On Sat, 17 Feb 2007 15:49:51 -0600, "Pete KE9OA"
wrote in
:


Sounds like it might be a little too hot for the band.....

I am not sure what you mean.


I have found that it's sometimes better to DE-sensitize the receiver
front-end on a noisy band. Tends to improve S/N.


I do have the service
manual, but I have never seen this kind of problem.


Never assume the manual is correct. Establish a baseline by injecting
a single tone post-AGC, run it up to 100% and measure. Then re-align
the AGC to your mic according to that baseline.

I might inject the signal right at the audio amplifier and work backwards.


Same difference.


For testing purposes, I am using a CT Systems Model 3000B Communications
Test Set. For the brief time that Tx audio is present, the modulation
envelope looks fine. This unit has a built in 100W dummy load.
RF output is flat across the band, at 4.5W, so it doesn't look like
somebody
tried to align the exciter/output stages.


Sounds like it may have an RF limiter. If you have a pulse generator,
hit the audio with 0.1ms spikes. That will also tell you if the audio
limiter has been disabled.

I'll give that at ry.


It looks like I will have to do some signal tracing with an o'scope.
Hopefully, I will be able to find that 5-pin mic plug so I can make up a
test cable. Thanks again for the advice!


No problem. If you ever get around to scanning the manual I wouldn't
mind having a copy.

I will do that. I only can scan up to 11 X 14 size, so the schematics will
be broken pages,but that should be fine.


No hurry. I only get a couple CB radios a year, so it's hard to tell
if or when I'll run across this model. There really should be a public
CB manual repository like BAMA..... maybe some day.

That sounds good.............I will probably scan the service info in the
next couple of weeks. now, about the receiver desensing.....I don't have any
problems with IMD in the receiver, and a diode attenuator is used in the
front end. I have found that IMD performance can be improved by changing
that general purpose switching diode in the front to a 1N5767 PIN diode. I
wanted to get started on this project this afternoon, but a friend stopped
over. I'll give it a whirl tonight. It will be quite a nice radio, once I
get through with it.

"Frank Gilliland" wrote in message
news
On Sat, 17 Feb 2007 15:49:51 -0600, "Pete KE9OA"
wrote in
:


Sounds like it might be a little too hot for the band.....

I am not sure what you mean.


I have found that it's sometimes better to DE-sensitize the receiver
front-end on a noisy band. Tends to improve S/N.


I do have the service
manual, but I have never seen this kind of problem.


Never assume the manual is correct. Establish a baseline by injecting
a single tone post-AGC, run it up to 100% and measure. Then re-align
the AGC to your mic according to that baseline.

I might inject the signal right at the audio amplifier and work backwards.


Same difference.


For testing purposes, I am using a CT Systems Model 3000B Communications
Test Set. For the brief time that Tx audio is present, the modulation
envelope looks fine. This unit has a built in 100W dummy load.
RF output is flat across the band, at 4.5W, so it doesn't look like
somebody
tried to align the exciter/output stages.


Sounds like it may have an RF limiter. If you have a pulse generator,
hit the audio with 0.1ms spikes. That will also tell you if the audio
limiter has been disabled.

I'll give that at ry.


It looks like I will have to do some signal tracing with an o'scope.
Hopefully, I will be able to find that 5-pin mic plug so I can make up a
test cable. Thanks again for the advice!


No problem. If you ever get around to scanning the manual I wouldn't
mind having a copy.

I will do that. I only can scan up to 11 X 14 size, so the schematics will
be broken pages,but that should be fine.


No hurry. I only get a couple CB radios a year, so it's hard to tell
if or when I'll run across this model. There really should be a public
CB manual repository like BAMA..... maybe some day.

On Sat, 17 Feb 2007 22:26:50 -0600, "Pete KE9OA"
wrote in
:

....... about the receiver desensing.....I don't have any
problems with IMD in the receiver......


IMD is minor compared to the negative effects of high input impedance
of the RF preamp, which is almost always too high. Reducing the 1st RF
gain (A = 5 to 10) will reduce the input impedance and make it less
vulnerable to noise -- especially the white background noise. You can
compensate by increasing gain in the IF stages. Then realign your
input matching network and =blammo=, weak signals are solid with
much less noise.


......., and a diode attenuator is used in the
front end. I have found that IMD performance can be improved by changing
that general purpose switching diode in the front to a 1N5767 PIN diode.


As long as the original isn't used for overload protection, I don't
see any problem with that. But then I don't see any advantage, either.
The band is just too noisy for it to make any significant improvement.
Still, I've got a crusty old Diamond 60 laying around somewhere, so
I'll give it a try anyway.

I understand what you are saying, but the RF amplifier should be conjugate
matched to 50 ohms anyway, in order to have uncondisional stability. I am
not sure what the noise figure of this system is, but it seems that the gain
distribution is such that most of the gain is in the 2nd I.F. strip anyway.
Even so, under 30MHz, in most areas the excess environmental noise is in the
15dB region so a receiver with a 12dB noise figure does just fine.
I remember the old Allied Model 2568 CB radio. This thing had quite a bit of
RF gain and relatively low I.F. gain. As soon as you connected an antenna,
it sounded like an FM unit. The problem with that design is that the AGC
voltage was derived from the RF stage with its relatively low selectivity,
in addition to the I.F. strip. Strong off channel signals would capture the
AGC loop and desense the whole system. Remember the old term "bleed over"?
You do have a good point about keeping the RF gain ahead of the mixer as low
as possible, since any gain ahead of the 1st mixer degrades the dynamic
range by that same amount. In my last contract with Motorola, we were using
mixers that had an IP3 of +40dBm so we were able to get away with having
some gain ahead of that mixer.


IMD is minor compared to the negative effects of high input impedance
of the RF preamp, which is almost always too high. Reducing the 1st RF
gain (A = 5 to 10) will reduce the input impedance and make it less
vulnerable to noise -- especially the white background noise. You can
compensate by increasing gain in the IF stages. Then realign your
input matching network and =blammo=, weak signals are solid with
much less noise.




......., and a diode attenuator is used in the
front end. I have found that IMD performance can be improved by changing
that general purpose switching diode in the front to a 1N5767 PIN diode.


As long as the original isn't used for overload protection, I don't
see any problem with that. But then I don't see any advantage, either.
The band is just too noisy for it to make any significant improvement.
Still, I've got a crusty old Diamond 60 laying around somewhere, so
I'll give it a try anyway.

On Sun, 18 Feb 2007 09:21:02 -0600, "Pete KE9OA"
wrote:

+++I understand what you are saying, but the RF amplifier should be conjugate
+++matched to 50 ohms anyway, in order to have uncondisional stability. I am
+++not sure what the noise figure of this system is, but it seems that the gain
+++distribution is such that most of the gain is in the 2nd I.F. strip anyway.
+++Even so, under 30MHz, in most areas the excess environmental noise is in the
+++15dB region so a receiver with a 12dB noise figure does just fine.
+++I remember the old Allied Model 2568 CB radio. This thing had quite a bit of
+++RF gain and relatively low I.F. gain. As soon as you connected an antenna,
+++it sounded like an FM unit. The problem with that design is that the AGC
+++voltage was derived from the RF stage with its relatively low selectivity,
+++in addition to the I.F. strip. Strong off channel signals would capture the
+++AGC loop and desense the whole system. Remember the old term "bleed over"?
+++You do have a good point about keeping the RF gain ahead of the mixer as low
+++as possible, since any gain ahead of the 1st mixer degrades the dynamic
+++range by that same amount. In my last contract with Motorola, we were using
+++mixers that had an IP3 of +40dBm so we were able to get away with having
+++some gain ahead of that mixer.
+++
***********

Conjugate match is needed for maximum power transfer. Nuetralization
helps extend stability over various mismatch condistions.

In a receiver RF front end it is preferable to match for best noise
figure and accept the gain. The less gain before the mixer the better.
The RF front end sets the noise figure for the whole receiver. The
gain of the RF Front end need only be high enough to overcome the
noise figure of the next stage and any losses it may present if any.

More often than not,CB calibur radios places far to much gain in the
first stages so that more simple IF stages can be used. Thereby
lowering costs.

james

On Mon, 19 Feb 2007 01:09:52 GMT, james wrote
in :

On Sun, 18 Feb 2007 09:21:02 -0600, "Pete KE9OA"
wrote:

+++I understand what you are saying, but the RF amplifier should be conjugate
+++matched to 50 ohms anyway, in order to have uncondisional stability. I am
+++not sure what the noise figure of this system is, but it seems that the gain
+++distribution is such that most of the gain is in the 2nd I.F. strip anyway.
+++Even so, under 30MHz, in most areas the excess environmental noise is in the
+++15dB region so a receiver with a 12dB noise figure does just fine.
+++I remember the old Allied Model 2568 CB radio. This thing had quite a bit of
+++RF gain and relatively low I.F. gain. As soon as you connected an antenna,
+++it sounded like an FM unit. The problem with that design is that the AGC
+++voltage was derived from the RF stage with its relatively low selectivity,
+++in addition to the I.F. strip. Strong off channel signals would capture the
+++AGC loop and desense the whole system. Remember the old term "bleed over"?
+++You do have a good point about keeping the RF gain ahead of the mixer as low
+++as possible, since any gain ahead of the 1st mixer degrades the dynamic
+++range by that same amount. In my last contract with Motorola, we were using
+++mixers that had an IP3 of +40dBm so we were able to get away with having
+++some gain ahead of that mixer.
+++
***********

Conjugate match is needed for maximum power transfer.


IMPEDANCE match... for maximum power transfer. A 'conjugate' match is
when the impedances are complex, which isn't always the case.


Nuetralization
helps extend stability over various mismatch condistions.


Lots of things can improve stability, but unless the amp/mixer is
oscillating the point is moot.


In a receiver RF front end it is preferable to match for best noise
figure and accept the gain. The less gain before the mixer the better.
The RF front end sets the noise figure for the whole receiver. The
gain of the RF Front end need only be high enough to overcome the
noise figure of the next stage and any losses it may present if any.


All true. But the point I was trying to make (which I incorrectly
assumed was already understood) is that any impedance matching device
or network between the antenna and the 1st RF can cause more noise
from IMD than the noise from the 1st RF or mixer, -especially- if that
matching device/circuit uses a ferrite core or solid dielectric, which
includes almost all CB radios. That's why strong signals can sometimes
be heard even when the source is several MHz away (often mistaken for
receiver overload).

The concept here is to reduce (or, ideally, eliminate) that impedance
transformation stage. A long time ago I built a common base (voltage
follower) RF preamp using eight transistors in parallel followed by
the impedance transformation stage (transformer). The input impedance
directly to the transistors is about 100 ohms, but I feed it directly
anyway. The difference is like night and day. I use it for lowfer work
these days.


More often than not,CB calibur radios places far to much gain in the
first stages so that more simple IF stages can be used. Thereby
lowering costs.


Like you said before, the first RF is all that matters.

On Sun, 18 Feb 2007 18:24:33 -0800, Frank Gilliland
wrote:

+++Conjugate match is needed for maximum power transfer.
+++
+++
+++IMPEDANCE match... for maximum power transfer. A 'conjugate' match is
+++when the impedances are complex, which isn't always the case.
***********

I have found that it is rare in the real world that impeadances are
not complex. Outside transimission lines, there is little that is not
complex. Then again when you conjugate match, the imaginary part of
the complex impedances is nulified and you are then left with the real
part.

james

I think we got off the original subject of this post, but that's ok. There
are NO problems with the Rx section. No IMD problems, no desense of the Rx
section. Although it is a cheap design, it isn't too bad. At least a
monolithic crystal filter is used as a roofing filter at the 1st I.F. to
protect the 2nd mixer from off-frequency signals. Most of the CB units use a
10.7MHz ceramic filter with a bandwidth of anywhere from 150 to 230kHz,
which is only going to reduce 2nd image response.
Below 30MHz, you don't really need to worry so much about
NF....................a 12dB NF is fine. As long as you are conjucate
matched, stability would be good. The main thing that can cause problems is
if you S11 parameters (input return loss) aren't high enough. If you have at
least a 10dB return loss for both your S11 and S22 parameters, you shouldn't
have any problems. Most RF amplifiers are going to be unilateral, so our S12
(input / output isolation) is going to be at least 25dB. It if isn't, the
designer of the equipment didn't know what he was doing, and doesn't have
any business doing that job in the first place.

Pete

Conjugate match is needed for maximum power transfer. Nuetralization
helps extend stability over various mismatch condistions.

In a receiver RF front end it is preferable to match for best noise
figure and accept the gain. The less gain before the mixer the better.
The RF front end sets the noise figure for the whole receiver. The
gain of the RF Front end need only be high enough to overcome the
noise figure of the next stage and any losses it may present if any.

More often than not,CB calibur radios places far to much gain in the
first stages so that more simple IF stages can be used. Thereby
lowering costs.

james

On Sun, 18 Feb 2007 09:21:02 -0600, "Pete KE9OA"
wrote in
:

I understand what you are saying, but the RF amplifier should be conjugate
matched to 50 ohms anyway, in order to have uncondisional stability.


I don't have the schematic for your radio in front of me, but if that
1st RF stage is like most CB radios it's common emitter. So the input
impedance is a lot higher than 50 ohms, and is matched to the antenna
with a transformer or LC network. Not exactly ideal.


I am
not sure what the noise figure of this system is, but it seems that the gain
distribution is such that most of the gain is in the 2nd I.F. strip anyway.
Even so, under 30MHz, in most areas the excess environmental noise is in the
15dB region.......


Are we talking 11m here?


so a receiver with a 12dB noise figure does just fine.
I remember the old Allied Model 2568 CB radio. This thing had quite a bit of
RF gain and relatively low I.F. gain. As soon as you connected an antenna,
it sounded like an FM unit. The problem with that design is that the AGC
voltage was derived from the RF stage with its relatively low selectivity,
in addition to the I.F. strip. Strong off channel signals would capture the
AGC loop and desense the whole system. Remember the old term "bleed over"?
You do have a good point about keeping the RF gain ahead of the mixer as low
as possible, since any gain ahead of the 1st mixer degrades the dynamic
range by that same amount.


The objective is not low gain but low input impedance. Closer to the
impedance of the feed, to keep the first impedance transformation as
small as possible. With a common emitter, the only way to do that is
by reducing the gain. And just at the first RF stage, not necessarily
everything else in front of the first mixer.


In my last contract with Motorola, we were using
mixers that had an IP3 of +40dBm so we were able to get away with having
some gain ahead of that mixer.


Most of the problems I've had with mixers came not from the mixers but
from unbuffered oscillators. Anywayz.....

I guess the question is if the radio works well enough as it sits. If
you can hear a signal buried in the band noise then that's about as
good as it gets. The only way I know to improve it is by matching the
impedance of the first RF to the antenna. Beyond that you'll need to
get a directional antenna.

I don't have the schematic for your radio in front of me, but if that
1st RF stage is like most CB radios it's common emitter. So the input
impedance is a lot higher than 50 ohms, and is matched to the antenna
with a transformer or LC network. Not exactly ideal.

This method has been used in the real world for many years, and it is still
being used. Better ways?


I am
not sure what the noise figure of this system is, but it seems that the
gain
distribution is such that most of the gain is in the 2nd I.F. strip
anyway.
Even so, under 30MHz, in most areas the excess environmental noise is in
the
15dB region.......


Are we talking 11m here?

Of course!



The objective is not low gain but low input impedance. Closer to the
impedance of the feed, to keep the first impedance transformation as
small as possible. With a common emitter, the only way to do that is
by reducing the gain. And just at the first RF stage, not necessarily
everything else in front of the first mixer.

As long as we are on that subject, an RF stage isn't even needed at
frequencies below 30MHz. As an example, you can use a Mini-Circuits SRA-3
doubly balanced diode ring mixer, that has only 4.77dB conversion loss at
11M. You also have approximately 35dB of port to port isolation. The only
advantage that an RF amplifier would provide in this situation is minimizing
1st LO radiation through the antenna port of the radio.


In my last contract with Motorola, we were using
mixers that had an IP3 of +40dBm so we were able to get away with having
some gain ahead of that mixer.


Most of the problems I've had with mixers came not from the mixers but
from unbuffered oscillators. Anywayz.....

I guess the question is if the radio works well enough as it sits. If
you can hear a signal buried in the band noise then that's about as
good as it gets. The only way I know to improve it is by matching the
impedance of the first RF to the antenna. Beyond that you'll need to
get a directional antenna.

Agreed.