Depends on how you look at it. Using 20MHz means you are defining a standard for the measurement which is required, otherwise everyone gets different results. And 20MHz means every scope can do it.
Tip with the mains cable is to get a big ferrite core ( old monitor CRT ferrite's work well here, or the LOPT core) and wind around 20 turns of the mains cable through it, or as many as you can. Gets rid of a lot of the higher frequency common mode hash. Big, bulky but does the job quite well.
This was super helpful. I'm not a EE and trying to make up for not having one on staff for our project. I didn't know anything about the 10X feature on the probe and we are using a Rigol scope (which I do quite like and find a good value so far).
Common mode noise can be very tricky to isolate, but isolating the scope usually works. But a lot of this noise you describe as "common mode" is mostly differential mode noise picked up by the probe. True common mode noise is usally due to multiple ground paths through the main power ground connections. That's why floating the scope usually works. Also, dont use a 10cm ground lead on the probe. you want the connections to the probe to be as short and direct as possible. That ground lead on the scope acts as an antenna, resulting in differential noise displayed on the scope. 10 years of low level A/D industrial control systems....
John Michael Stock In practice, there are a few ways to do it. 1- cut the ground pin on the power cord to the scope (not recommended, but people do it anyway), or use some adaptor that has no 3rd pin (GND). 2- use an isolation transformer to provide power to the scope, again with no GND connection. This is safer and is the prefered simple method. 3- use a battery-powered scope, if one is available.
Bob Lake Multiple ground paths usually have relatively high inductance (especially if they are wires/cables and not sheets or structure) and they don’t do much for high-frequency common mode noise. If you truly have have common mode noise induced via a connection, then a common mode choke fixes it if it is indeed common mode (it won’t if it’s not). You can also run a connection (say a multi wire cable) through a wideband current transformer and actually measure the common mode noise to validate the diagnosis. The scope has limited common mode rejection and thus acts as a common-mode-to-differential converter. Any unbalanced cabling, including plain coax, also acts as a common-to-differential converter (balun working backwards, you can make values from plain coax!), so the scope would need to have high frequency isolation at the inputs to stop the cable doing it. Running that coax through a few tightly fitting ferrite tubes made of appropriate material would have likely improved things quite a bit.
Bob Lake These things do almost zilch for high frequency noise. An isolation transformer, even with an electrostatic shield is simply a capacitor to whatever ground it’s hooked up to. Thus it doesn’t isolate much once you’re in megahertz, because that ground is just whatever you got, noise and all, with a hf inductance of the wires added. The power supply input filter on the scope also has caps from the secondary to ground input pin, so is transparent for such noise (it shunts local common mode noise from the switching transformer only and conversely it shunts the secondary to primary for such noise: it must do this unwanted task well to do its primary task well). To get a scope or other measurement setup reasonably isolated from common mode noise, the scope and the isolation transformer (with its electrostatic shield!) and the DUT must be all connected via low-inductance connections to a ground plane immediately underneath these devices (that implies that the scope’s connection to the isolation transformer must be two-wire only; the power input module must be directly strapped with braid to the ground plane). The DUT has to run from a battery or a similarly set up PS on the same ground plane. All connections must run directly on the ground plane, and all symmetric connections must be twisted (e.g. power pair from the transformer to the scope and the DUT PS must be twisted). Anyone who understands how valve circuits were designed on a chassis will recognize most of these techniques; they are old news. Substitute ground plane sheet for the chassis and there you go :)
I wish we could have seen the noise with and with and without the ground wire connected to the PS. I noticed near the end of the video (~18 min) that the ground wire was removed, then at the very end, the ground wire is connected again! I know that grounding and ground loops can significantly affect noise. Thanks for the video!
I was thinking of spending a bundle on an isolation transformer for my studio but after seeing this I think I'll replace all switch modes with linear types! Excellent video! :o)
I knew it! A POS Chinese AC Adapter! I've got one that can cause whine in a short wave radio from 10 meters away *with no antenna* connected to the radio. It has a date with the sledgehammer the next time I lay eyes upon it.
The probe is a HIGH Z device matching the input Z of the scope. In the other hand the coax is a low Z transmission line seeing on the PS side a low Z and on the scope side a very high Z, in other words it is acting as unmatched transmission line. Which is not a combination of inductance and capacitance picking up noise like an antenna. The BW of a matched coaxial line runs into UHF, broader than the oscilloscope and its probe. The probe's conductor is generally silver plated.
I have some LED controllers which use PWM dimming up to 50% and then are linear from 50-100%. I was told they were completely linear. That was difficult troubleshooting a flicker problem because in service they were set for less than 50%. On the bench I was testing them above 50%. Of course, the DC supply for the dimmer is switchmode. I am surprised you didn't talk about using a differential probe, or Ch1 minus Ch2 mode- the poor man's diff probe.
Great video. I have had this problem many times during my career. I try to always use a line (mains) filter on my scope, it does seem to help. I have a portable Tektronix THS720 that I used for many years. This problem got very bad with this scope and it turned out to be worn out capacitors in the power supply section of scope.
Dave, PLEASE get a message to our newest guru, (NOT Mr. Magoo!), Paul Carlson of Mr. Carlson's Lab. Tell him we need his help with the common mode noise problem... and ask him to design a lab bench WITHOUT any common mode noise! I dare ya! LOL de KQ2E [and you may wonder why he uses so many old scopes to measure with in his labs!]
You pulled the mains from the power supply, not from the wall. But I have had noise from the mains cable itself. Not sure if it's from the wall or from the cable itself. Changing the cable helped. It was a normal 220V 3 prong cable.
invaluable help for beginner and very funny too ! I got a new beginner scope and couldn't find why is the noise there at peaks even when probe was not connected to anything. Still I don't know where comes but I don't care for a simple beginner thing I'm doing now . :) cheers !
So when you're making noise measurements in your system the best solution is to power it from the same mains SMPS that it will use in the real world not a benchtop PSU. Even better if it's going to be powered by a battery, these kind of problems are a non runner - only internal SMPS issues need to be considered. All these kind of measurements on your system should be done in real world conditions.
My great wish is to get one day scope like you have in video. Now i have TDS2002B which is pretty modest in comparision with todays modern scopes such as Keysight/Agilent.
actually when I built my transmitter (very low range just for learning purpose) I used my oscilloscope as a receiver. Because it picks up all the signal 20 cm from the antenna. It shows how we can easly make a mistake with what we see on the oscilloscope...
If you have a 50 Hz hum in your audio signal and you pass it through a 200 Hz high pass filter the energy of the hum is attenuated. Some distortion does occur due to group delay and pass-band ripple, but this distortion is controllable by good filter design. So, in total, the quality of the audio signal is improved. If this were not true, no one would use high pass filters. You have to know something about the signal you are trying to measure. to measure it correctly.
Question that i would like to ask - Is it actually the coax that's picking up the noise or the electronics of the PSU like you initially suggested? Obviously, the electronics within the probe could dampen the noise detected.
Thanks for you nice video, and i have a question if we couldnt find high performance bnc in our lab, add a common mode choke between source and load will help us eliminate that led switching noise?
Is it just me ... or the noise that is picked up at 14:45 is totally matching his speech just like equalizer ... Is it possible to be captured by the camera that films and be transmitted in any way or the "antena" power supply is picking it up ?
I had a cordless phone power supply that was horrible. When it went into trickle charge it would short the mains across a thermal cutout, and and that would cut in and out. It was so bad it was actually throwing the 60 Hz off down to 54 or 52 Hz about every 1.5 seconds. It was so bad mains freq referenced clocks, stroboscopes, and other gear was not keeping proper time. My old turn table's strobe would walk backwards every 1 1/2 second. The turn table had enough momentum to run faster than the strobe lamp when the main freq dropped. I put some LEDs around some of my ham radio gear, and promptly ripped them off due to pulsed switching noise. It looked cool, but I was transmitting all that noise, which would wreak havoc at some broadcast freqs that had a harmonic resonance with the switching freq. I hate chasing that crap down!
Had something similar with a cheapo switching power supply. one of the caps inside was defective and it was putting noise onto the mains and affecting some X10 modules.
This is an issue I have to deal with at work for video and audio signals on very large shows... I end up having to do the same thing, running round looking for the guilty party.. and some events are in sports stadiums or large festival sites.. It can take a while!! Its exactly why I own a large stock of isolation transformers for both audio and video..
I thought it was going to be from your solar inverters. Is there a way to quieten down cheap inverters? I tried an inline ac filter and ferrites but it didn't really help.
Really cool. I was dealing with this problem earlier, but with a DC motor. Could you go into more information about how to stop the noise? Maybe by adding capacitors too? Thanks!
The problem is that each square pulse in electronic circuits sends a EMP that triggers currents in unshielded circuit traces and other bare metal structures. It's time to design circuits with well shielded traces, is it? Another solution is to decrease the slope of rising pulse edges, in components where it's not absolutely necessary, it reduces significantly the amplitude of the EMP.
He did that with the ground isolator. It's basically just a 1:1 transformer with no earth connection. Beyond that, you're picking up noise through your leads.
I can't seem to eliminate the common noise issue. I've even tried running my scope from a UPS with nothing else plugged in and I still get the the same recurring pattern shown in this video. Help?
Is the power supply case actually forming a complete metal shield from the review it seems like the front panel is made of plastic. If the internal transformer is the pickup if it was properly shielded would the performance be better? Normally those ESD filters should be grounded to the shell of the case to work ideally I think.
How do we determine whether the issue with the switching supplies was common mode or differential mode noise, if in the end that emission was radiated to the coax?
Not all of them, I don't think. Iirc Dave has a video up about a 50 MHz scope. I can remember him saying something like that in a video about the BW limit function though.
Interesting?!!! Oh man That was AWESOME 😍😍 I have guessed that the light may affect on the oscilloscope reading But not like this. Also the human body Man this was just awesome 😂❤️❤️❤️
I don't see how the supply meets its spec even if it's the transformer picking up external EM interference. It should be protected / isolated internal or whatever to deliver the spec under normal operating conditions.
Not only as a ham radio operator I especially have to warn about those more and more common powerline network adapters - Dave, could you show how ridiculous much noise these devices produce?
This is by far the worst thing for high capacity data circuits for telecom.. I have had to isolate this monster on so many circuits with at@t I could tell you in a moments notice as soon as I seen it on our scopes.. sometimes it's a bit$h to find.. The modems have Wi Fi so they can't be shielded like you would like to in order to suppress them from it.. Thanks I said Lights right off the bat... and the cable... yes,,, I didn't see the gear you had under the counter but I was skeptical.. great explanation !!
Hi. What if using an external dc block and configuring the agilent input to 1x, dc coupled and 50ohms terminated when using the coaxial cable? Remember its necessary to use external dc block because when in 50 ohms, the scope input is limited to dc 5Vrms. This is how I proceed with my noise mesurements, specially on onboard low voltage asic power rails. Please comment about it. I'm a fan of your channel. Kind regards
Help! I run my house off a 3000 watt 12 volt to 230-volt inverter from batteries, my inverter is good quality and pure sine wave, however, that pure sine I suspect is made up of fast switching high-frequency peaks. The household wiring seems to be acting as an antenna, the inverter is Earthed to the ground and sits on a large Aluminium plate in a special cabinet, I was thinking of installing a supply filter robbed from an old washing machine on the 230 volts out from the inverter. It drives me nuts, I have bought a beautiful MW/LW antique radio which sounds amazing running on its batteries until the inverter is put on and then it is totally screwed. any good ideas welcome.
Thanks a lot for the always extremely valuable videos. I have a question. If the psu is picking up noise from all around then the noise is indeed coming from the psu. I mean the psu is acting like an antenna. Am I wrong ? maybe a better shielding of the internals of the psu could help ? in your experience are there any psu with excellent isolation from this common mode noise ? Last question .... would a spectrum analyzer be a better tool to understand the nature of this noise ? thanks a lot again. Kind regards, gino
There are better shielded PSUs - if you make them yourself. They are special-purpose. Most PSU designs are made to be cheap, and there’s really no standard to specify them well, eg. common mode impedance between output and mains is not normally characterized, even through it’s an extremely important parameter in low-level work, especially in electrophysiology, electrochemistry and other high-impedance work.
@@absurdengineering Hi thank you very much for your kind and valuable advice I understand the complexity of the topic I opened some smps for computer and some had a kind of shield like Dell for instance The cheapest ones have nothing at all Kind regards, gino
Could it be a switching power supply running off the same power line that you've got your equipment set up to? I have a look at your lighting to if I was you.lol
The Qili power supply has a CE marking. Note the spacing between the C and the E. In this case the CE means "China Export". There is no way the power supply would pass European CE compliance testing. More detail here: support.ce-check.eu/hc/en-us/articles/360008642600-How-To-Distinguish-A-Real-CE-Mark-From-A-Fake-Chinese-Export-Mark
Depends on how you look at it. Using 20MHz means you are defining a standard for the measurement which is required, otherwise everyone gets different results. And 20MHz means every scope can do it.
Tip with the mains cable is to get a big ferrite core ( old monitor CRT ferrite's work well here, or the LOPT core) and wind around 20 turns of the mains cable through it, or as many as you can. Gets rid of a lot of the higher frequency common mode hash. Big, bulky but does the job quite well.
No, but it was the entry standard for a long time. That's why higher bandwidth scopes have 20MHz limit switches, the defacto standard
Hi just wanted to compliment you on your amazing EEVblog. One of the best of the best.
This was super helpful. I'm not a EE and trying to make up for not having one on staff for our project. I didn't know anything about the 10X feature on the probe and we are using a Rigol scope (which I do quite like and find a good value so far).
Common mode noise can be very tricky to isolate, but isolating the scope usually works.
But a lot of this noise you describe as "common mode" is mostly differential mode noise picked up by the probe.
True common mode noise is usally due to multiple ground paths through the main power ground connections. That's why floating the scope usually works. Also, dont use a 10cm ground lead on the probe. you want the connections to the probe to be as short and direct as possible. That ground lead on the scope acts as an antenna, resulting in differential noise displayed on the scope.
10 years of low level A/D industrial control systems....
+Bob Lake What does floating the scope mean. Is that some advanced isolation / filter system for the input power supply?
No ground reference. Ground "floats".
John Michael Stock
In practice, there are a few ways to do it. 1- cut the ground pin on the power cord to the scope (not recommended, but people do it anyway), or use some adaptor that has no 3rd pin (GND). 2- use an isolation transformer to provide power to the scope, again with no GND connection. This is safer and is the prefered simple method. 3- use a battery-powered scope, if one is available.
Bob Lake Multiple ground paths usually have relatively high inductance (especially if they are wires/cables and not sheets or structure) and they don’t do much for high-frequency common mode noise. If you truly have have common mode noise induced via a connection, then a common mode choke fixes it if it is indeed common mode (it won’t if it’s not). You can also run a connection (say a multi wire cable) through a wideband current transformer and actually measure the common mode noise to validate the diagnosis. The scope has limited common mode rejection and thus acts as a common-mode-to-differential converter. Any unbalanced cabling, including plain coax, also acts as a common-to-differential converter (balun working backwards, you can make values from plain coax!), so the scope would need to have high frequency isolation at the inputs to stop the cable doing it. Running that coax through a few tightly fitting ferrite tubes made of appropriate material would have likely improved things quite a bit.
Bob Lake These things do almost zilch for high frequency noise. An isolation transformer, even with an electrostatic shield is simply a capacitor to whatever ground it’s hooked up to. Thus it doesn’t isolate much once you’re in megahertz, because that ground is just whatever you got, noise and all, with a hf inductance of the wires added.
The power supply input filter on the scope also has caps from the secondary to ground input pin, so is transparent for such noise (it shunts local common mode noise from the switching transformer only and conversely it shunts the secondary to primary for such noise: it must do this unwanted task well to do its primary task well).
To get a scope or other measurement setup reasonably isolated from common mode noise, the scope and the isolation transformer (with its electrostatic shield!) and the DUT must be all connected via low-inductance connections to a ground plane immediately underneath these devices (that implies that the scope’s connection to the isolation transformer must be two-wire only; the power input module must be directly strapped with braid to the ground plane).
The DUT has to run from a battery or a similarly set up PS on the same ground plane. All connections must run directly on the ground plane, and all symmetric connections must be twisted (e.g. power pair from the transformer to the scope and the DUT PS must be twisted).
Anyone who understands how valve circuits were designed on a chassis will recognize most of these techniques; they are old news. Substitute ground plane sheet for the chassis and there you go :)
This is an amazing video, Dave. It makes me value my Tektronix 2245A scope, even more!
wish I found this video earlier. such a great help. I was having this trouble for years . thanks Dave
I wish we could have seen the noise with and with and without the ground wire connected to the PS. I noticed near the end of the video (~18 min) that the ground wire was removed, then at the very end, the ground wire is connected again! I know that grounding and ground loops can significantly affect noise. Thanks for the video!
I was thinking of spending a bundle on an isolation transformer for my studio but after seeing this I think I'll replace all switch modes with linear types! Excellent video! :o)
Prince Wessterburg , Watch out using an isolation transformer the wrong way! It can fry things, big time!
I knew it! A POS Chinese AC Adapter!
I've got one that can cause whine in a short wave radio from 10 meters away *with no antenna* connected to the radio. It has a date with the sledgehammer the next time I lay eyes upon it.
The probe is a HIGH Z device matching the input Z of the scope. In the other hand the coax is a low Z transmission line seeing on the PS side a low Z and on the scope side a very high Z, in other words it is acting as unmatched transmission line. Which is not a combination of inductance and capacitance picking up noise like an antenna. The BW of a matched coaxial line runs into UHF, broader than the oscilloscope and its probe. The probe's conductor is generally silver plated.
I have some LED controllers which use PWM dimming up to 50% and then are linear from 50-100%. I was told they were completely linear. That was difficult troubleshooting a flicker problem because in service they were set for less than 50%. On the bench I was testing them above 50%. Of course, the DC supply for the dimmer is switchmode.
I am surprised you didn't talk about using a differential probe, or Ch1 minus Ch2 mode- the poor man's diff probe.
Great video. I have had this problem many times during my career. I try to always use a line (mains) filter on my scope, it does seem to help. I have a portable Tektronix THS720 that I used for many years. This problem got very bad with this scope and it turned out to be worn out capacitors in the power supply section of scope.
Dave, PLEASE get a message to our newest guru, (NOT Mr. Magoo!), Paul Carlson of Mr. Carlson's Lab. Tell him we need his help with the common mode noise problem... and ask him to design a lab bench WITHOUT any common mode noise! I dare ya! LOL de KQ2E [and you may wonder why he uses so many old scopes to measure with in his labs!]
Great video. Another lesson learnt by a youngster. And another trap avoided.
Brilliant lesson on understanding switching noise pickup. Thanks a lot Dave!
You pulled the mains from the power supply, not from the wall. But I have had noise from the mains cable itself. Not sure if it's from the wall or from the cable itself. Changing the cable helped. It was a normal 220V 3 prong cable.
Very helpful Dave Jones. Thanks for the video.
I am loving the videos! A definite improvement. Thanks for tolerating the criticism dave!
invaluable help for beginner and very funny too ! I got a new beginner scope and couldn't find why is the noise there at peaks even when probe was not connected to anything.
Still I don't know where comes but I don't care for a simple beginner thing I'm doing now . :) cheers !
So when you're making noise measurements in your system the best solution is to power it from the same mains SMPS that it will use in the real world not a benchtop PSU. Even better if it's going to be powered by a battery, these kind of problems are a non runner - only internal SMPS issues need to be considered. All these kind of measurements on your system should be done in real world conditions.
Great investigation Dave! i learn so much from your channel that I wish i'd had the chance to learn at school. Keep it up mate.
My great wish is to get one day scope like you have in video. Now i have TDS2002B which is pretty modest in comparision with todays modern scopes such as Keysight/Agilent.
That failed LCD TV you saw in the Amiga 500 teardown, it was still plugged in and turned on under the bench
Thanks! I just tracked down common mode from a switching power supply using the video.
Great video. Your "grey beard" comments are always spot on.
actually when I built my transmitter (very low range just for learning purpose) I used my oscilloscope as a receiver. Because it picks up all the signal 20 cm from the antenna. It shows how we can easly make a mistake with what we see on the oscilloscope...
If you have a 50 Hz hum in your audio signal and you pass it through a 200 Hz high pass filter the energy of the hum is attenuated. Some distortion does occur due to group delay and pass-band ripple, but this distortion is controllable by good filter design. So, in total, the quality of the audio signal is improved. If this were not true, no one would use high pass filters. You have to know something about the signal you are trying to measure. to measure it correctly.
8:59 Why too much noise came when light turned off ?
Question that i would like to ask - Is it actually the coax that's picking up the noise or the electronics of the PSU like you initially suggested? Obviously, the electronics within the probe could dampen the noise detected.
If the noise is common on both wires why the scope shows the noise at all?
Thanks for you nice video, and i have a question if we couldnt find high performance bnc in our lab, add a common mode choke between source and load will help us eliminate that led switching noise?
What was the Qili Power supply powering? Quite a hefty beast at 4.2A.
Is it just me ... or the noise that is picked up at 14:45 is totally matching his speech just like equalizer ... Is it possible to be captured by the camera that films and be transmitted in any way or the "antena" power supply is picking it up ?
I had a cordless phone power supply that was horrible. When it went into trickle charge it would short the mains across a thermal cutout, and and that would cut in and out. It was so bad it was actually throwing the 60 Hz off down to 54 or 52 Hz about every 1.5 seconds. It was so bad mains freq referenced clocks, stroboscopes, and other gear was not keeping proper time. My old turn table's strobe would walk backwards every 1 1/2 second. The turn table had enough momentum to run faster than the strobe lamp when the main freq dropped. I put some LEDs around some of my ham radio gear, and promptly ripped them off due to pulsed switching noise. It looked cool, but I was transmitting all that noise, which would wreak havoc at some broadcast freqs that had a harmonic resonance with the switching freq. I hate chasing that crap down!
I dig these general practice type videos.
Had something similar with a cheapo switching power supply. one of the caps inside was defective and it was putting noise onto the mains and affecting some X10 modules.
Is the cable that's plugged into the PSU's earth-reference connected to anything in the later takes?
This is an issue I have to deal with at work for video and audio signals on very large shows... I end up having to do the same thing, running round looking for the guilty party.. and some events are in sports stadiums or large festival sites.. It can take a while!! Its exactly why I own a large stock of isolation transformers for both audio and video..
I thought it was going to be from your solar inverters. Is there a way to quieten down cheap inverters? I tried an inline ac filter and ferrites but it didn't really help.
Haha, I was sure that the culprit was the lights or something that had a switch mode supply. Thanks for sharing your knowledge :)
good .. example for common mode noise and probes quality
How about common mode suppression on the coax (some big ferrite rod or toroid)?
Or winding the coax to a coil. works a treat on lower frequencies.
You please help. The microcantroller crystal portion causes more noise . Please tell me the noise filter circuit
Interesting video. That is one fancy scope!
Add a clip on Common Mode Choke on the coax, and say 3 turns through the core. It will help. 43 material or similar will be good.
Really cool. I was dealing with this problem earlier, but with a DC motor. Could you go into more information about how to stop the noise? Maybe by adding capacitors too? Thanks!
The problem is that each square pulse in electronic circuits sends a EMP that triggers currents in unshielded circuit traces and other bare metal structures. It's time to design circuits with well shielded traces, is it? Another solution is to decrease the slope of rising pulse edges, in components where it's not absolutely necessary, it reduces significantly the amplitude of the EMP.
He did that with the ground isolator. It's basically just a 1:1 transformer with no earth connection. Beyond that, you're picking up noise through your leads.
Since the power supply was the cause. Does this mean that every signal measured and connected to the scope would pick up that 66khz waveform .
Jeez, Dave! You've been on TH-cam longer than TH-cam has been around it seems.
Thank you again for another fantastic video!
I can't seem to eliminate the common noise issue. I've even tried running my scope from a UPS with nothing else plugged in and I still get the the same recurring pattern shown in this video. Help?
Is the power supply case actually forming a complete metal shield from the review it seems like the front panel is made of plastic. If the internal transformer is the pickup if it was properly shielded would the performance be better? Normally those ESD filters should be grounded to the shell of the case to work ideally I think.
what could one add to the DC current outlet of the power supply to filter such noise?
How do we determine whether the issue with the switching supplies was common mode or differential mode noise, if in the end that emission was radiated to the coax?
Not all of them, I don't think. Iirc Dave has a video up about a 50 MHz scope. I can remember him saying something like that in a video about the BW limit function though.
What about using AC coupling to look at noise to get down to
Do you have a video on op-amp stability?
Interesting?!!!
Oh man
That was AWESOME 😍😍
I have guessed that the light may affect on the oscilloscope reading
But not like this.
Also the human body
Man this was just awesome 😂❤️❤️❤️
can i just use a no ground earth wire adapter?
Please help me in measure this noise using simulation tools like Ansys SIwave & perform EMI Filter analysis for a PCB Design
Great video Dave
So I can hook up my USB oscilloscope to 120v?
I don't see how the supply meets its spec even if it's the transformer picking up external EM interference. It should be protected / isolated internal or whatever to deliver the spec under normal operating conditions.
Not only as a ham radio operator I especially have to warn about those more and more common powerline network adapters - Dave, could you show how ridiculous much noise these devices produce?
18:00 times 1 probe is not like regular coax it has 1 MΩ input impedance and 100pF... this is like LPF
What type of filter did you used in the ac main line of the scope at the back?
Also I was wondering if having shielded mains cords would be of use.
This is by far the worst thing for high capacity data circuits for telecom.. I have had to isolate this monster on so many circuits with at@t I could tell you in a moments notice as soon as I seen it on our scopes.. sometimes it's a bit$h to find.. The modems have Wi Fi so they can't be shielded like you would like to in order to suppress them from it.. Thanks I said Lights right off the bat... and the cable... yes,,, I didn't see the gear you had under the counter but I was skeptical.. great explanation !!
Excellent stuff. Much appreciated
So, is there something you can put between your device and outlet to get rid of the noise? A transformer perhaps?
You can use differential subtraction mode with 2 probes, from Math menu.
So basically any portable scope that operates from battery will not have this problem?
I don’t really understand coax cables, why doesn’t it have to be 50 ohms terminated it this case?
Hi. What if using an external dc block and configuring the agilent input to 1x, dc coupled and 50ohms terminated when using the coaxial cable? Remember its necessary to use external dc block because when in 50 ohms, the scope input is limited to dc 5Vrms. This is how I proceed with my noise mesurements, specially on onboard low voltage asic power rails. Please comment about it. I'm a fan of your channel. Kind regards
Are there analog scopes that don't use CRT displays? Or is it possible to make a CRT not put out significant amount of noise?
Help! I run my house off a 3000 watt 12 volt to 230-volt inverter from batteries, my inverter is good quality and pure sine wave, however, that pure sine I suspect is made up of fast switching high-frequency peaks.
The household wiring seems to be acting as an antenna, the inverter is Earthed to the ground and sits on a large Aluminium plate in a special cabinet, I was thinking of installing a supply filter robbed from an old washing machine on the 230 volts out from the inverter.
It drives me nuts, I have bought a beautiful MW/LW antique radio which sounds amazing running on its batteries until the inverter is put on and then it is totally screwed. any good ideas welcome.
Thanks Dave! i learned quite a bit, as always.
Thanks a lot for the always extremely valuable videos. I have a question. If the psu is picking up noise from all around then the noise is indeed coming from the psu. I mean the psu is acting like an antenna. Am I wrong ? maybe a better shielding of the internals of the psu could help ? in your experience are there any psu with excellent isolation from this common mode noise ? Last question .... would a spectrum analyzer be a better tool to understand the nature of this noise ? thanks a lot again. Kind regards, gino
There are better shielded PSUs - if you make them yourself. They are special-purpose. Most PSU designs are made to be cheap, and there’s really no standard to specify them well, eg. common mode impedance between output and mains is not normally characterized, even through it’s an extremely important parameter in low-level work, especially in electrophysiology, electrochemistry and other high-impedance work.
@@absurdengineering Hi thank you very much for your kind and valuable advice I understand the complexity of the topic I opened some smps for computer and some had a kind of shield like Dell for instance The cheapest ones have nothing at all Kind regards, gino
Maybe some ferrite clamp around the power cable could help?
Fun video. Need to look at getting a new scope.
Just had a play, put my probe near my phone and ahh haa at 20mv you can watch an Australian talk 😂
Isn't there a scope with a feature that can just zero out the noise?
Hi ..
can u guide me how to design a MCUcircuit for noisy environment.Any Tips ..??
At the end of the video the link is private
Yes, the next video is still private, I'm being evil :-p
Main power in Australia (and most of the world) is 50hz, while 60hz is the standard in N. America.
What would happen if you put the power supply and scope in a Faraday cage?
What do you recommend for a power supplie for a student in elektronics currently running with an old PC PSU but it has a lot of ripple :(
"Hello mister Qili Power." Haha. Love it!
Try changing the output filtering caps (inside the PSU).
Why do they actually ALWAYS use relays to protect the output on/off ? instead of FETs ?
Is that CE logo of that Qili adapter a fake one?
"Greybeards" I resemble that remark. BTW you didn't discuss differential input measurement, or won't your scope do that.
what exactly is the "50Hz garbage" Dave talks about around 2:53?
Could it be a switching power supply running off the same power line that you've got your equipment set up to? I have a look at your lighting to if I was you.lol
Dave's creed : "Big trap for young players " :-)
Makes me feel young every time I make a mistake ^_____^
lmfaooo
The Qili power supply has a CE marking. Note the spacing between the C and the E. In this case the CE means "China Export". There is no way the power supply would pass European CE compliance testing. More detail here: support.ce-check.eu/hc/en-us/articles/360008642600-How-To-Distinguish-A-Real-CE-Mark-From-A-Fake-Chinese-Export-Mark