@7:25 an increase in level with an increase in frequency is usually a sign of a broken connection such as a pin etc. where there is a small gap in the transmission line.
Based on testing the break must be in the A10 module. I agree the frequency response looks like a broken connection. I'm thinking about taking the A10 module apart even if I can't fix it could be interesting to see inside.
only at 9:00 into the video, but increasing in accuracy at higher frequencies is usually a crack in the input coax (or semi-rigid coax), and now it's essentially AC coupled across the crack, with the coupling getting better as the frequency goes up. I've seen The Signal Path fix a few instruments with the same problem. It's about the best 'problem' you can ask for!
The coax cables have been tested and are ok. Yes a broken coax will attenuate low frequencies more than high frequencies as you mention, I was thinking that might be the case but no such luck.
@@EriksElectronicsWorkbench I have a broken 8593A I picked up recently for relatively cheap I need to dig into, and my 8594E (that I've had for a long time) power supply crapped out entirely a while ago - I haven't had the chance to really dig into either of them yet. The 8593A seems to have a problem with the -15V rail. I'm just hoping someone didn't shove a bad YIG into it or generally "parts mule" all the good stuff out and shoved a bunch of busted stuff into it, and pawn it off!
Be careful as there is one special coax cable in the high band section which is in black heatshrink which is indeed a filter and just looking like a coax… gave me headaches when I was repairing my trusty 8562a…
why did you open the case at 15:00? Would it be a valid test to switch to DC coupling and to switch manually to different attenuations, while applying the known signal? Or would that not tell us much?
I needed to connect to the output of the attenuator so that it could be tested with the power meter. For this troubleshooting I could no longer rely on the spectrum analyzer itself for measured values as I had to start isolating components.
I know, as you said, there is a cap, in the 'front end' that 'blocks' the D.C. ... but, isn't it better to use a "D.C. block" anyway. Also, if you are working with that rf power sensor, why are you not using an attenuator on it?
At times I do use a DC block when using the spectrum analyzer for normal testing purposes but also depends on the situation and what is being tested. I agree it is a good precaution. As you saw I use one later in the video.. absolutely needed then because the input cap was bypassed. The RF power sensor was being used well below the max power handling range so no need for any attenuators.
There are tech service manuals listed here www.keysight.com/us/en/product/8561E/portable-spectrum-analyzer-30-hz-to-65-ghz.html#resources but they do not have the electrical schematics. The only place I know to get the CLIP documents (component level information packet.. the schematics) is artekmanuals.com/manuals/hp-manuals/ but they do charge for them. $25 part number 5967-8582.
Good luck finding the spare! We hope to see you soon with the working one.
@7:25 an increase in level with an increase in frequency is usually a sign of a broken connection such as a pin etc. where there is a small gap in the transmission line.
to be honest I don't like your entertaining behavior and that's why I'm checking your channel sporadically 👽 But you are right 😉
Every time when the attenuation is the function of frequency it's 99% connection problem 😈
Based on testing the break must be in the A10 module. I agree the frequency response looks like a broken connection. I'm thinking about taking the A10 module apart even if I can't fix it could be interesting to see inside.
@@EriksElectronicsWorkbench but anyway thank you for this detail analysis sequence it's great! And thank you!
@@bfx8185 thanks for watching!
only at 9:00 into the video, but increasing in accuracy at higher frequencies is usually a crack in the input coax (or semi-rigid coax), and now it's essentially AC coupled across the crack, with the coupling getting better as the frequency goes up. I've seen The Signal Path fix a few instruments with the same problem. It's about the best 'problem' you can ask for!
Hmm, that sucks that it's the YIG, and not just a cracked coax - that's about the WORST problem you could have!
@@gorak9000 lol yeah I agree
The coax cables have been tested and are ok. Yes a broken coax will attenuate low frequencies more than high frequencies as you mention, I was thinking that might be the case but no such luck.
@@EriksElectronicsWorkbench I have a broken 8593A I picked up recently for relatively cheap I need to dig into, and my 8594E (that I've had for a long time) power supply crapped out entirely a while ago - I haven't had the chance to really dig into either of them yet. The 8593A seems to have a problem with the -15V rail. I'm just hoping someone didn't shove a bad YIG into it or generally "parts mule" all the good stuff out and shoved a bunch of busted stuff into it, and pawn it off!
@@gorak9000 Sounds like power supply troubles on both. At least start there on the 8593A and hopefully the remainder is in good shape.
i have found i always need a better spectrum analyzer to fix a broken spectrum analyzer :D
yes very true!
Remember to check the rigid coax from between the attenuator and the filter, the fault could be that.
Yes coax cables have been tested and are ok.
@@EriksElectronicsWorkbench that is a shame
Be careful as there is one special coax cable in the high band section which is in black heatshrink which is indeed a filter and just looking like a coax… gave me headaches when I was repairing my trusty 8562a…
Are you sure, you are not Mr. Carlson. You sound...despite the mic... just like him.
lol no relation
Lolll...the breath before speaking and the cadence are almost identical.
why did you open the case at 15:00? Would it be a valid test to switch to DC coupling and to switch manually to different attenuations, while applying the known signal? Or would that not tell us much?
I needed to connect to the output of the attenuator so that it could be tested with the power meter. For this troubleshooting I could no longer rely on the spectrum analyzer itself for measured values as I had to start isolating components.
I know, as you said, there is a cap, in the 'front end' that 'blocks' the D.C. ... but, isn't it better to use a "D.C. block" anyway. Also, if you are working with that rf power sensor, why are you not using an attenuator on it?
Okay, you mentioned you have a D.C. block on the sig gen. Usually, I keep one on the S.A. too.
At times I do use a DC block when using the spectrum analyzer for normal testing purposes but also depends on the situation and what is being tested. I agree it is a good precaution. As you saw I use one later in the video.. absolutely needed then because the input cap was bypassed. The RF power sensor was being used well below the max power handling range so no need for any attenuators.
Where can I find an electrical circuit for this family of spectrum analyzers?
There are tech service manuals listed here www.keysight.com/us/en/product/8561E/portable-spectrum-analyzer-30-hz-to-65-ghz.html#resources but they do not have the electrical schematics. The only place I know to get the CLIP documents (component level information packet.. the schematics) is artekmanuals.com/manuals/hp-manuals/ but they do charge for them. $25 part number 5967-8582.