong signal amplified by the RF
amplifier? Use a stronger mixer so that full RF gain can be applied
before the mixer. A strong mixer can be a problem with low power
battery powered equipment, but using a weak mixer in a mains powered
receiver is simply incompetence.


AGC in the front end prevents overdriving the mixer in areas with excessive RF, and producing Intermod.

If you use AGC to reduce front end gain due to some strong signals in
some nearby frequency, it will also reduce the gain for a weak signal
that you want to listen, thus the weak signal may be swamped in front
end noise when a strong nearby signal is present. A sharp filter
between RF stage and mixer helps getting rid of some strong signals.
Since the filter needs to be tunable, realistically the bandwidth is
at least 1 MHz (Q=100), so it helps only in reducing strong signals a
few MHz away.


A few MHz?t could be 200KHz away, which a single FM channel in the US.

The adjacent channel filtering is done in the IF filter. A capacitance
diode tuned filter between the RF amplifier and mixer will pass 1-2
MHz to the mixer. The mixer _must_ handle any strong signals 200 kHz
away, which is then handled by the IF filter. If the RF amp/mixer
can't handle that strong signal without gain reduction, the receiver
can't be used to receive a very weak signal only 200 kHz away from a
strong signal, no mater what fancy IF filtering is used.

A gain reduction in the RF amp will degrade the noise figure and it is
no longer capable of masking the noise of subsequent stages. In a
good receiver design, the total receiver noise figure is determined by
the noise figure of the first RF stage _provided_ that the gain is
high enough to mask the noise of the following stages.

Show me any FM tuner without a tuned RF stage.

I said nothing about omitting the RF tuned filters, but they are much
wider than the channel spacing. If some dynamic range claims are
given, look carefully at what frequency separation is used. Most
likely a few MHz is used, i.e. outside the RF filter bandwidth.

This is most easily seen with third order input intercept point (IIP3)
measurements. The 2f1 - f2 intermodulation product is nasty, since it
can fall directly on a weak signal you want to measure. If the
offending signals f1 and f2 are a few MHz away from the desired Rx
frequency, a tuned section in front of the mixer will take care of
them, but if the offending signals are within 1 MHz from the desired
signal, both can pass the RF filter generating intermodulation
products which fall on the IF frequency and thus also pass the IF
filter. Thus the IIP3 figures with a small separation of offending is
much worse compared to greater separation.

There were sites where a tunable trap with steep skirts was needed to attenuate an unwanted signal. That included schools when their local Cable TV company decided to scramble Ch3 with a strong CW signal in the guard band. Available MATV amps al overloaded, and wiped out the channels used for Educational TV. I designed and built a seven stage trap that didn't cause ripple in the level across the available channels.

I had no access to a spectrum analyzer so they were aligned with a signal generator and a 1n*@ UHF mixer diode as a detector. I also had to use a pair of 50 to 72 ohm transformers for impedance matching. The resulting notch was clean enough for them to feed the Ch 3 modulator into the system, and eliminate the switching between CATV and local. Physical components didn't exist for a 75 Ohm trap, so it was transformed to were it could be built. used glass piston trimmers, and a heavy, diecast Pomona box. Sheet bras was used to divide the interior into compartments, and to improve the isolation between filters. The seams were all soldered, including the covers.
 
On Fri, 21 Feb 2020 08:10:23 -0800 (PST), Michael Terrell
<terrell.michael.a@gmail.com> wrote:

You've never built a Telemetry system that had to follow a satellite from the ground, until it was in space, where there wasn't enough land to provide decent physical seperation.

So this was a full duplex link?

No, it wasn't. It was for Telemetry, not communications. Communications was provided by other equipment that was built in Europe.

So what was your problem ?

It sounds that you had a bad gain and selectivity distribution in the
Rx chain ?

What was the frequency separation between Tx and Rx ?

On the satellite, what was the physical R/Tx separation at the
antenna(s) ? I guess it is much smaller than the separation between
ground based Tx/Rx antennas.

When you have to use the same antenna for both Rx and Tx, you of
course need a notch filter against the Tx frequency inserted in front
of the Rx chain. In most cases, you also need a notch tuned at the Rx
frequency inserted into the _Tx_ chain to reduce the Tx phase noise
falling on the Rx frequency. A larger Rx/Tx frequency separation also
helps.


Sigh. My high school's Amateur Radio club built a two meter repeater in the late '60s. That included the required filters.

Using a two meter repeater with a common Rx/Tx antenna and 600 kHz
separation through a duplexer is standard practice, three cavities in
the Tx chain and three cavities in the Rx chain is enough.

Things got complicated due to multiple FM BC transmitters feeding a
common FM BC antenna a few meter from our 2 m repeater antenna. Most
problems were due to intermodulation products due to rusty bolts in
the nearby structures. Removing those obsolete structures helped a lot
for a few years. Unfortunately two FM BC stations were assigned
frequencies 600 kHz from each other, Unfortunately the third order
mixing product of those two FN BC signals and repeater Tx frequency
fell directly on the repeater Rx frequency, making the repeater
receiver more or less useless.

> AI was a Broadcast engineer where we combined Aural and Visual transmitters to a single antenna. Have you ever worked with one that would handle 200KW at the old Analog Ch55?

Rusty bolts will produce a lot of intermodulation products on sum and
difference frequencies :).
 
On Friday, February 21, 2020 at 3:57:06 PM UTC-5, upsid...@downunder.com wrote:
On Fri, 21 Feb 2020 08:10:23 -0800 (PST), Michael Terrell wrote:


You've never built a Telemetry system that had to follow a satellite from the ground, until it was in space, where there wasn't enough land to provide decent physical separation.

So this was a full duplex link?

No, it wasn't. It was for Telemetry, not communications. Communications was provided by other equipment that was built in Europe.

So what was your problem ?

Sigh. Being under a half mile from the launch site drove the LNA into saturation.


It sounds that you had a bad gain and selectivity distribution in the
Rx chain?

The receivers were tripe conversion, and had a LINEAR 0 to +5 AGC system allow an accurate Diversity system that could track to under 1/10dB, and the output level was adjustable over a 63dB range in 1/10dB steps. You can't seem to wrap your mind around the fact that agencies were lined up to buy these at around $20,000 each. They filed two rack spaces per Receiver or Combiner. NOAA, NASA and the ESA bought them, and one receiver was built to operate off 48VDC for the ISS.



What was the frequency separation between Tx and Rx ?

On the satellite, what was the physical R/Tx separation at the
antenna(s) ? I guess it is much smaller than the separation between
ground based Tx/Rx antennas.

When you have to use the same antenna for both Rx and Tx, you of
course need a notch filter against the Tx frequency inserted in front
of the Rx chain. In most cases, you also need a notch tuned at the Rx
frequency inserted into the _Tx_ chain to reduce the Tx phase noise
falling on the Rx frequency. A larger Rx/Tx frequency separation also
helps.


Sigh. My high school's Amateur Radio club built a two meter repeater in the late '60s. That included the required filters.

Using a two meter repeater with a common Rx/Tx antenna and 600 kHz
separation through a duplexer is standard practice, three cavities in
the Tx chain and three cavities in the Rx chain is enough.

Things got complicated due to multiple FM BC transmitters feeding a
common FM BC antenna a few meter from our 2 m repeater antenna. Most
problems were due to intermodulation products due to rusty bolts in
the nearby structures. Removing those obsolete structures helped a lot
for a few years. Unfortunately two FM BC stations were assigned
frequencies 600 kHz from each other, Unfortunately the third order
mixing product of those two FN BC signals and repeater Tx frequency
fell directly on the repeater Rx frequency, making the repeater
receiver more or less useless.

I was a Broadcast engineer where we combined Aural and Visual transmitters to a single antenna. Have you ever worked with one that would handle 200KW at the old Analog Ch55?

Rusty bolts will produce a lot of intermodulation products on sum and
difference frequencies :).

Of course. Even rusty barbed wire has crude semiconductors at their joints. Rusty tower bolts can also weaken to the point that they fracture and are ejected from a tower leg.

The tower at that TV station had two UHF TV stations, Eight, 50 KW FM transmitters. A trunking radio system with a large room crammed full of Motorola base stations, and a cluster of Government repeaters. The only named agency was the U.S. Forestry service. There was plenty of isolation on that 1700' tower. The lowest antenna was the FM antenna for the FM stations. It was at 1000 feet above ground.


Cracked waveguide with 185KW of RF can cause nasty burns, too. One tower worked received a long, narrow burn on his ass from just bumping int the corner of that several ton piece of brass.
 
On Saturday, February 22, 2020 at 12:26:46 AM UTC-5, Michael Terrell wrote:

> Cracked waveguide with 185KW of RF can cause nasty burns, too. One tower worked received a long, narrow burn on his ass from just bumping int the corner of that several ton piece of brass.

Hummmm.... a trip down memory lane.
Sounds like an Alan-Dick antenna.
 
On Saturday, February 22, 2020 at 5:02:09 PM UTC-5, mpm wrote:
On Saturday, February 22, 2020 at 12:26:46 AM UTC-5, Michael Terrell wrote:

Cracked waveguide with 185KW of RF can cause nasty burns, too. One tower worked received a long, narrow burn on his ass from just bumping int the corner of that several ton piece of brass.

Hummmm.... a trip down memory lane.
Sounds like an Alan-Dick antenna.

It was the big stick in Orange City, Florida :)
 
Jasen Betts <jasen@xnet.co.nz> writes:

There is a *tiny* one-pin Molex socket on the back. You get a
wire with a mating Molex in the same bag as the remote control
AAA batteries and such.

It kind of looks like 2 pins in this photo.
https://images-na.ssl-images-amazon.com/images/I/71yharzgGgL._AC_SL1500_.jpg
(on this page)
https://www.amazon.com/Sony-STR-DH790-7-2-ch-Receiver-Vision/dp/B079YW5VQL/ref=pd_sbs_23_t_0/146-2719398-7440836?_encoding=UTF8&pd_rd_i=B079YW5VQL&pd_rd_r=bf1d57f7-a5f7-4792-80f6-46d34ff4d9b5&pd_rd_w=dtda1&pd_rd_wg=nNYpn&pf_rd_p=5cfcfe89-300f-47d2-b1ad-a4e27203a02a&pf_rd_r=BTE46AHTCJ049ZBR9SSB&psc=1&refRID=BTE46AHTCJ049ZBR9SSB

It is a 2 position Molex but the is only one I could see on the receptacle.
There is only one pin in the plug; the one with the wire attached.

it seems the the connector on the antenna cable has only one position
populated though which hints that the other pin may be a ground
usable for for a 50 or 75 ohm antenna, unless the socket itself has
only one pin fitted.

Only 1.

https://www.sony.com/electronics/support/res/manuals/4726/55cd375e0f9e7103f2b6baa747ed9072/47269061M.pdf
page 25.

--
Jasen.
--
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close..........................
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433
 

Welcome to EDABoard.com

Sponsor

Back
Top