Re: HP 8566A large amplitude error in 2-22GHz range
No worries! I very much appreciate the help.
The behavior is sporadic. The first time I tried that (all the settings you gave). It jumped around at 90MHz span. Then it stopped and stayed more or less stable. Stepping down kept it pretty close to center. Then it would disappear entirely at either 10MHz span or 5MHz.
|
Re: HP 8566A large amplitude error in 2-22GHz range
Does the cal signal jump only when you change frequency span or does it jump if you just sit and watch it.
Also, lets disable the automatic bandwidth management. Try the same zooming expiriment with Res BW set to say 1MHz and video BW set to 100KHz for example. Part of the jumping may be due to the bandwidth filters being off freq - so lets eliminate them.
Sorry, I should have told you all this B4
Cheers!
Quoting huszaghmatt@...:
toggle quoted message
Show quoted text
Bruce, there are jumps at the same settings in your experiment. I've captured this in the following two images
The jumps don't last long, so it was difficult to capture in a photo, but there it is. The signal doesn't shift much on the first step down in span, but I completely lose it on the 2nd step down.
|
Re: HP 8566A large amplitude error in 2-22GHz range
Bruce, I also did your experiment starting at a wider frequency range, before your 2nd note. I think it might be worth including because the behavior is peculiar. Although, maybe this is what you were saying is not meaningful. Specifically, the peak frequency, which appears to be a spur, moves significantly in the first shift down in span.  
|
Re: HP 8566A large amplitude error in 2-22GHz range
Bruce, there are jumps at the same settings in your experiment. I've captured this in the following two images   The jumps don't last long, so it was difficult to capture in a photo, but there it is. The signal doesn't shift much on the first step down in span, but I completely lose it on the 2nd step down.
|
Re: HP 8566A large amplitude error in 2-22GHz range
On Tue, Dec 8, 2020 at 03:58 AM, Raymond Domp Frank wrote:
ISTR there's an el. cap. on the YIG drive module itself that was bad (leaky or low cap.) in one story I read, causing hunting and phase noise in the YIG. Could be a completely different SA model though.
I think my memory is coming back: It probably was a Tek SA and the C was there to prevent noise from reaching the Yig's "fine" current coil. About the size that Stephen H. mentions: 1 x 2.5 cm. Raymond
|
Re: HP 8566A large amplitude error in 2-22GHz range
PROBLEM - With the CENTER FREQUENCY set to 100 MHz you should keep the FREQ SPAN to 100MHz or less - the fct that the span is 2G is what is causing the apparent harmonics. Set Center freq = 100MHz, Freq span = 90MHz and then try the zooming epirment
Cheers!
Quoting huszaghmatt@...:
toggle quoted message
Show quoted text
|
Re: HP 8566A large amplitude error in 2-22GHz range
Lets try another experiment to evaluate the sweep DAC. Assuming the cal signal is actually 100MHz (Good idea -You should count it, but I expect it to be accurate) - BTW, have you counted the 10MHz standard in the SA or ar you using an external 10MHz source (recommended to have the counter and the SA using the same reference). It could be that the DAC tuning voltage is off causing the SA to think the cal signal is at 78 MHz (when in fact, it is 100 MHz). So try the following: CENTER FREQ = 100MHz; Marker PEAK SEARCH (marker should be on the calibrator freq); MKR--> CF (Now the the marker should be centered on the screen) - Now repeat the "zooming" in by reducing the FREQUENCY SPAN and let see if the CAL signal stays centered - watch for any position "jumps" as they may indicate DAC failures.
Cheers!
Quoting John Miles <john@...>:
toggle quoted message
Show quoted text
Good point there as well. The 100 MHz calibration signal comes from the same 10 MHz -> 100 MHz reference chain that drives everything else, so it's worthwhile to double-check it if you haven't already. If you were previously able to run the shift-W procedure successfully, it is probably OK, but still worth confirming. I've seen the 100 MHz crystals fail before.
Aside from that, hopefully fixing the YTO unlock condition will take care of the amplitude response problem as well. The four different segments in the 2-22 GHz band correspond to successively-higher LO harmonics (which is why the baseline floor has to be stepped up by several dB at the band boundaries.) In order to help the PLL lock quickly, the YTO is pretuned with a DAC to within a few MHz of the target frequency. There are pretune adjustments for the beginning and end of the 2-6 GHz fundamental range, where the math amounts to a classical Mx+b line equation -- meaning that there is an offset adjustment that establishes the beginning of the sweep near 2 GHz, and a scale adjustment that establishes the slope required to reach 6 GHz at the end of a full-band sweep.
Suppose that the scale adjustment is OK but the offset adjustment is out of whack. The YTO might still be close enough at the beginning of bands 1 and 2 to yield an IF that falls within the resolution bandwidth passband, explaining why the response appears OK there. But as higher LO harmonics come into play in bands 3 and 4, the offset error is multiplied as well, becoming large enough to position the IF several dB down on the RBW filter skirt.
In that case, fixing whatever is causing the pretune offset error will take care of the amplitude response problem, with no need to worry about the baseline adjustments or other aspects of the YTX calibration procedure, which IIRC is over 100 steps long.
-- john, KE5FX
From: [email protected] [mailto:[email protected]] On Behalf Of amirb
Sent: Monday, December 07, 2020 6:47 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8566A large amplitude error in 2-22GHz range
...As I mentioned before, a simple test is to make sure your Cal out signals is rock solid on another SA
that most likely rules out any problem in reference section.
|
Re: HP 8566A large amplitude error in 2-22GHz range
Those look fine, except for the frequency error.? Good chance that fixing the YTO unlock condition will bring the RF section back to 100%. ? -- john, KE5FX ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of huszaghmatt@...
Sent: Monday, December 07, 2020 7:24 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8566A large amplitude error in 2-22GHz range? Here's trying to zoom into 100MHz.
|
Re: HP 8566A large amplitude error in 2-22GHz range
Last image zooming in 
|
Re: HP 8566A large amplitude error in 2-22GHz range
Good point there as well.? The 100 MHz calibration signal comes from the same 10 MHz -> 100 MHz reference chain that drives everything else, so it's worthwhile to double-check it if you haven't already.?? If you were previously able to run the shift-W procedure successfully, it is probably OK, but still worth confirming.? I've seen the 100 MHz crystals fail before. ? Aside from that, hopefully fixing the YTO unlock condition will take care of the amplitude response problem as well. ?The four different segments in the 2-22 GHz band correspond to successively-higher LO harmonics (which is why the baseline floor has to be stepped up by several dB at the band boundaries.)?? In order to help the PLL lock quickly, the YTO is pretuned with a DAC to within a few MHz of the target frequency.? There are pretune adjustments for the beginning and end of the 2-6 GHz fundamental range, where the ?math amounts to a classical Mx+b line equation -- meaning that there is an offset adjustment that establishes the beginning of the sweep near 2 GHz, and a scale adjustment that establishes the slope required to reach 6 GHz at the end of a full-band sweep. ? Suppose that the scale adjustment is OK but the offset adjustment is out of whack.? The YTO might still be close enough at the beginning of bands 1 and 2 to yield an IF that falls within the resolution bandwidth passband, explaining why the response appears OK there.? But as higher LO harmonics come into play in bands 3 and 4, the offset error is multiplied as well, becoming large enough to position the IF several dB down on the RBW filter skirt. ?? ? In that case, fixing whatever is causing the pretune offset error will take care of the amplitude response problem, with no need to worry about the baseline adjustments or other aspects of the YTX calibration procedure, which IIRC is over 100 steps long.? ? -- john, KE5FX ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of amirb
Sent: Monday, December 07, 2020 6:47 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8566A large amplitude error in 2-22GHz range? ...As I mentioned before, a simple test is to make sure your Cal out signals is rock solid on another SA
that most likely rules out any problem in reference section.?
|
Re: HP 8566A large amplitude error in 2-22GHz range
First two images attempting to zoom to 100MHz signal.  
|
Re: HP 1331A XY - service manual?
On 12/7/20 9:56 PM, Toby wrote: New 1331A manual is posted. ? Umm, posted where?
Here:
via
Ahh. Thanks! -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
Re: HP 8566A large amplitude error in 2-22GHz range
|
Re: HP 8566A large amplitude error in 2-22GHz range
It looks like those images didn't send. Sending them 1 at a time. Here's the startup 
|
Re: HP 8566A large amplitude error in 2-22GHz range
Raymond,
Since this basic lineup is used in several models I¡¯ll tell you about my findings. ?I found a cap on one of the boards (And yes I don¡¯t remember which) on one board the caps + lead was melted through on the other it melted the trace. ?As I remember it was a smaller cap abt 3/8 x 1 or so.
Also thanks for the tip I have a 8672A coming in with high phase noise so I¡¯ll checking that for sure
THANKS! Regards, ? Stephen Hanselman Datagate Systems, LLC
toggle quoted message
Show quoted text
On Dec 7, 2020, at 18:58, Raymond Domp Frank <hewpatek@...> wrote:
?On Tue, Dec 8, 2020 at 03:47 AM, amirb wrote:
Assuming the LO level driving the harmonic mixer is OK (the correct level must be in the service manual)
then I think the YTF is not tuned correctly, either the YTF itself is bad which I dont think it's the case here
or again its driving circuit is bad.
A bit of a wild guess, since I don't know these SA's very well, but ISTR there's an el. cap. on the YIG drive module itself that was bad (leaky or low cap.) in one story I read, causing hunting and phase noise in the YIG. Could be a completely different SA model though. Raymond
|
Re: HP 8566A large amplitude error in 2-22GHz range
Also, John, sorry the order of the pictures is reversed. The one you asked about is last.
|
Re: HP 8566A large amplitude error in 2-22GHz range
It's worth reiterating that I saw a clear cal signal at 100MHz and -10dBm yesterday feeding the cal into the input. That's quite a sudden change that doesn't sound like a calibration issue. Also, the YTO unlock has been getting more and more frequent.
|
Re: HP 8566A large amplitude error in 2-22GHz range
John, here's an image right after start up with nothing connected. I believe it's the same as Siggi's post (apart from the unlock):  Here are also a series of images looking at the cal signal. Clearly it's way off and I wouldn't have expected all those spurs, but then that's just a guess. The frequency as reported by my oscilloscope is as close to 100MHz as I can measure and the output power is reading -10dBm. Obviously, because I'm seeing the signal at 80MHz, I couldn't zoom in to far to 100MHz. But, when I started zooming in to 80MHz, I could still see the signal clearly at a 10MHz span, but at 5MHz span it disappeared entirely. Hope this is helpful and happy to keep posting these images.    Also, it's good to hear that I should be checking voltages and caps before tuning anything. I think I may have misunderstood some of the previous notes.
|
Thanks for the pointer to the 1977 catalog online -- got it.
|
Re: HP 8566A large amplitude error in 2-22GHz range
Basically, the slope of the response in bands 3 and 4 (i.e., 12 GHz and up) is more than I'd expect to see from trimpot adjustments that have aged slightly out of spec.? Something is causing you to lose several dB of (apparent) gain at the beginning of the third and fourth bands.? The trimpot(s) in question may just be intermittent, or there could be more to it... but unless someone went nuts with a diddle stick, this won't be an adjustment issue alone. ? The 'YTO unlock' message, on the other hand, could potentially be a simple adjustment problem.? Once you have a counter or other means of following the alignment guidelines -- and after you've checked the caps and power rail voltages -- ?I'd suggest going through the YTO pretune calibration procedure in the manual and see how far you get. ?If you reach a step that can't be completed, that will tell you a lot about where to look for the problem. ? When you look at the screen right after power-up, does the display resemble Siggi's post from 1:29 PM?? See if you can post a photo comparable to his, with nothing connected to the input. ?? ? -- john ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of huszaghmatt@...
Sent: Monday, December 07, 2020 6:29 PM
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8566A large amplitude error in 2-22GHz range? John, thanks for the info!
The signal is still there actually. I can see it if I connect that output to an oscilloscope. It just doesn't show up when I do recall 8, which I guessed might be another symptom of the YTO unlock issue. But, take my guesses with a grain of salt since they're not very well informed at this point...
Would you mind clarifying why you believe this is a component failure, rather than just a tuning issue? I very much appreciate your concern that if I tune these and that turns out not to be the issue that may end up being a lot of work in the wrong direction. In any event, it seems like checking the caps is a harmless place to start.
|
Dave McGuire
Rick - WA6NDR