Products > Test Equipment
DC Block for spectrum analyzer: minimum safe low cutoff frequency
<< < (3/4) > >>
bdunham7:

--- Quote from: joeqsmith on June 01, 2022, 12:19:14 am ---Attaching to a DC source can certainly induce a fast transient.   Same for removing it.

--- End quote ---

True, but a series capacitor won't do much about that.  The DC block in front of an SA or other instrument will prevent thermal damage from the slow transient only if it limits the total energy per event to less than the damage threshold.  To calculate that you need the capacitance of the DC block, the maximum DC voltage you expect the system to survive, the SA or device input impedance and the minimum Joules to explode it.
matthuszagh:

--- Quote from: bdunham7 on June 01, 2022, 01:31:29 am ---The DC block in front of an SA or other instrument will prevent thermal damage from the slow transient only if it limits the total energy per event to less than the damage threshold.  To calculate that you need the capacitance of the DC block, the maximum DC voltage you expect the system to survive, the SA or device input impedance and the minimum Joules to explode it.

--- End quote ---

Please correct me if I've misunderstood, but it sounds like this is saying that the event of a DC voltage applied to the input through the block (RC highpass filter) is short enough to assume the input circuitry approximately doesn't dissipate any energy as heat to its surroundings, so we only need to consider the total energy of the signal applied to it. If that's true, is it temperature rise that ultimately kills the input? If so, how would one determine the temperature threshold? If not, how do we calculate the minimum damage energy?
bdunham7:

--- Quote from: matthuszagh on June 01, 2022, 02:23:53 am --- If that's true, is it temperature rise that ultimately kills the input? If so, how would one determine the temperature threshold? If not, how do we calculate the minimum damage energy?

--- End quote ---

I suppose, in the same way that blowing any overloaded component usually involves temperature rise at some point, perhaps with overvoltage as an occasional exception.  But knowing the amount of energy it takes to fry the inputs of an SA would require knowing all of the details of the input circuitry.  For a 50-ohm oscilloscope input, it is just a 50R resistor that can probably take some surge and the input as a whole doesn't mind DC at all.  That's easy.  An SA with a 0VDC input spec probably has something close to the input like a diode mixer that can't tolerate any DC, or at least not much, from a low-impedance source. 

So to answer your original questions 1) you can't model anything as a 50R resistor here unless you know that is where the issue with DC is, like in an oscilloscope.  and 2) obviously you found one, but as to whether a 9kHz DC block is 'safe' for an SA with a 10MHz lower bound is unknown and would depend on the DC voltage applied in any case.  I would assume that a given block would be adequate for an SA with a minimum frequency  in the same ballpark--say a 5MHz block for a <10MHz SA--and a reasonable applied DC bias.  What's reasonable?  No idea.  Probably at least the maximum peak AC voltage.  And in many cases, probably a lot more.

Edit:  I just realized that the DC block you posted is SMA not Type N, so that's pretty amazing that they can fit a ~350nF capacitor in there.
bob91343:
It's almost always temperature rise that destroys input circuitry.  Exception would be if you are working with voltages high enough to cause arcing.

Even a zener diode can handle a large transient as long as its duration doesn't cause local overheating.  When a device fails from a transient, it's because local heating melted something.

I have some old analog power supplies that I use on the bench.  When powering circuitry with them, I disconnect the supply output before shutting it off because many of them have outputs that rise during turn off.  A bit off topic, but if you have a feel for what's going on, you can guard against some things.
jonpaul:
Decades ago a Solar LISN with dual L1/L2 capability via switch was coupled to a TEK SA 50 Ohm.


The SA front end mixer blew instantly as the switch applied a 1/2 Vline transient, with a large series cap perahps 470 nF.

Both a 1..10 MHz HP and transient limited are needed.

Check the SA specs for max V and max power, nomally 500 mW

Jon
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod