Siglent 1104X-E and Leo Bodnar pulse generator causes WiFi problems

[807]

I have a wifi dongle in the back of my 1104X-E which I use mainly for taking screen shots on my pc using the web interface. It also connects to an NTP server at start up to provide accurate date & time. I noticed when connecting the Leo Bodnar 10MHz fast risetime pulse generator to channel 1 while viewing the web interface, that it lost wifi connection. Removing the pulse generator, or removing it's power instantly re-established the connection.

Connecting the pulse generator to channels 2 & 3 caused no problem, but channel 4 was sometimes OK & sometimes not. I've powered it through the pc USB port, a mains USB adapter & a power pack. All with the same results.

Anyone else notice this, or have any ideas as to why this is happening? 

[tautech]

I always power the Bodnar pulser directly from the scope USB port.
Have you tried ?

[807]

I didn't think of that. I'll give it a try later.

[807]

I tried powering the pulse gen from the scopes own USB, but still the same. I tried connecting the wifi adaptor to the front USB, but still no joy. I notice that the LED on the adaptor flashes rapidly when working. It flashes slowly when the pulse gen is attached (about once per second) indicating no data being received or sent.

[807]

OK. It looks like the scope is acting as an aerial when the pulse generator is connected to it. There must be some kind of leakage when connected to channels 1 & 4, but not when connected to 2 or 3.

If I hold the pulse gen about 1metre from the scope & touch the centre pin of the BNC on the pulse gen with a short screw driver, it kills the wifi connection. Of course, the fast rise time pulses will have plenty of harmonics, but I didn't think they would extend into the 2.4GHz band with enough strength to wipe out the wifi signal. The scope is only about a metre from the router.

I might probe around with the spectrum analyzer later.

[tautech]

Are you terminated into 50 Ohms ?

[807]

Are you terminated into 50 Ohms ?

Initially I wasn't, but I did try a 50 Ohm feed through terminator. It reduced the interference on wifi, but was still blocking it every now & again.

Plugging the pulse gen into the spectrum anayzer showed +4.8dBm @ 10MHz & -44dBm in the 2.4GHz band.

[ebastler]

In a way, I am surprised that there are not more problems with WiFi on-board in a scope. I mainly would expect problems going the other way: The WiFi transmitter interfering with measurements at low signal levels, causing spurious FFT frequencies etc. 

Maybe that's a reason why not many scopes support WiFi dongles? Even some models which are known to have the drivers installed don't let you configure WiFi, at least not officially. (Siglent's SDS2000X+ or Rigol's complete DHO series for example.)

[Calambres]

¿Perhaps using an USB extension cable to place the WiFi dongle somewhere else? 

[tautech]

Yes well who didn’t think a 30ps rise time pulse generator wouldn’t generate RFI ?

Maybe if it was in a metal enclosure RFI can be contained.
Hopefully I’ll get some chance to make some tests in a day or two when we return to base.

[807]

¿Perhaps using an USB extension cable to place the WiFi dongle somewhere else? 

That's something I haven't tried yet. I'll give it a go, out of interest.

It's not really a problem though. It was more of an observation, & confused me a bit when it first happened. I thought Windows had crashed on my pc 

If I want to take screenshots of the pulse, I can just use channels 2 or 3. For anything else, all channels seem OK.

[807]

Yes well who didn’t think a 30ps rise time pulse generator wouldn’t generate RFI ?

Maybe if it was in a metal enclosure RFI can be contained.
Hopefully I’ll get some chance to make some tests in a day or two when we return to base.

I tried putting the pulse gen close to the dongle. It was less than a centimetre before it blocked the signal, so I dont think it's being unshielded that's causing it. Perhaps it's something like the PCB track lengths being resonant at 2.4GHz on channels 1 & 4 inside the scope that's amplifying the signal? 

Did Dave ever do a teardown of this scope? Be interesting to see where the channel inputs go.

[tautech]

Be interesting to see where the channel inputs go.

Into shielded enclosures and then into ADC’s.

I have some ideas for tests when at the bench again in the next few days. 800km flight first………

[tautech]

Not seeing a problem here. 

2 captures attached, one from member hendorogs SSA3075X-R with RF loop around the USB-B power input into a 50 Ohm terminated Bodnar pulser.
Analyzer trace setting = Peak Max hold
Preamp ON showed no additional info so was left OFF.

SDS1104X-E capture from webserver via WiFi dongle and LAN to PC.
DSO ~8m from WiFi access point.
Ubiquity airCube router
https://store.ui.com/us/en/products/acb-ac
Scope continues to operate remotely as normal with little/no latency.

[807]

Thanks for trying it out yourself.

Did you try unterminated in channel 1? That was the worst case for me. It blocked the signal 100% of the time. Channel 4 was intermittent, sometimes OK, sometimes blocked. Using the 50 Ohm termination helped. Channel 1 then became intermittent & channel 4 worked for most of the time.

[tautech]

Did you try unterminated in channel 1?

No, the Bodnar pulser is not intended to be used in that manner for RT tests.