Have you looked at the 44.545 MHz injection signal with the spectrum
analyzer? What do you see there? Any spurs on it? I'm not exactly
sure how you have it set up...are you saying you have disconnected the
output of the 1st mixer so that it doesn't reached the 2nd mixer? If
so, what does the 44.545 signal mix with to get the 455 KHz 2nd IF?
Also, if the input to the 2nd mixer is simply disconnected, is the input
port of the 2nd mixer terminated somehow? Also, are you feeding the
proper level of 44.545 to it (not overdriving it are you??)? Not sure
what level is required for the SBL-1.

Scott
N0EDV

john wilkinson wrote:
Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.

The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

Scott wrote:
Have you looked at the 44.545 MHz injection signal with the spectrum
analyzer? What do you see there? Any spurs on it? I'm not exactly
sure how you have it set up...are you saying you have disconnected the
output of the 1st mixer so that it doesn't reached the 2nd mixer? If
so, what does the 44.545 signal mix with to get the 455 KHz 2nd IF?
Also, if the input to the 2nd mixer is simply disconnected, is the input
port of the 2nd mixer terminated somehow? Also, are you feeding the
proper level of 44.545 to it (not overdriving it are you??)? Not sure
what level is required for the SBL-1.


+7 dBm.

Leon

On Tue, 27 Jun 2006 21:37:44 +0100, john wilkinson
wrote:

Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.

The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

You have spurs on one or both of the inputs to the SBL1. Passive
doide mixers will not generate spurs like that though they may pass
them through if the sources are dirty.

Also the IF could also be oscillating and your see ing that
contribution. If its the 100db IF duscussed before, thats a lot of
gain and I'd hope its bypassed and shielded very well.

Allison

Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.

The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

On Wed, 28 Jun 2006 19:48:20 +0100, john wilkinson
wrote:

You have spurs on one or both of the inputs to the SBL1. Passive
doide mixers will not generate spurs like that though they may pass
them through if the sources are dirty.

Also the IF could also be oscillating and your see ing that
contribution. If its the 100db IF duscussed before, thats a lot of
gain and I'd hope its bypassed and shielded very well.

Allison
Hi Allison,

I have disconected the second mixer from the output of the first IF amp.
That amp is powered down, and disconnected from any supply.
The LO input to the mixer is from a crystal oscillator, at 7-8dBm.
The output from the mixer is into the spectrum analyser only, not the
second IF.

This morning I did an experiment. If I left the RF port unterminated, the
output was quite clean on the analyser.
If I put a 50 Ohm term to gnd, the spurs appear.

Either the LO is dirty and your not seeing it due to overdriving the
analyser or the stages after SBL1 are dirty.

What I'd expect with the input of the SBL1 terminated and the LO
applied is the LO (40db or so down) some second and third harmonic
also way down and if you go far enough down some noise.

Maybe it is on the LO itself, but looking at the LO on the analyser it
looks OK to me.

Ok whats at the output of the SBL1? Or more specifically whats
between the SBL1 and the Analyser?


Allison

john wilkinson wrote:



Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.

The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

A couple of observations.

If you simply disconnected the second mixer from the 1st mixer then it
sounds like you do not have the output of the 1st mixer properly
terminated. Unless, that is, the spectrum analyzer provides a good 50ohm
termination in the probe. The SBL-1 is a double-balanced mixer but in order
to get the "balance" to work right you must properly terminate *all* ports.
Otherwise all kinds of impedances can be "thrown" back to the other ports -
causing all kinds of problems with attached LO's and RF stages.

Regularly spaced spurs from DC to 455khz does not sound like spurious mixing
products. It sounds more like an oscillator being driven into saturation
causing square waves to be produced which are providing regular harmonic
outputs. This could be a product of poor terminations on the mixer ports.

Directly connecting the output of a SBL-1 to the input of another mixer,
even another SBL-1, is not recommended either. The SBL-1 does not have
built-in terminations. A good, symmetric, 50ohm 3db pad between the stages
would provide a much better setup and would not significantly impact your
overall noise factor. Be sure and connect your spectrum analyzer to the far
side of the pad.

tim ab0wr

On Tue, 27 Jun 2006 17:03:06 +0000, nospam wrote:

On Tue, 27 Jun 2006 21:37:44 +0100, john wilkinson
wrote:

Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.

The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

You have spurs on one or both of the inputs to the SBL1. Passive
doide mixers will not generate spurs like that though they may pass
them through if the sources are dirty.

Also the IF could also be oscillating and your see ing that
contribution. If its the 100db IF duscussed before, thats a lot of
gain and I'd hope its bypassed and shielded very well.

Allison
Hi Allison,

I have disconected the second mixer from the output of the first IF amp.
That amp is powered down, and disconnected from any supply.
The LO input to the mixer is from a crystal oscillator, at 7-8dBm.
The output from the mixer is into the spectrum analyser only, not the
second IF.

This morning I did an experiment. If I left the RF port unterminated, the
output was quite clean on the analyser.
If I put a 50 Ohm term to gnd, the spurs appear.

Maybe it is on the LO itself, but looking at the LO on the analyser it
looks OK to me.

Thanks,
John

On Wed, 28 Jun 2006 07:43:20 -0500, tim gorman
wrote:

john wilkinson wrote:



Hi,
If I feed my second mixer into a spectrum analyser, and inject a 44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.


The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

A couple of observations.

If you simply disconnected the second mixer from the 1st mixer then it
sounds like you do not have the output of the 1st mixer properly
terminated. Unless, that is, the spectrum analyzer provides a good 50ohm
termination in the probe. The SBL-1 is a double-balanced mixer but in order
to get the "balance" to work right you must properly terminate *all* ports.
Otherwise all kinds of impedances can be "thrown" back to the other ports -
causing all kinds of problems with attached LO's and RF stages.

Regularly spaced spurs from DC to 455khz does not sound like spurious mixing
products. It sounds more like an oscillator being driven into saturation
causing square waves to be produced which are providing regular harmonic
outputs. This could be a product of poor terminations on the mixer ports.

Tim, the LO is 45.455 (or 44.545) either way there should be no LF
outputs from the mixer below the LO injection unless it's a spur on
the LO.

Directly connecting the output of a SBL-1 to the input of another mixer,
even another SBL-1, is not recommended either. The SBL-1 does not have
built-in terminations. A good, symmetric, 50ohm 3db pad between the stages
would provide a much better setup and would not significantly impact your
overall noise factor. Be sure and connect your spectrum analyzer to the far
side of the pad.

Thats true though his testing is with a 20db IF after it and one would
hope that does present a 50ohm match to the IF port. Usually if the
poarts are badly matched the mixing spurs are greater and the port
to port isolation go to pot. If the only thing going in is the LO
then likely the LO is the source but if there is gain flollowing then
the gain stage is also suspect.

I've worked with enough SBL1 and MD108s and all their similar DBMS
to know the SBL1 in this case is not the offendor Even if
misterminated. Likely causes are the 45mhz filter is looking reactive
to the IF amp and the IF is unstable (makes a fine OSC) or there is
some other source of RF that is unaccounted for (spurs).


Allison
Kb!GMX

What kind of spectrum analyzer are you using? What are the front end
attenuator settings on the spectrum analyzer? What is your reference level?
All of these things, especially the front end attenuator settings, can
affect your result. If the 1st mixer in the spectrum analyzer is driven into
nonlinearity, you can see all kinds of IMD products that are generated in
the analyzer itself.
Also, why are you doing this test with the RF port of the mixer terminated
in 50 Ohms? This type of termination is only used when doing either return
loss tests or noise balance tests.
If you are driving the SBL-1 mixer with +7 to +9dBm of LO level and you are
driving the RF port with a signal that is at least 20dB below the 1dB
compression point of the SBL-1, the main things you should see at the I.F.
port are a supressed LO signal, an upper sideband that is about 7dB lower in
level than the RF injection level, and a lower sideband that is also about
7dB lower in level than the RF injection level.
The 1dB compression point on the SBL-1 is about 0dBm, so I would set the
level to the RF port at -20dBm. If you go above this level, you will see
higher conversion loss, and possibly more IMD products at the I.F. port of
the mixer.
In conclusion, make sure that the IMD isn't being generated in your test
setup, and make sure that you are not exceeding the recommended RF input
level to the SBL-1.

Pete

wrote in message
news
On Wed, 28 Jun 2006 07:43:20 -0500, tim gorman
wrote:

john wilkinson wrote:



Hi,
If I feed my second mixer into a spectrum analyser, and inject a
44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output freq.


The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

A couple of observations.

If you simply disconnected the second mixer from the 1st mixer then it
sounds like you do not have the output of the 1st mixer properly
terminated. Unless, that is, the spectrum analyzer provides a good 50ohm
termination in the probe. The SBL-1 is a double-balanced mixer but in
order
to get the "balance" to work right you must properly terminate *all*
ports.
Otherwise all kinds of impedances can be "thrown" back to the other
ports -
causing all kinds of problems with attached LO's and RF stages.

Regularly spaced spurs from DC to 455khz does not sound like spurious
mixing
products. It sounds more like an oscillator being driven into saturation
causing square waves to be produced which are providing regular harmonic
outputs. This could be a product of poor terminations on the mixer ports.

Tim, the LO is 45.455 (or 44.545) either way there should be no LF
outputs from the mixer below the LO injection unless it's a spur on
the LO.

Directly connecting the output of a SBL-1 to the input of another mixer,
even another SBL-1, is not recommended either. The SBL-1 does not have
built-in terminations. A good, symmetric, 50ohm 3db pad between the stages
would provide a much better setup and would not significantly impact your
overall noise factor. Be sure and connect your spectrum analyzer to the
far
side of the pad.

Thats true though his testing is with a 20db IF after it and one would
hope that does present a 50ohm match to the IF port. Usually if the
poarts are badly matched the mixing spurs are greater and the port
to port isolation go to pot. If the only thing going in is the LO
then likely the LO is the source but if there is gain flollowing then
the gain stage is also suspect.

I've worked with enough SBL1 and MD108s and all their similar DBMS
to know the SBL1 in this case is not the offendor Even if
misterminated. Likely causes are the 45mhz filter is looking reactive
to the IF amp and the IF is unstable (makes a fine OSC) or there is
some other source of RF that is unaccounted for (spurs).


Allison
Kb!GMX

This didn't sound like the problem if I read it correctly, but...
To test for analyzedr generated effects, switch-in 10 dB of attenuatin in
FRONT of the analyzer mixer (not in the analyzer IF). If the garbage drops
more than 10dB (like 20 or 30) then you are over driving the analyzer. If
it is coming from outside, then you'll se dB for dB changes.

73, Steve, K9DCI

Ain't the math of nonlinearity great?



"Pete KE9OA" wrote in message
. ..
What kind of spectrum analyzer are you using? What are the front end
attenuator settings on the spectrum analyzer? What is your reference
level?
All of these things, especially the front end attenuator settings, can
affect your result. If the 1st mixer in the spectrum analyzer is driven
into
nonlinearity, you can see all kinds of IMD products that are generated in
the analyzer itself.
Also, why are you doing this test with the RF port of the mixer terminated
in 50 Ohms? This type of termination is only used when doing either
return
loss tests or noise balance tests.
If you are driving the SBL-1 mixer with +7 to +9dBm of LO level and you
are
driving the RF port with a signal that is at least 20dB below the 1dB
compression point of the SBL-1, the main things you should see at the I.F.
port are a supressed LO signal, an upper sideband that is about 7dB lower
in
level than the RF injection level, and a lower sideband that is also about
7dB lower in level than the RF injection level.
The 1dB compression point on the SBL-1 is about 0dBm, so I would set the
level to the RF port at -20dBm. If you go above this level, you will see
higher conversion loss, and possibly more IMD products at the I.F. port of
the mixer.
In conclusion, make sure that the IMD isn't being generated in your test
setup, and make sure that you are not exceeding the recommended RF input
level to the SBL-1.

Pete

wrote in message
news
On Wed, 28 Jun 2006 07:43:20 -0500, tim gorman
wrote:

john wilkinson wrote:



Hi,
If I feed my second mixer into a spectrum analyser, and inject a
44.545MHz
LO, with no power to the first IF stage, I see spurs at about 48-50KHz
intervals on the mixer output, from 0 to the 455KHz normal output
freq.


The mixer is an SBL-1.

Any ideas as to where these are comming from?

Best regards,
John

A couple of observations.

If you simply disconnected the second mixer from the 1st mixer then it
sounds like you do not have the output of the 1st mixer properly
terminated. Unless, that is, the spectrum analyzer provides a good 50ohm
termination in the probe. The SBL-1 is a double-balanced mixer but in
order
to get the "balance" to work right you must properly terminate *all*
ports.
Otherwise all kinds of impedances can be "thrown" back to the other
ports -
causing all kinds of problems with attached LO's and RF stages.

Regularly spaced spurs from DC to 455khz does not sound like spurious
mixing
products. It sounds more like an oscillator being driven into saturation
causing square waves to be produced which are providing regular harmonic
outputs. This could be a product of poor terminations on the mixer
ports.

Tim, the LO is 45.455 (or 44.545) either way there should be no LF
outputs from the mixer below the LO injection unless it's a spur on
the LO.

Directly connecting the output of a SBL-1 to the input of another mixer,
even another SBL-1, is not recommended either. The SBL-1 does not have
built-in terminations. A good, symmetric, 50ohm 3db pad between the
stages
would provide a much better setup and would not significantly impact
your
overall noise factor. Be sure and connect your spectrum analyzer to the
far
side of the pad.

Thats true though his testing is with a 20db IF after it and one would
hope that does present a 50ohm match to the IF port. Usually if the
poarts are badly matched the mixing spurs are greater and the port
to port isolation go to pot. If the only thing going in is the LO
then likely the LO is the source but if there is gain flollowing then
the gain stage is also suspect.

I've worked with enough SBL1 and MD108s and all their similar DBMS
to know the SBL1 in this case is not the offendor Even if
misterminated. Likely causes are the 45mhz filter is looking reactive
to the IF amp and the IF is unstable (makes a fine OSC) or there is
some other source of RF that is unaccounted for (spurs).


Allison
Kb!GMX