Engineer It - How to test power supplies - Measuring Noise
Вставка
- Опубліковано 3 лип 2024
- TI's Bob Hanrahan demonstrates how to measure noise when testing power supplies.
For more videos on testing power supplies, check out:
Overview: www.ti.com/testingpoweroverview
Measuring efficiency: www.ti.com/measuringefficiency
Measuring stability: www.ti.com/measuringstability - Наука та технологія
Would have been nice to see the actual difference between using a long ground wire and your method to see how much the difference actually makes!
Nice oscilloscope!
Hi Bryan-
Thanks for the comment. The oscilloscope was a new LeCroy HDO6000. The scope utilizes qty 4, 12-bit resolution 2.5 Gigasample per second ADC’s from Texas Instruments. The video was recorded in a remote lab which were fortunate enough to be able use for a few days. Bob found the 12-bit resolution made a new level of measurement possible. He could clearly see things that are distorted with the 8-bit scope, or simply not visible.
that about probe it self.
1x10 ?? M ohm
or coax 50 termineted
Thanks bob in this video you taught me some important tricks, I will treasure them!!
This is great. Please post more useful videos like this.
Thank you so much for such valuable video!
Thank you. Simple basics for a guy who likes to learn
Glad you like it! We invite you to watch some of our other "how-to" videos, including Robert Kollman's Power Tips. They are all available on TI's power management video channel.
Is there any app can i use to measure the affected transistors in my mother board?
Thank you for the great video.
This is a great video. Has Bob Hanrahan made any more videos?
That's a hell of a scope.
Ron: A Spectrum Analyzer identifies RF problems an oscilloscope helps pinpoint the cause/solution through time domain details. Fundamental/harmonics or mixing of ripple/transient noise can result in interference. A linear reg or LC filter after a Switch-Mode PS (SMPS) reduces this noise but ripple noise/harmonics can move from RF by moving the SMPS oscillator frequency. Transient noise can reduce/move with an RC snubber. Specifics vary with design. For more: ow.ly
LhUy
I agree, the oscilloscope can measure level, but where in the frequency spectrum are those levels. That is my concern. Noise from a supply can be a problem for many circuits like simple audio circuits, but for for RF applications the noise that is often from a switching power supply oscillator can play havic on RF circuits. I did enjoy the video, just thought would give more info about RF noise.
How would you measure much lower noise levels, like for a linear supply, where the noise is way below 1mVpp? I mean if the measured signal is below the noise of the oscilloscope, what sort of probe do you recommend?
Thanks for the video! :-)
Flexing on us with that scope I see.
@Mario Kyree why am I smelling a whiff of scammy spam? Oh that's because InstaPwn is a scam! Time to report it.
You is great man naice video am wery happy
Thanks sir.
Hello, can this same procedure be effective with breadboard power supplies? Let me know. I look forward to hearing from you 😎
Very interesting. What type of scope is that?
Thank you...
That is one badass oscilloscope
+Hydranix HNx 500Mhz @2.5Gsps badass? We have a 100GHz @ 240Gsps in our lab. That's badass.
+Yo Da 10-100Zi?
Hydranix HNx Yeah! They have it in a half-height rack with other instruments which together costed more than several lambos! Pretty insane
and the probes cost more than a couple Toyotas right? As afraid as I would be to touch that thing, I want to play with it X)
That is true, although all RF grade cables and interconnect at those frequencies is severely priced! There are several engineers working on just making optimum use of the oscilloscope for HF measurements at the moment. At mm-wave frequencies instrumentation is a big PITA (or, "an art", if you prefer), and discerning between the noise floor and intermodulation distortion of the circuit under test and the internal circuitry inside the oscilloscope is really tough. I also learned just today that there is some considerable trickery going on for the highest frequency ranges. It's not a normal scope in several other ways than just the price haha.
Well I'm outgunned with that 11k scope lol. Guess I'll just live with my ripple haha. Either way helpful to know a little bit of what to look for on a smps.
I have seen the tip and barrel method for the first time, it seems that this method requires some special equipment to be present on the PCB to make the measurement possible.
I am wondering how to check my power supply, on how much amps I get from it. There is not any label. I measure the voltage and it's 13.8. I connected my transceiver and on receive mode everything is fine, when I pressed the PTT it looks like everything going to die. Please help.
Regards
more please
Hello sir, i want to measure a Signal and i Don t know where from is comming the noise! My probe is pasive one. I use a 3,57Mhz. The Signal is looking like your.
Ok the tech switched from longer ground to barrel ground method. But you're measuring at completely different points. Capacitor filters out most of the noise. So longer ground noise measurement should have been made same spot as pin & barrel test point.
I am wondering the same thing, but why not just use the springy ground probe to also measure the high frequency noise? That noise is still within the 100MHz range.
Nice video. One question, maybe a noob one. You've measured noise without a load applied (load is OFF, Power supply shows 30mA). Does load applied affect the results or is always a measurement done without load? Regards
yes, I am wondering the same thing.
Fantastic video. A touch-screen oscilloscope though? Jeez!
But what about those spikes every cycle?
Are they to be totally disregarded?
What are they, even?
Hi Bodragon, Those very high frequency transient spikes are caused by a combination of EMI pick-up right at the probe, and IR voltage spikes during switching off (dead) times. The EMI pick-up is caused by the probe and/or return picking up a magnetic field caused by the fast switching edge rates, this contribution is mostly not real since it's the probe itself that's receiving the energy. On the other hand you will also see a spike during the dead times when both pass transistors within the regulator are off, caused by current flow continuing into a high Z. If your load is sensitive to these spikes (depending on your application), they can be minimized by reducing dead time, if possible. But in any case much can be removed with proper decoupling capacitance placed very close to the load. Series inductance can also be used to reduce the higher frequency noise but comes with ESR that will cause an IR drop across the inductor that must be considered. One of the many tradeoffs within a design. Thanks for the comment.
For a sec I legit thought he said his name was Bob hammer-hands
pepe's world pro tip: use a spectrum analyzer for seeing noise. its much more intuitive to see the noise especially at various frequencies. also place things in a shielded faraday cage & close the lid. i do it on pepe's world: ua-cam.com/video/bgyUGBi-ZCc/v-deo.html
imagine if tesla or edison could see this
You look like George Clooney!😊
Thanks, LOL!
I came here to see the oecilloscope. Teledyne making good oscilloscopes, but nobody (including teledyne lecroy) reviewing the product. Awful marketing
The test was done by a tech, obviously not an engineer. The good part is it shows the typical noise of a power supply output, but what about noise at say 14.1 MHz. This is most concern for radio ops...will the power supply interfere with my rcvr??? A spectrum anal is needed...scopes at this levels are useless for RF. Again a tech not an engineer doing the test.
Give the guy a break, Im sure he's surrounded by more engineers you can poke a stick at.. let alone all his years of experience and training with a survivor and leader in the industry.. TI
What makes you better than him if you seemingly don't know the answer to your question yourself?
Hes an engineer.