EEVblog #983 - PART 2: Piezoelectric Oscilloscope Issue Followup

A quick followup to the piezoelectric oscilloscope and vibration issue shown previously. This time talking about probe loading and probe microphonics.
Kemet Clip On Lead Frame (CoLF) Capacitors:

Support the EEVblog through Patreon!

EEVblog Amazon Store (Dave gets a cut):

[whidbeyman]

Here's an obscure but possibly related phenomenon: I designed a product that has a rigid board PCB assembly inside an ultrasonically welded plastic housing. During U/S welding, the microcontroller-supervised electronics are in a sleep state. In a small percentage of units, the microcontroller and/or some of its I2C peripherals experience static register corruption, so that the device fails to come out of its sleep state properly. Much effort has gone into troubleshooting these failures. Removed from the housing, the assembly passes all tests and x-ray inspection of solder joints and components reveals no damage. We are now looking at the possibility that the bypass MLCCs (amounting to about 30 uF, mostly 2.2uF) are generating enough microphonic voltage on Vcc to corrupt memory registers. Anyone out there have any experience with cap microphonics during U/S welding? The size and density of the assembly makes it unthinkable to go to any dielectric with less volumetric capacitance than X7R. And, yes, U/S welding was probably not the best choice for sealing the enclosure, but I didn't make that decision, and it's a bit late to do a total enclosure redesign.

[emcgon]

I winced in pain every time you stabbed the screen with that pointer 😱

[jpmiller]

When one connects a scope and 64 channels of a logic analyzer to a multimillion dollar server under development and wait days for some rare failure such a thing can be damned relevant. I chased many false triggers to electrostatic events. Just movement nearby can sometimes trigger a scope or logic analyzer or present false data. I found a microphonic capacitor causing a real time clock to miscount time due to fan vibration. But it never occurred to me to suspect mechanical shocks to the instrument itself. Some of my diagnostic failures may have been due to this effect.

[TonyPirc]

I don't think I've ever seen someone hit a scope so much without the requisite swearing.

[fakenews]

I showed your first video to Keysight and I think they are having a fit, they denied their scopes did this . Now they are making their own video . My old analog scope goes crazy when I tried this .

[glasslinger]

We had to use the older Tektronix 465 and other early scopes when we were doing work onboard ships at sea. The bench continuously vibrated at certain engine speeds! A lot of the new scopes are microphonic!

[andrewkowalczyk]

Wondering if you could use a capacitor and an ADC as a crude bump sensor for a small cheap project...

[dougspurell]

Final feedback to Dave and all from Keysight. Through an email and phone conversation, they have accepted my test method and the noise measured, is from the connectors. My thanks to Dave for allowing a different view. As well many thanks to Keysight who I am responsible for their efforts with another manufacturers instrument.

[morte]

Does the 32G scope's 50om front end have this kind of problem?

[RebounD]

Second link in the description is an edit link. :)

Youtube changed the links in the video manager:
- click on video for video link
- click on name for edit link

[monchiabbad]

Does this mean we could setup a global seismic network using all active oscilloscopes ?

[srmeister]

Dave, would it be possible to "use" these mlcc's as actual microphones? can u crank up the resolution high enough to see, say the sound from a speaker placed right next to the scope, on the screen?

[dougspurell]

Good job Dave on the micro-phonic issues of MLCC capacitors. I will investigate the same symptoms with my Tektronix 2235 and report back if anyone is interested.

[MichaelBRothschild]

I have an RTB2004 which I have now been using for almost a week and have not noticed this effect on the Analog inputs at anytime whilst using the touch screen. I have also been using a wired mouse from the front usb port and the user interface is great with the use of a mouse, especially the cursors for measurements and also moving the waveforms around the screen. Dave's scope appears to be more susceptible to this problem than my one for whatever reasons (I have been using my scope down at 2mV a division without bandwidth limit). The scope has a few software bugs which I am sure will be sorted with further firmware updates (new version already on R&S website). So far I am delighted with the scope, the digital decodes are great and far better than on my original Rigol DS2072.

[bwack]

I keep thinking, "yes you can MLC with less microphonic effect", but we are down in the 1mV/div range. High sensitivity, low noise amplification. What would be the alternative to MLC .. ? As for the high impedance probe, would an active probe be better?

[AttitudeGames]

Dave when you was looking for the internal 50 Ohms load, you didn't scroll down that menu, if you look you have a scroll bar on the right of that menu.

[bluedeath]

How does your latest dumpster dive find do with this sort of thing?

[Elektronik-]

@ 6:40 The same amplitude @ 5 mV and @ 50 mV on then open channels!? Why? Is the problem not located in the frontend? This is an extraordinary scope - 10 bit, and a great display for a very good price (expecially the 2002 version) - why you don't do a "serious" review?

[lebanbo]

Hi Dave, would you like to give us a talk about how signal that carries data from the capacitor of a DRAM cell to the latch of a register works on the circuitry perspective please ? I know how sense amplifier works, I have trouble in visualized seeing the process of how the signal which carries data is generated and sent to the register, and how register receives it and traps it into the latch of register . Thanks in advance.

[mariusvasileingeaua]

Just bought one RTB2004, 3 days ago because of you. You should get some kind of reward from R&S :) Thanks Dave!