#46 - HP 8590A spectrum analyzer repair - part 3 of 3
Вставка
- Опубліковано 2 сер 2024
- In this part another problem with this analyzer was found, confirmed and fixed. It turned out that the input attenuator Wavetek 0955-0329 was defective in a strange way: the 20 dB pad in it attenuated lower frequencies only by 15 dB, but worked fine at higher frequencies.
Forum:
www.eevblog.com/forum/repair/h...
Previous videos:
Part 1: • #43 - HP 8590A spectru...
Part 2: • #44 - HP 8590A spectru... - Наука та технологія
You know, there are only a couple of channels that I follow arduously, TSP and Feedback Loop. Your approach and knowledge are above and beyond any other. And the quality of the videos makes me think it takes you a long time to produce. Thank you for this
I just finished part 3. Excellent job ! Seems like every time I buy a parts unit I end up fixing it and thus don't have a parts unit to harvest . It is just too painful to cannibalize instruments that only a few years ago we could only dream about owning. Your mixer repair is very encouraging as I have passed up some great units because of blown mixers. Thanks for all these great videos.
Thanks for a good video! It would be interesting to see you open up the broken attenuator, now that you got a new one.
Glad that the physical space inside the analyzer was big enough to accept the larger attenuator box! Nice fix, thanks for sharing!
Hello, a series of fantastic and very educational videos, we need channels like yours, please continue ... you are doing great, thanks so much for sharing with the technical community!
Congratulations on this analysis and repair! I watched the 3 part video series on this repair. There are many of these still for sale on eBay with 'issues'. I realize now that they can be easily diagnosed and repaired to save $$. I intend to give this a try because these T&M equipment is very expensive fully functioning and calibrated.
Enjoyed the series on this repair, Great job!
Re the attenuator, I had a defective one (suspect burned out with too much power) - the casing needs heat from a small torch to take apart, it does contain surface mount resistors which I replaced ok (I was surprised when I worked out it was 10-20-30dB steps and not binary !) I also had to change the diodes with what I could get hold of (about 15 years ago, I may try your modern replacement types) thanks for the good videos
Very enjoyable series. I too have looked for a part for some old piece of gear, only to end up buying another unit ;-)
You must repair test equipment as a business ? As much as you know about them and some things you said like you've seen certain things before making me think that's what you do lol. You're very good at repairing test jigs and that's such a blessing. Enjoying your videos always great content.
brilliant work great channel. ..regards from the UK ☺
in the manual, it says the LO output socket should also be terminated with 50R
Excellent! Got there in the end.
Before HP made digital spectrum analyzers they made ones with direct-view storage tubes. Quite an interesting thing.
Does the ESG-D2000A have a white electroluminescent backlight? If so the odd backlight colour may be a sign of aging.
You are correct about the E4406A being an FFT based unit, being a vector signal analyser and not a proper swept-LO based spectrum analyser. VSAs are typically cheaper than a regular spectrum analyser. BTW this is actually one of the last HP instruments to use a PA-RISC processor, and it runs HP-UX.
now that you have a working attenuator: will you try to open the broken one with force?
Are those SMA connectors on the attenuator in/outs?
Good job.
Now that you have a good attenuator, please consider dissecting the bad one. Very curious as to what's inside.
This is very cool! I think you are as obsessed as I am with fixing stuff like this. I considered purchasing another broken unit, but since I already have a broken Tektronix 2710, perhaps I should fix it, first.
Please share the details of your test results.
I performed the same tests and discovered that my attenuator is in fine shape. However, I am getting an occasional confidence test failure "VID-BW FAIL 100" and "VID-BW FAIL 30". Also, started doing some of the simple performance verification tests. The unit fails the Resolution Bandwidth switching test only at 100KHZ and by 1.08dB. It also fails the frequency drift test (out by 800KHz). Not sure where to go from here...
Unfortunately that is not how you verify the attenuator after install. Good fix and you verified it before putting it in.
great job from uk
great vidoes comrade
Thank you for great videos. Do you know how to shift the frequency back into the exact frequency, i.e. 1.002 GHz -> 1.000 GHz? It would be great if you could do a part-4 with this adjustment.
These old models are not that precise
Great video series, one questions for the replacements of the attenuator. The replacement analyser has 75 Ohm instead of 50, how does it come that the attenuator works for this in the 50 ohm environment.
Regards from Hannover Germany.
I did not explain in the video. The 75 ohm version of the analyzer has a75-to-50 ohm matching pad right at the input connector. The calibration output also has a small difference on the board to make it 75 ohm matched. Otherwise the analyzers are identical.
I purchased an 8590a on Evil Bay a year ago. The 10db step has been dinged from too much RF. Ironically, it already had the 0955-0453 attenuator installed, so I guess the original was replaced. All the other hardware is 50 ohm.The documentation on this early version machine is very poor.
That unit with the good attenuator was a great bit of luck !...it's been a long time, did you ever get curious and open the bad one ?...cheers.
I'm going to try your trick of buying some expensive piece of equipment that doesn't work cheap on Ebay and fixing it so I found a Boonton 9200A RF Millivoltmeter for $22 that normally costs hundreds of dollars. Maybe it works and just needs cleaning up, wish me luck! Do you have any experience with that model?
I have never seen one. I wish you luck.
Thanks. BTW, it doesn't come with any probes and they are very expensive but the schematic in the manual makes them look easy to make with just a couple of resistors, caps and diodes. It doesn't give any part numbers though so I wonder what kind of diode would be best to use? It probably needs something better than a regular 1N34A and preferably surface mount if possible. Got any suggestions?
I really don't know.
Great set of videos on the HP 8590 repair!
In reference to the Boonton question posted by steve c, there is some info on the eham.net site that might be useful at this link:
www.eham.net/ehamforum/smf/index.php/topic,22475.0.html
Seems that the original diodes were somewhat of an exotic item, however the info should be of use in seeing if a modern replacement might be a workable choice for hobby / experimental use.
I wonder why the salvaged attenuator was not 75 ohm because the unit had 75 ohm impedance.
Yeah, I did not explain this point very well in the video. The 75 ohm analyzer is different in just two places: there is a 75-to-50 ohm matching pad right at the front input, and the third converter module is slightly different (a couple of passive parts have different values) so that the calibration output from it is 75 ohm matched.
I just thought it must have a wideband matching (from 10KHz to 1.5GHz, so it's over 4-5 decades of frequencies) and eventhough having a wideband matching with a low impedance ratio of 1.5 is not a huge challenge still I was expecting the whole front end being designed for 75 ohm
So, when are we going to see more of these excellent repair videos specially test equipment. I just watched this one for maybe the 4th time...
Crack open that old WaveTek attenuator. Maybe find some one with a milling machine and face off on side. We are all curious to see what $200+ used ebay bucks get's you?