EEVblog  

Be sure to watch Part 2: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-0Iv2TnI0C98.html

It's ok, Dave. If you don't want this microphonics plagued scope anymore, I'd take it :3

It's the Seismograph option!!!

Me : "Doctor, it hurts when I do this" Doctor : "Don't do that"

watching a crazy aussie, slapping scopes for 18 minutes. what i am doing with my life?

*Clicks on new EEVblog video to hear Dave's latest ramblings* *Amazed to see how sensitive today's instruments are* *Learned about SMD multilayer ceramic capacitors* Thank you Dave 🙂

I once had a similar problem with a Bit-Error Rate Tester (BERT) I designed back in 2002. The problem was phase shift in my 2.5GHz clock, which was produced by a PLL that multiplied up the 156MHz system clock produced by a Temperature-controlled crystal oscillator (TCXO). There were 16 PLL chips, and each one got it's own 3.3V supply from a low-noise LDO regulator. LDO regulators typically require a few microfarads on their output for stability, and I initially used ceramic caps. When I tapped on the side of the chassis with a screw driver, the 3.3V rail went up (I think) by about 30uV, and it lasted about 1ms. This caused a problem known as frequency pushing, where the output freq of a PLL changes slightly in response to supply rail perturbations. I replaced the ceramic caps with plastic film caps, and the problem went away completely. Why would such a tiny freq change affect the timing so much? Because phase is the integral of frequency. A 1ms supply transient lasts for 156k clock cycles at 156MHz.

I think if someone started taping me with a stick and poking me in my eye, I would trigger too! :-) Poor things.

You are killing me Dave! Pulling out a couple dozen scopes, and I cant even come close to affording a scope.

While I nodded, nearly napping, suddenly there came a tapping, As of some one gently rapping, rapping at my oscilloscope door. "'Tis some visitor," I muttered, "tapping at my oscilloscope door - Only this and nothing more."

I'll take it for free, if you want to get rid of this shocking problem :-)

So is this what they call slapstick?

nice to see what actually occurs when I repair my TV by slapping the side of it

"planned a quick video" we both know that you're incapable of quickies.

As an experiment, do you think you could datalog the interference and extract an audio signal out of this? It would be interesting to see if the microphonic effects of the MLCCs are sensitive enough to actually work as a usable microphone!

It got me curious and I tried the same on my Rigol 1054Z. My results were different. Tapping the chassis did not trigger anything, even harder ones. Tapping the BNC did, but I had to tap it relatively hard. I bought it just a month ago, maybe different hardware version ..

EEVblog #983 - A Shocking Oscilloscope Problem! (Gone wrong!)(Gone sexual!)(You won't believe what happens next...)

Not only MLCC do this, also chips do it, especially if they are in small packages. Opamps "drift" with mechanical stress. On chip oscillators (ring oscillators, etc.) have microphonics. The timebase thus might be significantly microphonic, which could also be evaluated.

hitting the BNC isn't a test of the mlcc because you're transferring charge between the pen and the sense circuitry, which is what it is designed to do. tapping the case though, wow, that's something to bear in mind for the future!

Not sure because I am not front end designer, but may be differential noise cancelling techniques need to be applied in order to cancel such piezo effect. At least for those scopes having touch screen and Dave like users. OR passives need to be reinvented to kill such effects. This is the place where you will find so many interesting things.

Ceramic caps in the signal chain. Pretty typical. That’s where special construction techniques come in. It’s possible to work around this problem, but it requires quite a bit of experimentation to get it right.

Interesting to see that results differ so much and that some budget models/brands (like the siglent) perform way better than some very expensive models/brands.

No such issues on my Siglent SDS 1022DL :) Hit the crap out of my scope at different settings and didn't get such false readings.

Awsome Video as always! I have seen almost every single episode while playing around over the years. Now I'm on the ee train, too. Thanks Dave :)

Just tried this on a HMO3054 and no issue. I find I tend not to whack the oscilloscope anyway :-) Same with the Spectrum analyzer - no pugilistic tendencies towards them, too much money to replace or repair :)

Dave, I will call the manufacturers tech support in the morning, to get more info and feed it back. Agilent / Keysight is my first choice, but I will try the others if not satisfied. The caps used in my Tek 2235 are 3 types. Dipped Tantalums, disc style, and stick style. All are through-hole of course. I have the service manual for the unit with schematics.

I know that with my 20 year old Tek 2024b scope, if I hold the supplied probe tip within a few inches of the back of the scope it's noise a-go-go. But I haven't yet tried using it as a set of bongo drums while capturing any signals.

Been in electronics for years, this is nothing new... But good to show the young ones...

It might be interesting to do these tests with the BNC inputs shorted or with a scope probe to see if there is any difference.

Just a second... is it April 1st yet in Australia?

I got over *2Vp-p* when I used plastic fork on my toy-scope from owon sds5032e

Mr Jones, you must answer truthfully! When did you stop beating your scopes?

Several years ago, my employer was considering a R&S signal generator for noise floor measurements, there were even two R&S sales engineers present when one of our lead techs demonstrated to the R&S types that noise floor measurements could not be done because of the high noise level on the signals from the R&S generator, the he demonstrated the same test using a very old HP 8640B sig genny, noise floor measurement successful.

The TELEDYNE is beautiful

Expecto oscillorum!

So dave, "There's scope for improvement." :D

My Keysight DSOX1202X is immune from tapping on the scope itself BUT I have noticed environmental capacitance effects holding the probe and tapping my feet produces about a 50 volt DC impulse that peaks just as my foot reaches the floor, the velocity of rise increases as proximity to the floor, very much what I would expect in a capacitance situation with a DC voltage gradient in the room.

The DSOX1102G displayed "Trig'd?" when Dave whacked it with his hand.

How do you know that the problem isn't being caused by triboelectric noise from the input connectors?

Actually, the new Owon XDS series scopes have 12 bit resolution. If you go for the 200MHz version, you get 14 bit resolution. I just picked up the XDS3201A through Amazon for $550 US. 12 bits, with the battery, DMM, touch screen, decoding and other options included. So far so good.

I've had a similar effect when flinging a mic cable around at fast paces, as if I used it as a whip. If you turn the gain up on the console to full (most analog boards will have about +75dB of gain on the Mic inputs) you can actually here the effect going through the board.

My DS1054Z does it very similar to yours: CH1 higher sensitivity than the other channels.

It seems like RIGOL fixed it a little bit. Triggering by nocking on the case seems to be a bit harder. The BNCs are still a lot microfonic.

You have a very good set of measuring instruments. It's so cool! I would also like to have so many instuments.

Would be interesting to see if the most sensitive ones could even be affected by music or loud noises.

I dont often use my scope without the leads plugged in. Perhaps the damping when the mass is increased takes the noise floor down. Its the old Mass Spring Damper problem from control theory.

Terminate the BNC inputs and repeat the test?

Isn't this the case that when you leave BNC open, not mechanically bonded, then the BNC connector makes the capacitor out of itself (except of the capacitors issue of course)?

couldn't it be the bnc connectors, which jiggle around if you shock the scope? This will also change the capacity. You should check it with bnc cables connected.

I always suspect from this near April 1st.

I have now been extensively using my RTB2004 for three days using both the touch screen and physical buttons and I have not seen this problem once. I have had two analog chan. probes connected and the digital logic channels for decode operations. I must also say having had my previous Rigol DS2000 series (not A) for over three years I have never notice that problem once. I would like to see the effect with probes connected?

The oscilloscope is just showing the signal you inject into the system whenever you perform percussive maintenance on a device you are trying to get working.

Ceramic capacitors are by no means the only source of these transients. If you take a coaxial cable that obviously has no ceramic caps in it, connect it to the scope and tap the far end with a pencil, you will still trigger the scope. Plastic-metal interfaces also generate microphonic transients when tapped.

try the RTO2024 2GHz BW, 20GSa/S, even better. Cost about $15k with trade-in. Awsome!! Can't wait to see some tear down of that.

Now you have your own set of scope-o-drums, just hook the trigger out signals onto a synthetizer.

could you throw some rubber buffers around the PCB on the inside to try and dampen this? or some rubber around the BNC connectors?

All joking aside, thanks for the heads up.

Positively surprised by the siglent!

One more comment. Mikeselectricstuff said this behaviour is not there if a probe is plugged in. This is correct, at least on my Tek 2235.

Question, why does oscilloscopes costs so much??? Because of their status as specialised equipment?

From the title, I thought it was going to be mains coupling or something like that.

Re-thinking the culprit of the noise phenomena. My Tektronix 2235 has as a build date 1984 (33 years ago). At least for my scope it isn't from MLCC's.

So this can affect my reading if I am making measurements during an earth quake?

"I _think_ that's every scope I have in my lab, I'm not entirely sure" - talk about hoarding ;-)

you could try to find the rsonance frequency of the mainboard with a speaker and a frequency generator

The real question. Can this be turned into a drum machine?

Wow,I have never considered this phenomena before, would make delicate measurements in a high vibration environment rather difficult, maybe not a concern for bench-top models but in a portable that may see use in an industrial setting..... it would be interesting to see if the fluke portables, for example are effected, maybe with harmonics of 50/60hz as found in equipment or spaces one might need to measure.

Digital Scopes have always done this. We always locate our scopes on a shelf above the workbench. On some spectrophotometers I used to work on that had glass photocells we even had to stop talking to make sensitive measurements.

So the capacitors act like capacitive microphone capsules? Awesome!

- Doctor, when I press in this place, it hurts me a lot. What should I do? - Don't press! 🤣

He definitly smashed his spys headphones. :) Dave revealed DSOs as the most vulnurable autospy bugs in the lab!

Interesting video Dave. Do you get the same effect on Analogue scopes?

I just now recalled some snippet of memory related to coax connectors and the insulation used in them. I am off to bed, but if anyone has any insight about this, it would be great to see it. The typical insulation in the BNC connectors is polyethylene for cheap ones or Teflon for higher quality. If the noise is coming from the connectors themselves, that would be a sweet albeit ponderous solution. Again if this is what is happening, the loading in the probe makes sense as per Mikeselectricstuff.

You may wack when you are ready, Gridley.

There's a different way to love oscilloscopes as well! Watch dungeon master Dave getting scopes aroused! From gentle to firm spanking and severe poker action, this is new! Ages 18 and above for this very different scope porn only!

Can you sample audio without a lead then? Turn up a good bass amp!

I think Dave is just trying to get a day or so ahead of April Fools' Day again. He does this every year. :(

wooww... A probeless scope? and touch screen at the same time? woow...

Just imagine trying to measure a VLV signal when something might vibrate the floor and bench.

im sure there is a reason you cant, but is it possible to mount the bnc connectors with braided wire to connect to the pcb? like how a voice coil on a speaker attaches? also maybe attach the pcb with rubber grommets like how a cd rom drive is mounted internally? would that reduce vibrations enough to work?

My Teledyne LeCroy WaveRunner 64Xi has a surprisingly good isolation. I've set the trigger level to 1mV and no matter how or where I whacked it (the case, all the BNCs, the screen) but there is no capture. One thing to note, however, the fan is pretty loud and for this reason I believe the engineers might have paid a special attention to this matter when they designed the front end. UPDATE: I used 1MΩ impedance inputs.

The optional earthquake detection feature is not a bug, it's a feature!

I dare say the oscilloscope maker forgot to glue in the pin mount caps and put rubber sills on the connectors. We get the piezo electric effect in telecommunications equipment with cheap telephones. The fix is to glue all the pin mount caps with hot glue. You can isolate the problem caps by just tapping them with a pen or small plastic screw driver.

My Uni-T UTD2102CEX has the same "problem", but I have to hit pretty hard to get 2-3mV spikes. It seems to be a high-frequency thing. It's easier to trigger with a light pen than with a reversed pair of scissors (steel blades in my hand, plastic handle on scope). Scissors have more mass so produce a lower frequency signal? Anyway, with the amount of force I have to tap the scope with, I don't see an issue with this scope to be honest. Nobody hammers their scope while taking measurements, right?

You missed Post #1000 for this really interesting discovery, Dave !?

I have several scopes they all share this problem with varying amplitudes in the spikes none are as sensitive as as your rs scope

Clickbait factor is high on that one... Next: Read corresponding blogpost: "You won't believe the horrors this oscilloscope caused! Click here to read the whole truth!"

I am somewhat reminded of a certain "the scope does this when I stand up from my chair" video

Lessons that i learned: - oscilloscopes aren't supposed to be used as drums. - oscilloscope probes aren't supposed to be used as drumsticks

Good to know when you're on the field with rattling machinery around!

But what happens if you throw it on the ground?

18:10 that's the finger-tap ID feature. ^__^

This could be a new eevblog featurette: Send Snapshots of Your Microphonics!

Whack it enough times, Dave, and it will trigger. Picking up the frequency of your whacks could bring a premature trigger...careful!

This may be a naive comment, but does leaving the inputs floating make them more susceptible to this?

+EEV blog hi Dave just wanted to ask if you have any old digital osciloscopes for sale thank you and keep up the grate work on your videos :)

Try it with the Tektronix 2225 analog oscilloscope.

Just tried with a lousy Hantek DSO 5102B and I get almost nothing. It does trigger sometimes but only by tapping the BNC connectors and even at 2mV/div , there is barely any ampltude in the burst. Doesn't mean it's a good scope, far from it, but still ...

Hi Dave, I was just wondering.. aren't the BNC channels supposed to be vibration sensitive .. I mean .. if I connected my BNC probe to an accelerometer sensor to measure vibrations I would want highest probe sensitivity .. or perhaps this is a real problem where there is no avenue to redesign probes and BNC front ends .. would like some clarity on that.. or perhaps you can come up with a Gizmo like ucurrent that would counteract this effect if it is indeed deemed undesirable...

Whats a good scope for a back yard tinkerer?