Assistance with HP853A and 8559A spectrum Analyzer

[AdrianH]

I wonder if anyone here is familiar with this spectrum analyzer. The unit works good enough for me but one thing that I have noticed is that although the bottom range is 0.01 (10 MHz) to 3 GHz it does not start to display until the input signal is above approx 100 MHz.

The frequency display tallies with the input frequency and the span seems OK , but things like the 35Mhz at -10dBm can not be used for reference etc.

Could it be a simple knob/control tweak?

Adrian

ps I should add that if I select the Alt IF function then I need to be around 120 MHz before I see any signals?

[Radio Wrangler]

Do you get the normal big response at zero hertz (no input applied)?

Can you set up a span which goes a few dozen MHz negative IE to the left of the zero hertz response?

David

[AdrianH]

Hello David thanks for responding

I can get the response of zero although it does seem smallish?

Here is a picture, 10MHz per div -30 ref level no input looking at zero MHz With the frequency display and the centre of the screen matching -9 Mhz is the furthest I can get to, if that makes sense?



Cheers

Adrian

[Radio Wrangler]

Yes, I agree, that looks quite weedy. You should be seeing a strong gaussian shaped response large enough too see the trace go clean without the raggedyness of noise on the upper parts.

Microwave spectrum analysers are like their 2GHz-ish brethren with extra stuff grafted on. The lowest band is an up-converter with a lowpass filter in the RF path handling mixer image. So there's an IF in the2-3GHz region, that gets down converted to about 300MHz and thence to 20 or 21.4MHz.

when you select a higher band it becomes a down converter with a different mixer. A yig tuned bandpass filter handles images, and the microwave mixer is switched into the 300 ish MHz IF that already exists.

The same YIG LO is used, and higher bands work from harmonics of the YIG LO, with the RF yig filter preselecting the wanted band and fixing images.

So the LO YIG feeds a distribution amplifier/splitter and feeds both the lowband mixer and the high bands mixer.

Having the zero response there at all and in the right sort of place, shows the YIG LO is going, and on roughly the right frequency.

The zero response is the YIG LO crossing the high first IF. So not enough LO is leaking into the IF at the first mixer, or the IF is deaf. he zero response needs no RF path working, so that rules out the input attenuator, switch and filter. B(e)ggered mixers can leak more, unless both diodes are ruined by someone transmitting into the input. The LO distributor could be giving low power, or the 2-ish GHz IF onwards could be deaf.

To diagnose, you're either going to need another microwave analyser or a microwave signal source.

David

[AdrianH]

Thanks again David. Sounds like I could be in a bit of a pickle then, I have various bits of elderly test gear, but not another analyser or a microwave source, I have an old signal test set an R&S SMPF2 that can go to 1000Mhz and an old IFR 1000A sat on the floor, the R&S has not been powered in a while. Nothing that will really get me up in the 2 to 3 GHz region.

Just tried an old Racal 9915 counter and found it dead to the world.

Think I will need to do a post in request for services.

Adrian

[Radio Wrangler]

There's still a bit you can do.... checking all the power supply rail voltages, sticking a stub antenna on the input and seeing if the higher bands will find you any wifi or cellphone signals. If the lowband mixer is goosed, then the high bands could still be there, unless they toasted the input attenuator at the same time. It all helps you zero in a bit.

I've got most of the structures and frequency plans of HP's later generations of analysers in my head... I was on the design team for some of the variants. The 8559 and 8558 were just a bit before my time at HP.

David

[AdrianH]

If I stick a bit of wire in the input socket I can certainly see inputs on all bands, problem is some signals are there on all the bands no matter what button is pressed 1to 3 GHz 6 to 9 GHz etc. I think these are the local DVT transmitters such as Winter Hill etc

I can spot the 5 GHz and the 2.4 GHz frequencies or what I am assuming are them, The 5 GHz diminish somewhat when I turn off my router in the room. The5 Ghz are not there on the 0 to 3 GHz range.

I need to strip the case down a bit to get at the power supply test points so will do that tomorrow, I do not have any extender for the plug in and I have to pull the unit out of my desk to work on it. They may be on the back panel, not sure.

Adrian

[Cruisin Marine]

I have never used these units, but my old 141 Frame 8553/8552 etc. SA used to give very low 0Hz oscillator level if the rear plug in plugs or sockets needed a good clean and a re seating.

Worth a look perhaps.

[AdrianH]

I can clean it when I strip the case down.

Adrian

[AdrianH]

With the case top off I can check the 100 Volt, 15 Volt and -12.6 Volt rails are all spot on.

Cleaned the connector with a swab and Servisol, still same unfortunately.

Adrian

[Jon_G4MDC]

Did you get anywhere with this Adrian?

The fact you saw the same signals on all the different GHz bands is not a surprise. The 8559 does not have a YIG preselector filter so it's wide open. The band switch only changes the mixer bias to favour certain harmonics. I forget but it might alter the amplitude calibration too.

I have had the frequency setting pots (The fat blue Allen Bradley ones) drift in my 8559. It meant the LO YIG didn't tune where the front panel display indicated it should. Luckily it was just an anno domini effect. They just needed to be worked from end to end a bit to clean the track and pick up and then reset as described in the manual.
I see that about 10 years ago David, Radio Wrangler, pointed out that actually the display is only a digital Voltmeter reading the tune Voltage. That is exactly right and the pots I mention are what calibrates that.

I'm interested in the behaviour of your 853A please. Does it all work on both Analog and Digital modes? Do the trace and scale intensity controls do what they should?

[AdrianH]

Hello Jon; No I did not get any further with this issue, all I know is that I could set the frequency to match a synth signal generator, but that the amplitude on the display would drop considerably around 100 to 110 MHz input and nothing was seen below that frequency.

last time I tried it the 853A section was working fine, could store a trace display it and also have an analogue display on the B switch, or either way around. It is not something I use a lot. Just a nice to have. I like HP gear it is just big and heavy a bit like my 182C scope.

Adrian

[Jon_G4MDC]

They certainly keep our back muscles in shape Adrian and many thanks for the reply.
I am battling an HP853A with an 8558 Plug In so it is only capable of 1500 MHz.
It was working well but lately something bad has happened in the Display Unit.

I wonder if you knew that you can get digital self test screens by powering on while holding then releasing the Plot Grat button. My display is currently missing Z modulation (beam intensity) control so the screens are there but they are not being correctly written. I am trying to figure out how the beam intensity circuit works and what might be wrong.

An enquiry at HP/Agilent groups.io hasn't turned up any eureka answers yet so I'm trying to figure it out for myself. It might yet end up in the components and circuits section since I'm finding it hard to fathom it out. Glad yours works and long may it continue!

[AdrianH]

Jon if you were closer you could certainly borrow mine if it would help, doubt I would getting down your way any time soon though.

Adrian

[Jon_G4MDC]

It's true we are a bit far apart. I have a few strange things observed with mine on slow sweeps which would be good to test on another instrument. It's off topic here though

I also have an HP8559A plug in a 182T frame. If you think there are any useful comparisons that could be made with it please shout.

Off hand I can't think what might be the matter to cause that deafness at 100-110MHz. Input semi rigid or the step attenuator maybe? I wonder if you have tried going in at the 3dB Attenuator before the mixer - W2 in the attached?

[AdrianH]

Hi Jon, the deafness starts at around 100 to 110 MHz going down in frequency, so basically from 100 MHz down I see nothing rather that the 10 MHz which is spec. Above 110 MHz it is fine and flat, so either capacitive somewhere or things need aligning.

The analyzer is in the garage at the moment not beeing used, I need the space on my desk. It is missing the side panels as I think at one time it was rack mounted and someone decided to keep them.

I had it suggested not to put my 8559 into the 182C frame for some reason I can not remember.

If there is anything you would like to me to try and video for you let me know I am sure I could do something?

Adrian