Author Topic: HP8590L spectrum analyzer calibration signal issues (Read 4756 times)

Bjoern · « **on:** July 10, 2018, 06:16:05 am »

After I replaced two capacitors on the analog board I got the unit up and running. However the calibration routine seems to have issues with finding or using the calibration signal.
I made a YouTube video showing the issue and would very much appreciate your input:

Many Thanks !!

borghese · « **Reply #1 on:** July 10, 2018, 06:30:08 am »

Can you specify which capacitors have you changed?
Maybe some loose connector? The calibrator output is -30dBm.

Bjoern · « **Reply #2 on:** July 10, 2018, 10:44:19 am »

Hi and many thanks for the prompt reply !

I do not have a schematic of the instrument therefore its a bit hard to say which of the caps I changed without opening it up again.
When I bought it the +15v were down and I found that on the analog board two of the tantalum caps were shorted. I assume they are filtering the main 15v supply of the board. When I changed them to electrolytic the SA powered up.

Do you know on which board the calibration signal is generated ?
Would I have to find the calibration signal that is accessible from the back or the one from the front panel ?

Bjoern · « **Reply #3 on:** July 10, 2018, 11:46:08 am »

Luckily I found the schematics within a large manual.
I remember the Caps to be C40 and C50 (see schematics) which seems to be on part of the board that might cause the error.
Maybe the OP-Amps did suffer ?

What do you think ?

SaabFAN · « **Reply #4 on:** July 11, 2018, 12:18:05 pm »

I think your general idea where the problem is (Yig Coil-Drivers) is correct. An unmodulated signal shouldn't look like it does on your screen. Which means either the coil-driver itself, or the control circuits for the driver are bad / horribly out of calibration. Have you tried adjusting any potentiometers or measured them? What do the control-voltages for the YIG look like?

Btw. How are the Supply-Voltages doing? Voltage ok? Excessive Ripple? If one Tantalum has gone bad, there's a chance others are causing problems as well.

Also with the Cal-Signal supposed to be -30dBm, it's quite a bit lower on your screen. With strong harmonics (don't know if they are supposed to be there). I'd suggest you build a simple diode-detector (if you don't have any equipment to measure the cal-signal) and measure the output-power. Doesn't have to be precise, just a rough estimate to see if the signal is in the right area.

Bjoern · « **Reply #5 on:** July 11, 2018, 12:36:00 pm »

I am just a few days away from finishing my Signal Tracer project.
Its based on Carlson's Super Probe (Youtube) and I hope that it will help me find some faulty cap's.

borghese · « **Reply #6 on:** July 11, 2018, 08:19:53 pm »

The level of the calibration signal is -20dBm; my apologies.
Your analyzer settings do not allow you to view the 300MHz calibration signal; the start frequency is 400MHz; there is also a spurious signal at about 410 MHz.

Can you post a photo at full span?
This is my analyzer's screenshoot

Bjoern · « **Reply #7 on:** July 12, 2018, 02:16:32 am »

Hi attached is my screenshot, does it give any clues of were to search ?
I hope that my BNC cable connection on the back panel (RF in/out) is acceptable ?

borghese · « **Reply #8 on:** July 12, 2018, 05:32:52 am »

Hi
Have you a signal generator for testing the frequecy response? Also the local oscillator feedtrough is very low, check the amplitude and stability of LO. First check all power suppy.

Bjoern · « **Reply #9 on:** July 14, 2018, 09:37:25 am »

I checked all the voltages which are OK.
I checked the cal output on the backside which gives a nice 10Mhz signal and is OK as well.

Attached are more screenshots of the calibration process, I hope that it gives further clues what might be at fault.

Bjoern · « **Reply #10 on:** July 14, 2018, 11:44:54 am »

I am thinking, that maybe not the calibration signal has an issue but the input of the signal analyzer.
I connected my Signal Generator and fed a simple 10 Mhz none modulated signal into the SA input.
Attached are the screenshots which show that the unit definitely has issues.

Probably these issues do not allow a proper calibration, or do you think that the misreadings are due to a failed calibration ?

Where could I start searching, are there modules I could skip (inside) in order to sort of trace my 10 Mhz signal ?

borghese · « **Reply #11 on:** July 14, 2018, 02:38:54 pm »

Can you post a screenshot with start -5MHz e stop 45MHz and center 10MHz, span 100khz and RB 3khz? Also you have a lock off message

Bjoern · « **Reply #12 on:** July 14, 2018, 02:57:22 pm »

I tried to match your request as good as possible but some settings change others.
I start by setting the center f to 10Mhz when I set the start -5 and end 45 then the center has changed.

The attached is a close match.

What is the RB 100khz setting I cant find it (what does RB stay for ?).

I noticed the Lock Message but I am not sure why it is there or how to get rid of it.
Hope the photo helps identify the error.
Highly appreciate your support !!!

Bjoern · « **Reply #13 on:** July 14, 2018, 03:27:37 pm »

Sorry I noticed that the reference BNC cable on the back was disconnected.
I did the same test again with the signal connected at the back.

borghese · « **Reply #14 on:** July 14, 2018, 03:46:21 pm »

The signal dont look bad.
Start -5 stop 45 is the same as center 20 span 50. RB is resolution bandwith under the button SWP/BW.
New setting:
Centre 20 span 50. AT 50 dB; take a shot.
Centre 10M, span 100khz; screenshot. Wait 60 minutes or more for a warmup
You can say tanks with the button in the right upper window

Bjoern · « **Reply #15 on:** July 14, 2018, 03:55:59 pm »

OK I will do the other measurements after warmup,
can I ask why you think the signal doesnt look bad, I connected a 10 Mhz signal gen (sine wave, no modulation) to the input,
I would expect low (no) noise floor and a sharp peak at 10 Mhz instead I see all this mess ??!!

borghese · « **Reply #16 on:** July 14, 2018, 04:31:04 pm »

The image of post #12 look good.
Try with a signal generator a 10MHz O dBm. SA setting AT 20 .
Seem as you have some problem with lock as there is a lot of modulation

Bjoern · « **Reply #17 on:** July 14, 2018, 04:36:48 pm »

I took your recommended settings.
For the 10Mhz center where we do not see the noise floor the image is not stationary, the spikes are moving from right to left.

Bjoern · « **Reply #18 on:** July 14, 2018, 04:42:54 pm »

Hi, I am not sure I can follow.
QUOTE:
Try with a signal generator a 10MHz O dBm. SA setting AT 20 .
Seem as you have some problem with lock as there is a lot of modulation

My Signal Generator doesn't allow amplitude settings in dBm (or do I set this at the SA?).
Which setting do you want me to set to 20 ? The Amplitude ?

Bjoern · « **Reply #19 on:** July 14, 2018, 04:46:48 pm »

This is the screenshot

Bjoern · « **Reply #20 on:** July 14, 2018, 05:05:31 pm »

I found this regarding the LOCK OFF msg:

φ
LOCK OFF appears briefly during
the CAL FREQ routine, during instrument preset, or when the
frequency value is changed; this is normal and does not indicate a
problem.

Bud · « **Reply #21 on:** July 14, 2018, 05:23:01 pm »

But it seems to stay and to me is an indication some PLL has a problem. You should look in the schematic where they have a PLL.

borghese · « **Reply #22 on:** July 14, 2018, 05:44:00 pm »

In the post #12 the level of the 10 MHz signal seems OK; maybe you also have problems with the attenuator or bandwidth filters and PLL of course

Bjoern · « **Reply #23 on:** July 24, 2018, 04:33:15 pm »

I did a CONF TEST and the following attached errors were displayed,
would that point me to the right board to look for hardware errors ?

borghese · « **Reply #24 on:** July 24, 2018, 08:22:39 pm »

When was the last time you saw the analyzer work or is it a recent purchase?
Have you checked that all the voltage are between the limits?
It seems to me that you are not very familiar with the analyser


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: HP8590L spectrum analyzer calibration signal issues (Read 4756 times)

Share me