#43 - HP 8590A spectrum analyzer repair - part 1 of 3

A good old HP 8590A spectrum analyzer, which was picked up from eBay for parts or repair. It turned out to have a bad mixer in the first converter. A matched pair of schottky diodes was dead. A replacement was found and installed. The analyzer seems to be back in business. Perhaps, it will need some adjustments though.
Forum:
www.eevblog.com/forum/repair/h...
Service manual:
www.keysight.com/upload/cmc_up...
Avago (Broadcom)HSMS-28xx datasheet:
www.mouser.com/ds/2/678/AV02-0...[1]-908864.pdf
Diodes on Mouser:
HSMS-2802 - the ones I used:
www.mouser.com/ProductDetail/B...
HSMS-2812 - the ones I think I should have used:
www.mouser.com/ProductDetail/B...

[@PapasDino]

Nice fix! Hopefully you'll replace the diodes with the other family units in the future and share the result with us.

[@cambridgemart2075]

Just doing a similar repair on an 8590B and the Avago HSMS2822 is the diode used in the 1st mixer, so you made an HP / Agilent approved mod!

[@mechmania8450]

Glad your YTO was OK.

[@andypalm7061]

This was very interesting. Good info here. I am in the market for a an older unit I can learn from. Thanks for the great vid.

[@ruhnet]

Thanks for this great video!! I always enjoy seeing people troubleshoot RF circuitry.

[@perrooceaniko2005]

that is 50% knowledge 25% accute common sense and 25% art !!! don't you agree ?

[@smash7777-]

Nice repair video. I didn't have a clue about the instrument but was still nice to watch. Thanks for sharing.

[@DIY-electronics-automation]

Clear and concise description. Thank you

[@uploadJ]

Excellent! Have 8591A that still works, plus 141T with the 8555A RF plug in series too.

[@KerryWongBlog]

I was betting that you would eventually get yourself a spectrum analyzer. Nice work!

[@feliccieandorro408]

by the way, wonderful this repair .. smart thinking on the SMD DIODES !! I guess is BETTER than original solution !!

[@keithnoneya]

Love the way you think with alternative parts. I keep telling my fellow workers at work. Just because it calls for a specific part doesn't mean that's the only part that will work. When they wrote the manual they had to put SOMETHING that met the specs in there. If we don't have one check the specs and go find a part that meets those specs or close to it. Then install it and test it, if it works great, if not keep looking! Thanks for the the 3 video series, I watched all three. Now it's time to cut open that attenuator and fix it. Best Wishes n Blessings. Keith PS where did you get your training? also if you run across some old HP programable brick attenuators and need a specific pad let me know we have some junk ones lying around at work. Kieth

[@ronjones4069]

Thanks for sharing you success!

[@egbertgroot2737]

Very nice to watch this!

[@fixthetempo7276]

Very very interesting! Keep going on!

[@netdudeuk]

Another great repair video thanks.Not far from a thousand subs.

[@danishnative9555]

Anyone else notice the P/N on the RT Duroid circuit board in the A4 1st mixer? HP 08558-20004. Needless to say, this early version 8590 has some things in common with the HP 8558B spectrum analyzer.

[@feedback-loop]

I believe that the first converter from 8558 analyzer is almost identical except it does not have the 6 dB pad in it.

[@howardhiggins9641]

Very enjoyable. Thank you.

[@wadehicks9270]

Great job

[@amikhimji]

Great video thanks

[@AECRADIO1]

CHECK BACK PANEL FOR THE 10MHz. CLOCK TO THE REFERENCE CONNECTOR. BOTH MUST BE CONNECTED.

[@uploadJ]

Good point.

[@rougersierralabinstruments3122]

Nice video! Thank you. I have an 8596E SA. Frequency calibration passed, Amp cal failed at low amp signal. Measuring cal peak with this SA, normal w/ correction removed; distorted w/ correction. Any advice?

[@jorgecasagrande5421]

Hello! I have one of these (8562B) for sale... It was unused for a long time and had a problem with the power supply when turning it on. It was working fine before that. I'll give you a good price the way it is. It is still possible to recover it for use or even for spare parts to repair others.

[@ToltecMerc]

Nice job finding replacement diodes considering you had little information to work with.

[@feliccieandorro408]

he was very smart rat on this substitution !! it is IMPOSSIBLE to get a more accurate match pair than that !!!

[@SureshKumar-nk2ok]

sir i need to remove crt and extend it to monitor pls do help us

[@chinonsoezeobi1687]

Do you have a contact or repair center for 8591A around the DMV area?

[@douro20]

Do you still have this instrument or did you pass it on when you got the 8563E? I believe this was one of HP's first fully self-contained digital sampling spectrum analyzers.

[@feedback-loop]

I sold this unit.

[@cavan5]

Hi. Your 3 part video was great and it has tempted me to have a go at a HP8590a repair that I have for some time now. It has a number of problems. First the monitor (TR-60S1A.Panasonic )has a voltage problems which is collapsing slowly the horizontal width on the display and further to that vertical lines that are masking the display of the signals on the screen. Can get the Hp schematics on the Keysight website but can't find the third party Panasonic monitor diagram... Can anyone point me in this direction ? My test equip is limited but would like to check out the monitor first.

[@feedback-loop]

www.ko4bb.com/getsimple/index.php?id=download&file=HP_Agilent/HP_8590_Spectrum_Analyzer/HP_TR-60S1A_CRT_Service_Manual_for_HP8590a_Spectrum_Analyzer_Service_Manual-panasonic_tr-60s1a_chassis_y21_sm.pdf

[@cavan5]

Many thanks for the link 👍

[@amikhimji]

great video thanks, I have Hp 8590A unit,it passes self cal and output cal is 300Mhz -20 dbm, but when i want to measure +18DBM 5VPP Sinewave with 50 ohms impidance atteched its giving me +7.9 dbm( not 18dbm) would you know what it could be? thanks Ami

[@barleybillagain7320]

You are compressing the first mixer. You should not use more than -10 dBm input. Use a 30 dB attenuator.

[@mikedillon729]

i'am curious. You seem extremely well versed. Where did you get all your experience?

[@feedback-loop]

From a lot of tinkering, I suppose. :)

[@mikedillon729]

hahaahahhahahaah, That's Great! I enjoy watching bring that stuff back to life!

[@hobbymcu]

I have been trying HP8920B Communication test set to test bandpass cavity filter, but unfortunately, I am unable to do it. Do you have any idea how to setup and use HP8920B to test the bandpass cavity filter? I would appreciate your help very much.

[@uploadJ]

Info should be in the operating manual for 8920?

[@davidricardogarciagarcia5994]

Hola amigo tengo un equipo analizador de espectro IFR 2399 tiene una Falla para arrancar envía un mensaje en la pantalla YIG Cal Fail que puedo hacer la tarjeta o circuito de 10 Mhz esta ok que posible daño tengo por favor ayuda saludos

[@Samucas81]

Hello, I need firmware for my hp 8594A, could you help me? Thanks.

[@k1mgy]

Although the diodes in my 8590A seem fine, the 6db attenuator seems a bit weird. The attach point between the 147 ohm and the 38 ohm resistors measure a short to gnd. So do both of the diodes (on the left side nearest the filter). Lifted resistors and they're measuring fine. Anyone have a schematic of this converter? Not in the service manual, unfortunately.

[@feedback-loop]

How did you test the diodes? Did you lift one of them? I don't think it is possible to test them in circuit in this case.

[@k1mgy]

Yes, I tested these in circuit. So I just lifted one end and tested again. Both diodes test OK (using Fluke 87 in diode test mode, which shows approx 0.310 V junction forward biased). What is weird is the fact that the lower 147 ohm resistor in the 6db pad (assume here it's a PI network pad) is at GND. I lifted the resistor and its fine.

[@k1mgy]

Indeed the issue with my 8590a was the 1st converter... I think Feedback Loop has perfected an excellent repair technique for the 1st converter. Congratulations!! I bet there are bunches of these old analyzers which could be brought to life! Replaced diode with HSMS-2802-BLKG, although also had the HSMS-2812, but elected to try what worked. Also touched up soldering in the 1st converter, then reinstalled. Both FREQUENCY and AMPLITUDE calibrations are now successful and confidence test also OK! BUT... ... calibration signal reads 301.7MHZ on the 8590a. Checked also with RF source at 146.000 MHZ which reads 147.380MHz on 8590a. Cal signal into very good frequency counter is 299.815052 MHz, so we know it's within spec. This was the case before fixing the 1st converter, too. Indeed the Frequency cal worked fine before... it was the amplitude cal which failed before the fix. Looks like everything is out 1.38 MHz. I thought the frequency calibration routine would line things up properly. Is there something else that needs adjusting to bring the received frequency in alignment with what's actually being applied?

[@k1mgy]

Here are some photos of various tests and cal data, plus a video of the confidence test. Note that the Video BW fail messages went away once the unit had warmed up sufficiently. goo.gl/photos/Nrfvn4ftE7b62npU9

[@feedback-loop]

Wonderful! I am glad it worked for you. As to the frequency accuracy, these old models are not very accurate. According to specs +-5MHz + 1% of span.

[@christopherlange4475]

You are a Rf brain surgeon! Amazing manipulating the rf voodoo magic monitor.

[@douro20]

He's mainly a software guy; I think he picked up RF along the way.

[@kaveag]

у нас стоит такой в лаборатории, но у него частота постоянно плывет, мне сказало что это вроде как нормально, потому что он у него вроде как не синтетизирован, вы не замечали такого на вашем ?

[@kaveag]

+kaveag I thought you speak Russian, because you have a little accent from a eastern Europe, but apparently it's not the case, anyway, i enjoy a lot your videos, very very interesting repairs you do, we could think that it's very easy to repair something when we are watching you, but i imagine it's not, it should take some time and require some serous background in electronic stuffs as well !

[@feedback-loop]

+kaveag You are right about the accent. I just didn't have a chance yet. Regarding the time base drifting, I am not an expert in this RF stuff (yet). This model is quite old and must be quite simplistic. I didn't see any precision frequency references there. So, yes, I would imagine it might drift all over the shop.

[@kaveag]

+FeedbackLoop Ok :) I have a question, do you know why in SA we often do up conversion then down conversion? Down conversion i understand, but for up conversion not sure .. is it to ease resolution bandwidth filtering ?

[@feedback-loop]

+kaveag The reason is that the intermediate frequency (IF) of the first mixer must be outside of the tuning range. So it is chosen to be higher than the highest input frequency. This model has 1.8GHz input option, so the IF must be higher than that. They picked 2.05GHz. I am not sure why this particular frequency. Perhaps the decision was affected by availability of some parts like YIG oscillators and low pass filters.

[@kaveag]

+FeedbackLoop Ah Ok! Thank you! I'm trying to understand how does a SA works ..

[@erenjaeger5344]

Do you still got the manual? The link in your description is unfortunately not valid.

[@feedback-loop]

This video is almost 6 years old. Try Google. It usually helps.

[@erenjaeger5344]

@@feedback-loop I did, but I can’t find a service manual for the HP8590A. You still got it?

[@feedback-loop]

​@@erenjaeger5344 drive.google.com/file/d/1yc3Sttqug-gUPVkei80fXkajMvTDQ2di

[@electronics_engine]

Hi

[@egbertgroot2737]

Can i find the HP8590A doc you are using on online?

[@feedback-loop]

there is a link in the description

[@egbertgroot2737]

@@feedback-loop Thankyou!

[@PlasmaHH]

Its always funny how the prices for such things differ around the world, around here you would have to pay 500EUR for it...

[@frac]

High frequency electronics is pure dark magic. Gold demonic symbols doing amazing things.

[@feliccieandorro408]

hello people … need a honest help from the experts here … need to buy a used spectrum analyzer .. I have 2 in sight .. one is like the one AGILENT on the video, 7ghz like 12 years old, LCD COLOR model and also, same price, also 7ghz, LCD color too, rodhe@schawrz … I am located in USA … what brand do you think is wise to buy .. ? both are 7ghz range, both color LCD .. .more or less same specs .... expert comments are welcome ...!!

[@PaulHigginbothamSr]

Looks like a partially successful repair and you can now use the instrument especially after recalibration. I am sure it will be good for about 3 years and will degrade faster than the other diodes. Both to one's originally in the unit and the ones there now. So without that other Agilent you will be unable to permanently fix your older analyzer. Will work for a while though.

[@feedback-loop]

How did you reach the conclusion that the diodes will degrade faster? Why 3 years?

[@Arek_R.]

I'm first!