HP 8560e Spectrum analyzer input coupling

[xzswq21]

Hello, HP 8560e has an internal attenuator with the range from 20dB 0dB to 70dB for 30Hz to 2.9GHz and the Spectrum Analyzer supports from 30Hz to 2.9GHz.

* when I use AC-coupled does the SA support from 30Hz? or it supports from 100KHz?
* when I use internal attenuator the Spectrum Analyzer is in AC-coupled or DC-coupled?

for example I set the internal attenuator to 40dB, now the Spectrum Analyzer covers from 30Hz or 100KHz?

Thanks

[Ice-Tea]

You need to use DC coupled below 100kHz.

The attenuator won't make a difference I would think.

[xzswq21]

The attenuator won't make a difference I would think.

When I enable 50dB internal attenuation and the input signal is 1KHz it's important to know the answer of my question and we must be sure.

[Bud]

Check the device user manual, it must be in the specs section.

[xzswq21]

Check the device user manual, it must be in the specs section.

I checked the user manual I couldn't find it.

[Ice-Tea]

The attenuator won't make a difference I would think.

When I enable 50dB internal attenuation and the input signal is 1KHz it's important to know the answer of my question and we must be sure.

- I don't see a logic reason why the attenuator would make a difference
- The manual does not note there's a difference
- If that does not suffise: measure it.

[tkamiya]

When input is DC coupled, spectrum analyzer input amplifier does not tolerate any DC voltage.  Yours say 0.2V in specifications, but it's really zero.  You'd be foolish to risk applying any DC voltage to it.  If you blow front end, it will cost you just as much as the whole SA to repair it.

Use A/C couple as much as you can.  If you are new to this, you might even want to put on DC isolator on the input connector.

An only time you'd want to use DC coupling is when you want to measure very low frequency with limited amplitude, minimum of 30Hz and zero DC bias.  Remember, really slow AC is same thing as DC.   I've owned various specAn for few years now.  So far, I never needed DC input.  I actually have DC input (8568 has a separate input connector) zip tied, so BNC won't couple) disabled.

For such a low input frequency, FFT will be more suitable instruments.

Please be careful.  Oooops will cost you plenty!

[rf-loop]

How this can be unclear to OP. Every single detailed info is all available, only need read.

Also OP claim intput attenuator is 20 - 70dB. -- No.  It is 0 - 70dB.  So OP need read user manual.  RTFM ! 

This principle is also same in 8590E series front end.

And before you know enough about these instruments I will also recommend: do not set attenuator to 0dB !

Inside attenuator there is (of course first) AC/DC switch and after then switches for 10, 20 and 40dB attenuator chips but I do not remember these chips order, so long time I have done repair/overhaul last one.


RF front  simplified  block diagram.

[tkamiya]

OP is obviously new to SA.  It is a complicated, and delicate equipment.  Yes, everything is in manuals, but it's a daunting to read the whole thing and pick up the information you need.

Please be kind to him/her. 

[Electro Fan]

OP is obviously new to SA.  It is a complicated, and delicate equipment.  Yes, everything is in manuals, but it's a daunting to read the whole thing and pick up the information you need.

Please be kind to him/her.

+1

---

And xzswq21, as recommended above if you can, put an inline DC block on the input and leave it there at least until you are very confident in the operation of the SA.

[xzswq21]

And xzswq21, as recommended above if you can, put an inline DC block on the input and leave it there at least until you are very confident in the operation of the SA.

Most of dc block covers from 100KHz and above, I think I should build a DC block with a 1 or 10uF capacitor but these capacitor are X7R and not NP0 , so I think the capacitor makes some distortions at high frequencies

[tkamiya]

It's kind of important to keep your expectations in check....  It's not possible to cover the entire frequency range of your SA and keep absolute flat response.  Using capacitors, it's really not possible to expect a flat response from 30Hz to say 1MHz.  There are different technique to do that. 

At 30Hz, I tend to recommend being extra careful as DC offset starts to play a big part.  Use a too big of cap will risk damaging the input from discharge.  So either way, you are putting your SA at risk.  What I usually do when there is a least bit of doubt is to observe the signal using oscilloscope first, AC coupled, then DC coupled before connecting to SA.   

Realizing there will be limitations, what is your goal?  What frequency range is your interest?  What power level?  Placing an attenuator in front of input will afford you some level of protection, too.

[xzswq21]

At 30Hz, I tend to recommend being extra careful as DC offset starts to play a big part.  Use a too big of cap will risk damaging the input from discharge.  So either way, you are putting your SA at risk.  What I usually do when there is a least bit of doubt is to observe the signal using oscilloscope first, AC coupled, then DC coupled before connecting to SA.   

in my application, the frequency range starts from 1KHz or 10KHz but usually DC blocks covers above 100KHz, so I think I should build a customized DC block for myself, can I use NP0 capacitors?

[xmo]

"... can I use NP0 capacitors?"
-------------------------

How does a zero drift temperature coefficient affect the issue?

If you did build a DC block with a 1 microfarad capacitor, at 1000 Hz it would have a capacitive reactance of 159 ohms.  Don't you think that would have a serious impact on the accuracy of your measurements?

You said your sensor is 50 ohms.  Just make sure there is no DC on the device before connecting the analyzer.