EMI on FET probe

[hpw]

I deal currently with an EMI issue using a 1GHz Cal Test CT4121 FET probe. 50E input was selected on used scope.

1. I liked to measure the noise of a switched power supply as 12V battery to +5V & +/- 15V for audio as using in the 100mA range.

   - it showed some noise even without powering the switcher

2. Then simple shorting the FET probes (using shorted given connections) showed similar frequency burst at about 650kHz

   - switched other gear as in addition WiFi & LAN/WAN & lamps off  >> no influences

   - added a EMI blocker (black part on picture) what lowered the bursts about 30% only 

3. Then put the probe into a metallic case & top shielding all grounded to scope GND pin

   - it did not help

   - not grounding the shield picked additional hum

4. Even using TEK 2465B shows equal burst behavior at 650kHz


5. Using ordinary probes did not show equal signals

The beef questions: does someone have some Keysight or TEK or LeCroy FET probes with similar experiences?

Or and help how to get ride as may using symmetric probes?

While it looks like the probe cable detects any surrounding EMI

So now 

[2N3055]

Try adding more ferrites on the cable that goes to the scope (after amplifier).
It does seem that probe is not very well shielded..
You are powering from the battery ?

[hpw]

Try adding more ferrites on the cable that goes to the scope (after amplifier).
It does seem that probe is not very well shielded..
You are powering from the battery ?

Yes, using 9V battery for FET probe using USB power does not change much!

Also manufacturer claims amplifier EMI shielded & not using any switcher but the cables??

The benefit using the EMI block from input connection to amplifier was much larger...

So IMHO the cables is may the culprit / antenna

[T3sl4co1l]

Ferrite bead near the body, lowered SW-band noise by a whole 3dB?!  That's shockingly high, actually.

Is it more with the bead near the scope?  Does it improve with more beads?

Is the probe actually floating (no ground or signal connection, tips shorted as shown) or is there another connection somewhere?  -- Can you show the whole setup from probe to scope?

Does it vary with orientation i.e. is there any direction around the scope which has more or less intensity?  Sniff it out.  Also try sniffing around with the probe tip open-circuit (i.e. as a crude E-field probe).

The repeat rate looks plausible for any sort of SMPS, but the magnitude and ringing suggest terribly failed snubber and filter (or, completely absent).

Tim

[hpw]

Ferrite bead near the body, lowered SW-band noise by a whole 3dB?!  That's shockingly high, actually.

Is it more with the bead near the scope?  Does it improve with more beads?

Is the probe actually floating (no ground or signal connection, tips shorted as shown) or is there another connection somewhere?  -- Can you show the whole setup from probe to scope?

Does it vary with orientation i.e. is there any direction around the scope which has more or less intensity?  Sniff it out.  Also try sniffing around with the probe tip open-circuit (i.e. as a crude E-field probe).

The repeat rate looks plausible for any sort of SMPS, but the magnitude and ringing suggest terribly failed snubber and filter (or, completely absent).

Tim

Thank you for your input! 

I use now 3 beads, 2 at the end of the input part and one at the BNC connection parts. This is so far the quest mode. Also need with lower inner diameters.

The setup is POWER GND (1) to a star distribution connected a shielded metallic box as DUT too.

The DSO is connected from same power array as (1)...

the pickup-ed noise gets as the largest as soon I connect the FET signal pin to the metallic box...

in other words, it is not the ideal connection for measurement as the GND from to DSO should go to the star distribution.
So I may measure the switcher noise of the DSO against the GND.

Any better is, using an ordinary probe 1/10 what me 

Have to order new beads what requires 1..2 month 
 

[hpw]

Try adding more ferrites on the cable that goes to the scope (after amplifier).
It does seem that probe is not very well shielded..
You are powering from the battery ?

As my 1GHz Cal Test CT4121 FET probe is actually the EQUAL as TESTEK TT-AF 1200 brand...

In a later message may around 2020, you claimed to have TESTEK 1GHz active probe too 

In other words, does your active FET probe is sensitive to any radiations in the 1MHz regions??

Now seen on https://xellers.wordpress.com/electronics/1ghz-active-differential-probe/ having similar noise 

BTW: Is I see, the suppression blocks are for above 10MHz... 

[2N3055]

Try adding more ferrites on the cable that goes to the scope (after amplifier).
It does seem that probe is not very well shielded..
You are powering from the battery ?

As my 1GHz Cal Test CT4121 FET probe is actually the EQUAL as TESTEK TT-AF 1200 brand...

In a later message may around 2020, you claimed to have TESTEK 1GHz active probe too 

In other words, does your active FET probe is sensitive to any radiations in the 1MHz regions??

Now seen on https://xellers.wordpress.com/electronics/1ghz-active-differential-probe/ having similar noise 

BTW: Is I see, the suppression blocks are for above 10MHz... 

Sorry, I don't have that anymore..
I only have Siglent SAP1000.... If I had some time this week I will try to see how that one behaves for your use case ...From what i remember from previous use it is quite clean...

[hpw]

@2N3055: From what i remember from previous use it is quite clean...

Yes, as using for digital signals as 3.3V, looked good and any better with rise & overshoots 

Than looking for +%V or +/- 15V power switching I get even without battery & star grounding & probe shorted or opened and simple connected  to DUT GND those signals.

Also this Cal Test allows +/- 40V input may swings, so be careful with your...

I am still waiting for any Keysight or LeCroy owner measurements...

[hpw]

Well well,

as using a LeCroy WM 8600A I do not have/get those RF signals at all 

So best wishes to Siglent Turtles/Followers 

Hp