|
Hi,
I am troubleshooting a signal generator and the service manual says to check TP1 for 500MHz at +2-3dBm.
TP1 appears to be just a probe location on a trace (see attached image).
My question is what sort of probe should I attach to my spectrum analyser (8566B) to check the test point on the trace? Can I use a oscilloscope probe? If so, do I attach the probe ground lead?
I¡¯m used to using straight BNC-based input so this has me a bit unsure how to proceed.
I don¡¯t want to get an incorrect reading or worse, damage my SA!
thanks
tony
|
dBm is a power reading, so it cannot be made without knowing
the impedance at the point of measurement. They probably mean
dBm @ 50 ohms.
I think you will be able to observe that the test point is at
some point on a trace that is designed to be a transmission line.
You will probably also see that the test point is surrounded by
ground plane.
Smart manufacturers will often design in a simple probe with a
simple connector at test points such as this, however, that is
not always done because of the line disturbance that a probe
and unused connector may cause at the test point.
What I would do is cut the extra connector off of a scrap 141
hardline jumper, and machine the cut end so that it can safely
engage both the test point and the ground plane that I assume
will be surrounding the test point. Attach the homemade 50 ohm
probe to a flexible cable that goes to your SA's input, and make
your measurement.
-Chuck Harris
toggle quoted message
Show quoted text
On Sun, 19 Feb 2023 06:38:36 -0800 "Tony" <tonycox01@...> wrote: Hi,
I am troubleshooting a signal generator and the service manual says
to check TP1 for 500MHz at +2-3dBm.
TP1 appears to be just a probe location on a trace (see attached
image).
My question is what sort of probe should I attach to my spectrum
analyser (8566B) to check the test point on the trace? Can I use a
oscilloscope probe? If so, do I attach the probe ground lead?
I¡¯m used to using straight BNC-based input so this has me a bit
unsure how to proceed.
I don¡¯t want to get an incorrect reading or worse, damage my SA!
thanks
tony
Attachments:
photo-2.jpg:
/g/HP-Agilent-Keysight-equipment/attachment/132707/
|
Do you have a scope? If so try that first. Even if the BW is lower than 500MHz you should be able to see some signal and it will give you DC levels. That should let you determine if there are SA damaging voltages present. Given the power levels you suggest that should be ok for the SA. HP had some special cable assay¡¯s that would plug into the two ¡°sockets¡± shown. A 10X probe should be OK, AFTER checking for DC voltages and I would ground the probe. ?From the picture I¡¯d suspect that the other pin is a ground connection. Not knowing what the schematic says on that little circuit it¡¯s hard to guess. ?I think it looks like the black diode might be a PIN switch with the ¡°on¡± bias supplied through the glass diode. ?Another question would be where does the fat trace leading to the left of the black diode go? ?It is also DC blocked, so that could mean the PIN diode idea is OK. If that trace goes to an off board connector you could try sampling after the DC blocking cap or if it¡¯s a connector just pull that connector and see what¡¯s there. ?
Hope this helps a bit. ?Just make sure there are no DC voltages anywhere you try to monitor with the SA.?
Steve
|
A x10 scope probe won't work well with a SA because the probe is made up of a 9M series resistance that works with the 'scopes 1M input resistance to make x10. The SA has a 50R input resistance, so instead of x10 you get x 180000.
|
"Tony" <tonycox01@...> writes: I am troubleshooting a signal generator and the service manual says to check TP1 for 500MHz at +2-3dBm.
TP1 appears to be just a probe location on a trace (see attached image).
My question is what sort of probe should I attach to my spectrum analyser (8566B) to check the test point on the trace? Can I use a oscilloscope probe? If so, do I attach the probe ground lead?
I¡¯m used to using straight BNC-based input so this has me a bit unsure how to proceed.
I don¡¯t want to get an incorrect reading or worse, damage my SA! If you have it, I'd probably use a high-frequency passive probe for this. For example, the PMK PML series: I think Keysight and others just rebrand this for their high-frequency passive probes. The 50ohm-input terminated 10:1 and 100:1 versions will get you up to 1.5 GHz. If you use this with a scope you'll have to use it's 50ohm input or add a 50ohm pass-through termination. Of course, your SA is already 50ohm input so nothing additional needed there. If you use 100:1 division, the resulting signal will be on the weak side for an oscope, but your SA will see it fine. Of course you can also use a suitable active probe if you have that, but check it's damage levels first and make sure the signal is within those before using it (that actually goes for the 10:1 passive probe too, which can only handle 12V peak). In the likely scenario you don't have these, I'd just use something you do have (as others have suggested), such as a normal 10x probe and realize the attenuation at 500 MHz will be significant. You can solder in a coax cable too, but realize that you'll be loading down the circuit pretty badly. The capacitance of the cable will also present a significant load to the circuit. You might be able to solder a resistor in series to decrease the loading, but I don't know how easy it will be to get high-frequency performance doing this. As for damaging the SA, put a good DC block in series with the input. That should prevent damage in the case of DC voltages being present. I actually just leave DC blocks on my 8566's at all times and take them off if I ever need to see below the 10 MHz cutoff. If you're still worried, you can probe the signal first with other means (eg an oscope) to check for large DC voltages or even large AC swings, but this all seems unlikely to me. Matt
|
Tony
In years past I made probes from two resistors and a length of coax. The probe was made of a series
5100 ohm resistor that connected to the center of the coax cable. A 51 ohm resistor was placed across
the junction of the coax and the 5100 ohm resistor to ground. If you want to have a more precise
probe use two paralleled 10 K resistors for the series resistor and two paralleled 100 Ohm resistors
for the shunt 51 ohm. Using the resistors with the spectrum analyzer, you should have a fairly wide
bandwidth probe up to a GHz with -40 dB gain loss. So for a 0 dBm signal, your spectrum analyzer
will display -40 dBm. The error using this arrangement will be less than 0.1 dB and 5 K plus of resistance
shunting your transmission line on the board will have negligible effect.
Regards
Chuck
|
If you can find ont, I recommend attaching an HP FET probe to your SA. If you don't have one, it is a good idea to start looking for one. I find them a valuable resource for trouble shooting.
Note: the passive probe is unlikely to work.
Cheers!
Bruce
Quoting Matt Huszagh <huszaghmatt@...>:
toggle quoted message
Show quoted text
"Tony" <tonycox01@...> writes:
I am troubleshooting a signal generator and the service manual says to check TP1 for 500MHz at +2-3dBm.
TP1 appears to be just a probe location on a trace (see attached image).
My question is what sort of probe should I attach to my spectrum analyser (8566B) to check the test point on the trace? Can I use a oscilloscope probe? If so, do I attach the probe ground lead?
I¡¯m used to using straight BNC-based input so this has me a bit unsure how to proceed.
I don¡¯t want to get an incorrect reading or worse, damage my SA! If you have it, I'd probably use a high-frequency passive probe for
this. For example, the PMK PML series:
I think Keysight and others just rebrand this for their high-frequency
passive probes.
The 50ohm-input terminated 10:1 and 100:1 versions will get you up to
1.5 GHz. If you use this with a scope you'll have to use it's 50ohm
input or add a 50ohm pass-through termination. Of course, your SA is
already 50ohm input so nothing additional needed there. If you use 100:1
division, the resulting signal will be on the weak side for an oscope,
but your SA will see it fine.
Of course you can also use a suitable active probe if you have that, but
check it's damage levels first and make sure the signal is within those
before using it (that actually goes for the 10:1 passive probe too,
which can only handle 12V peak).
In the likely scenario you don't have these, I'd just use something you
do have (as others have suggested), such as a normal 10x probe and
realize the attenuation at 500 MHz will be significant. You can solder
in a coax cable too, but realize that you'll be loading down the circuit
pretty badly. The capacitance of the cable will also present a
significant load to the circuit. You might be able to solder a resistor
in series to decrease the loading, but I don't know how easy it will be
to get high-frequency performance doing this.
As for damaging the SA, put a good DC block in series with the
input. That should prevent damage in the case of DC voltages being
present. I actually just leave DC blocks on my 8566's at all times and
take them off if I ever need to see below the 10 MHz cutoff. If you're
still worried, you can probe the signal first with other means (eg an
oscope) to check for large DC voltages or even large AC swings, but this
all seems unlikely to me.
Matt
|
"Bruce" <bruce@...> writes: If you can find ont, I recommend attaching an HP FET probe to your SA.
If you don't have one, it is a good idea to start looking for one.
I find them a valuable resource for trouble shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe (like the PML series) can absolutely work up to 500 MHz and beyond. In fact, you can make them work up to several GHz with proper design. See the Andrew Zonenberg probes, for instance (). Of course, the passive probe approach mandates some sort of voltage division, otherwise this won't work. Sticking some coax between your test point and the oscope will just load down the circuit with a bunch of parallel capacitance. A typical 10:1 or 1:1 oscope probe designed for high input impedance scope termination won't work though. Don't confuse passive probes designed for 1M scope input impedance with those designed for 50ohm scope input impedance. You probably know this, but if you do use a scope and not the SA, make sure the reference point can be connected to earth ground. If not you'll need something to isolate it. The typical approach would be to use a differential probe. Matt
|
Matt -
I agree - I was assuming a probe designed for Hi-Z and am unfamiliar with any others.
Cheers!
Bruce
Quoting Matt Huszagh <huszaghmatt@...>:
toggle quoted message
Show quoted text
"Bruce" <bruce@...> writes:
If you can find ont, I recommend attaching an HP FET probe to your SA.
If you don't have one, it is a good idea to start looking for one.
I find them a valuable resource for trouble shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe (like
the PML series) can absolutely work up to 500 MHz and beyond. In fact,
you can make them work up to several GHz with proper design. See the
Andrew Zonenberg probes, for instance
(). Of course, the passive probe
approach mandates some sort of voltage division, otherwise this won't
work. Sticking some coax between your test point and the oscope will
just load down the circuit with a bunch of parallel capacitance.
A typical 10:1 or 1:1 oscope probe designed for high input impedance
scope termination won't work though. Don't confuse passive probes
designed for 1M scope input impedance with those designed for 50ohm
scope input impedance.
You probably know this, but if you do use a scope and not the SA, make
sure the reference point can be connected to earth ground. If not you'll
need something to isolate it. The typical approach would be to use a
differential probe.
Matt
|
If the OP's description of the set point being 2-3 dBm, is accurate, you can do nothing else but use a 50 ohm probe to make the connection. dBm is totally undefined at the arbitrary high impedances of FET probes. dBm is totally undefined at the arbitrary impedances of something like the Andrew Zonenberg probes. dBm is totally undefined at the arbitrary impedance of using a resister in series with the center lead of a coax cable... The only way the high impedance route could work is if the OP, or the MFR, used dBm in an imprecise way, and really meant something else. If the MFR did that, he likely would have the sort of probe he intended listed in the necessary calibration equipment table for the instrument. -Chuck Harris On Sun, 19 Feb 2023 10:29:50 -0800 "Matt Huszagh" <huszaghmatt@...> wrote: "Bruce" <bruce@...> writes:
If you can find ont, I recommend attaching an HP FET probe to your
SA. If you don't have one, it is a good idea to start looking for
one. I find them a valuable resource for trouble shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe
(like the PML series) can absolutely work up to 500 MHz and beyond.
In fact, you can make them work up to several GHz with proper design.
See the Andrew Zonenberg probes, for instance
(). Of course, the passive
probe approach mandates some sort of voltage division, otherwise this
won't work. Sticking some coax between your test point and the oscope
will just load down the circuit with a bunch of parallel capacitance.
A typical 10:1 or 1:1 oscope probe designed for high input impedance
scope termination won't work though. Don't confuse passive probes
designed for 1M scope input impedance with those designed for 50ohm
scope input impedance.
You probably know this, but if you do use a scope and not the SA, make
sure the reference point can be connected to earth ground. If not
you'll need something to isolate it. The typical approach would be to
use a differential probe.
Matt
|
"Chuck Harris" <cfharris@...> writes: If the OP's description of the set point being 2-3 dBm,
is accurate, you can do nothing else but use a 50 ohm
probe to make the connection.
dBm is totally undefined at the arbitrary high impedances
of FET probes.
dBm is totally undefined at the arbitrary impedances of something
like the Andrew Zonenberg probes.
dBm is totally undefined at the arbitrary impedance of using
a resister in series with the center lead of a coax cable...
The only way the high impedance route could work is if the OP,
or the MFR, used dBm in an imprecise way, and really meant something
else. If the MFR did that, he likely would have the sort of probe
he intended listed in the necessary calibration equipment table for
the instrument. Ok, fair enough. I had assumed 2-3 dBm meant into a 50ohm load presented by whatever is downstream of this. If it's 2-3 dBm when you load it externally with 50 ohms, then yes of course you'll want to load it. It does seem a little odd to ask you to change the circuit behavior when probing though, which is what will happen if you load it down by an external 50 ohms. But, admittedly, I haven't seen this sort of test point much so maybe I'm just not familiar with the typical procedure here. Minor nit: passive probes like the Zonenberg probes aren't "arbitrary impedance". They're designed for specific division ratios into a 50 ohm input, so they have a very well-defined impedance, at least over the specified bandwidth. Matt
|
Both SAs and Scopes measure voltage. If a signal level is CORRECTLY specified in dBm, the impedance must be 50 ohms unless otherwise stated.
All that is necessary to do is the mathematical conversion from power to voltage (works in general if the particular impedance is known).
An easy way to avoid the maths ias to use the FET probe to measure the output of a working signal generator (with appropriate termination). I find the HP FET probes a valuable asset for debugging - protects the SA from DC and overloads (to some degree).
Cheers!
Bruce
Quoting Chuck Harris <cfharris@...>:
toggle quoted message
Show quoted text
If the OP's description of the set point being 2-3 dBm,
is accurate, you can do nothing else but use a 50 ohm
probe to make the connection.
dBm is totally undefined at the arbitrary high impedances
of FET probes.
dBm is totally undefined at the arbitrary impedances of something
like the Andrew Zonenberg probes.
dBm is totally undefined at the arbitrary impedance of using
a resister in series with the center lead of a coax cable...
The only way the high impedance route could work is if the OP,
or the MFR, used dBm in an imprecise way, and really meant something
else. If the MFR did that, he likely would have the sort of probe
he intended listed in the necessary calibration equipment table for
the instrument.
-Chuck Harris
On Sun, 19 Feb 2023 10:29:50 -0800 "Matt Huszagh"
<huszaghmatt@...> wrote:
"Bruce" <bruce@...> writes:
If you can find ont, I recommend attaching an HP FET probe to your
SA. If you don't have one, it is a good idea to start looking for
one. I find them a valuable resource for trouble shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe
(like the PML series) can absolutely work up to 500 MHz and beyond.
In fact, you can make them work up to several GHz with proper design.
See the Andrew Zonenberg probes, for instance
(). Of course, the passive
probe approach mandates some sort of voltage division, otherwise this
won't work. Sticking some coax between your test point and the oscope
will just load down the circuit with a bunch of parallel capacitance.
A typical 10:1 or 1:1 oscope probe designed for high input impedance
scope termination won't work though. Don't confuse passive probes
designed for 1M scope input impedance with those designed for 50ohm
scope input impedance.
You probably know this, but if you do use a scope and not the SA, make
sure the reference point can be connected to earth ground. If not
you'll need something to isolate it. The typical approach would be to
use a differential probe.
Matt
|
?
One type of probe that could be used for this purpose is a passive probe with a bandwidth of at least 500 MHz and a maximum input voltage that exceeds +2 dBm. A passive probe is typically less expensive and simpler to use than an active probe, but it can introduce some loading and attenuation to the circuit under test.
Another option is an active probe that has a frequency range that covers 500 MHz and a maximum input voltage that exceeds +2 dBm. An active probe uses a built-in amplifier to boost the signal and minimize loading effects, but it can be more expensive and may require additional power.
In addition to selecting an appropriate probe, it's important to properly calibrate the spectrum analyzer before taking measurements. This typically involves setting the reference level, center frequency, and resolution bandwidth to appropriate values based on the signal being measured.
|
?
You may also use a differential probe, it is important to pay attention to the probe's input impedance and the common mode rejection ratio (CMRR) of the probe and oscilloscope. The input impedance of the probe can affect the signal being measured, and the CMRR is important for rejecting noise and interference in the measurement.
|
Hi Tony ? Just looking at your post and the excellent advice given I think there are a few pointers I can give you as I strongly advise you invest in the following ? I use DC blocks on my spectrum analysers when probing around to protect the input ? Depending on the job but a useful tool is a short length of semi rigged one end the centre conductor as a probe the other end a SMA? this is connected to a DC block then coax to the SA ? DC block suggest you look out for a HP 0960-0092 this is infect a DC return use the dc return on the SA side it also acts as a handle for the above although it spec is to 18 GHz it works well above 20 GHz and is a semi permeant fixture on my 26 GHz analysers? cost about ?40 Not good for specific levels but close enough and gives a value to work with ? Active probe suggest you look out for a HP1120A they work to 500 MHz? cost about ?60 - ?100 depending on parts it comes with ? Paul ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Andrew
Sent: 19 February 2023 19:28
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] Advice on how to use an SA for test point troubleshooting ? One type of probe that could be used for this purpose is a passive probe with a bandwidth of at least 500 MHz and a maximum input voltage that exceeds +2 dBm. A passive probe is typically less expensive and simpler to use than an active probe, but it can introduce some loading and attenuation to the circuit under test. Another option is an active probe that has a frequency range that covers 500 MHz and a maximum input voltage that exceeds +2 dBm. An active probe uses a built-in amplifier to boost the signal and minimize loading effects, but it can be more expensive and may require additional power. In addition to selecting an appropriate probe, it's important to properly calibrate the spectrum analyzer before taking measurements. This typically involves setting the reference level, center frequency, and resolution bandwidth to appropriate values based on the signal being measured.
|
Hi Bruce,
I agree, but the person specifying that you look at that point needs
to define how. And, I am pretty sure they did... if one looks.
Reading minds is a perilous task when you are up to your hips inside
of an instrument.
-Chuck Harris
toggle quoted message
Show quoted text
On Sun, 19 Feb 2023 12:24:06 -0700 "Bruce" <bruce@...> wrote: Both SAs and Scopes measure voltage. If a signal level is CORRECTLY
specified in dBm, the impedance must be 50 ohms unless otherwise
stated.
All that is necessary to do is the mathematical conversion from power
to voltage (works in general if the particular impedance is known).
An easy way to avoid the maths ias to use the FET probe to measure
the output of a working signal generator (with appropriate
termination). I find the HP FET probes a valuable asset for debugging
- protects the SA from DC and overloads (to some degree).
Cheers!
Bruce
Quoting Chuck Harris <cfharris@...>:
If the OP's description of the set point being 2-3 dBm,
is accurate, you can do nothing else but use a 50 ohm
probe to make the connection.
dBm is totally undefined at the arbitrary high impedances
of FET probes.
dBm is totally undefined at the arbitrary impedances of something
like the Andrew Zonenberg probes.
dBm is totally undefined at the arbitrary impedance of using
a resister in series with the center lead of a coax cable...
The only way the high impedance route could work is if the OP,
or the MFR, used dBm in an imprecise way, and really meant something
else. If the MFR did that, he likely would have the sort of probe
he intended listed in the necessary calibration equipment table for
the instrument.
-Chuck Harris
On Sun, 19 Feb 2023 10:29:50 -0800 "Matt Huszagh"
<huszaghmatt@...> wrote:
"Bruce" <bruce@...> writes:
If you can find ont, I recommend attaching an HP FET probe to
your SA. If you don't have one, it is a good idea to start
looking for one. I find them a valuable resource for trouble
shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe
(like the PML series) can absolutely work up to 500 MHz and beyond.
In fact, you can make them work up to several GHz with proper
design. See the Andrew Zonenberg probes, for instance
(). Of course, the passive
probe approach mandates some sort of voltage division, otherwise
this won't work. Sticking some coax between your test point and
the oscope will just load down the circuit with a bunch of
parallel capacitance.
A typical 10:1 or 1:1 oscope probe designed for high input
impedance scope termination won't work though. Don't confuse
passive probes designed for 1M scope input impedance with those
designed for 50ohm scope input impedance.
You probably know this, but if you do use a scope and not the SA,
make sure the reference point can be connected to earth ground. If
not you'll need something to isolate it. The typical approach
would be to use a differential probe.
Matt
|
Chuck -
Gotta agree with the mind reading comment. Thanks!
What I suspect is the "test point" is probably a connection between circuit elements (amp, splitter, modulator,etc.) and they specified the signal level as a diagnostic (Take a look at the system level 8340 diagram for example) and that accuracy is not terribly important. In these cases, if you can break the circuit, use a power meter as the 50 Ohm termination. If it is a PC trace, a Hi-Z probe is a better option because there is already a 50 ohm termination.
WHAT WE SHOULD ASK FOR is a picture of the circuit diagram (no more mind reading) and then we could make better recommendations.
Cheers!
Bruce
Quoting Chuck Harris <cfharris@...>:
toggle quoted message
Show quoted text
Hi Bruce,
I agree, but the person specifying that you look at that point needs
to define how. And, I am pretty sure they did... if one looks.
Reading minds is a perilous task when you are up to your hips inside
of an instrument.
-Chuck Harris
On Sun, 19 Feb 2023 12:24:06 -0700 "Bruce" <bruce@...> wrote:
Both SAs and Scopes measure voltage. If a signal level is CORRECTLY
specified in dBm, the impedance must be 50 ohms unless otherwise
stated.
All that is necessary to do is the mathematical conversion from power
to voltage (works in general if the particular impedance is known).
An easy way to avoid the maths ias to use the FET probe to measure
the output of a working signal generator (with appropriate
termination). I find the HP FET probes a valuable asset for debugging
- protects the SA from DC and overloads (to some degree).
Cheers!
Bruce
Quoting Chuck Harris <cfharris@...>:
If the OP's description of the set point being 2-3 dBm,
is accurate, you can do nothing else but use a 50 ohm
probe to make the connection.
dBm is totally undefined at the arbitrary high impedances
of FET probes.
dBm is totally undefined at the arbitrary impedances of something
like the Andrew Zonenberg probes.
dBm is totally undefined at the arbitrary impedance of using
a resister in series with the center lead of a coax cable...
The only way the high impedance route could work is if the OP,
or the MFR, used dBm in an imprecise way, and really meant something
else. If the MFR did that, he likely would have the sort of probe
he intended listed in the necessary calibration equipment table for
the instrument.
-Chuck Harris
On Sun, 19 Feb 2023 10:29:50 -0800 "Matt Huszagh"
<huszaghmatt@...> wrote:
"Bruce" <bruce@...> writes:
If you can find ont, I recommend attaching an HP FET probe to
your SA. If you don't have one, it is a good idea to start
looking for one. I find them a valuable resource for trouble
shooting.
Note: the passive probe is unlikely to work. While a FET probe is nice, a suitable high-frequency passive probe
(like the PML series) can absolutely work up to 500 MHz and beyond.
In fact, you can make them work up to several GHz with proper
design. See the Andrew Zonenberg probes, for instance
(). Of course, the passive
probe approach mandates some sort of voltage division, otherwise
this won't work. Sticking some coax between your test point and
the oscope will just load down the circuit with a bunch of
parallel capacitance.
A typical 10:1 or 1:1 oscope probe designed for high input
impedance scope termination won't work though. Don't confuse
passive probes designed for 1M scope input impedance with those
designed for 50ohm scope input impedance.
You probably know this, but if you do use a scope and not the SA,
make sure the reference point can be connected to earth ground. If
not you'll need something to isolate it. The typical approach
would be to use a differential probe.
Matt
|
It would b useful to know the make and model of the subject signal generator. The list of required test equipment should reveal all.
From the picture it clearly is a "proper" test point with a pair of female contacts. I'd guess that it is feeding a PIN switch and the test requires a 50R terminated measurement with the switch off. That is a guess, RTFM.
Robert.
|
Hi,
Thanks for all the replies - I'm a bit like the rabbit in the headlights - and beginning to realize how much I don't know!
As a bit of background, the signal gen is a HP 8656A and has no output. Looking at the output on an SA I see some sort of haphazard signal that appears to be sweeping rapidly over the entire output range, but there is no stable output anywhere - so it is pretty broken. Working through the T/S manual I have verified that all the power rails are good and the attenuators all seem to be working (no sticking etc).
The manual next suggests setting the frequency to 500MHz at 0dBm and checking the oscillator input to the main output amplifier stage (A6). This should be 500MHz @ -3 to +4 dBm.? It is the input from the main oscillator to the dividers,filters and output amplifier. I believe this checks whether an appropriate signal is entering the A6 board but it not leaving it correctly.
This is the stage at which I posted my original question regarding probing the trace as the manual suggests doing this with an SA and I've never used a probe on my SA.
thanks
Tony
|
Hi, my guess would be that the high frequency loop is not locking. Instead it sweeps over its range 490 to 990 MHz over and over again. There are three phase-lock loops in this generator. Try to locate the respective tuning voltage for each loop and you may find if any loop is not locked. It is far easier to measure the tuning voltage using a scope compared to measuring a 500 MHz signal. And from there you can continue searching for what is failing.
G?ran
|
KB2LMN
Paul Bicknell
G?ran Krusell