Great . I enjoyed it ... I think it was better to add the 0.2ohm to the inductance but will not change much .. this is expected as the resistance always damps the resonance .... Some times they call them parallel /series resonance ...
I have seen this being used - rather than use an unpredictable aluminum electrolytic (that has temperature dependent ESR) use a more predictable low ESR cap, but ad an extra predictable low value resistor so that the system is stable at any temperature.
Another way of get a proper noise reduction or decoupling is to use a little ferrite or resistor in series. I guess it's also a little relative. In the last case you showed there was an anti resonance, but overall it was better than the electrolitic capacitor except for that one resonance point. So in the end you still improved the overall noise.
Well, you would add a ferrite to prevent noise going over the supply line, that won't prevent the noise at the source - its still a good idea to short circuit the noise source with components that have an effect on high frequency noise - like low value capacitors. Anyway, there are certain circuits (usually only high performance IC's like processors) that have a dedicate section in the datasheet describing the impedance their supply pin should see - for example below 0.1R for frequencies below 50MHz. This is needed to keep the supply stable enough for proper operation.
Now what about if i have a DC bias on these capacitors? With MLCC capacitors, there is a big decrease in capacity. That and other factors that are not ideal and not exactly known, make it kinda useless to simulate in my opinion.
You can either change the simulated capacitor value - based on the expected derating from DC biasing (ex use 6uF of capacitance instead of 10uF - the nominal value); or you can use DC-bias capacitor simulation models - some manufacturers do offer these, but the simulation will be quite slow since the capacitor modeling is highly complex.
Can this also work for example in lm3886 v+ and v- audio circuit, The decoupling was chosen to be 4.7uF , 22uF and 1000uF. www.circuitbasics.com/wp-content/uploads/2016/10/How-to-Design-a-Hi-Fi-Audio-Amplifier-With-an-LM3886-Circuit-Schematic-with-Diode-2.png
This is a great video showing an effect of bypass caps that not many other resources give credit to. Thanks from an electrical engineering student!
I've been searching for an explanation like this one for a long time. Thank you!
Glad it was helpful!
Great . I enjoyed it ... I think it was better to add the 0.2ohm to the inductance but will not change much .. this is expected as the resistance always damps the resonance .... Some times they call them parallel /series resonance ...
I have seen this being used - rather than use an unpredictable aluminum electrolytic (that has temperature dependent ESR) use a more predictable low ESR cap, but ad an extra predictable low value resistor so that the system is stable at any temperature.
Excellent topic. I never really gave thought to this effect, but understanding it could really save the design. Thanks for presenting that.
best teacher
Good movie, thank you!!
Nice explanation
U are great 👌👌👌
you are amazing
Please do a video on reduction of noise on PCBs in analog circuits.. shielding grounding etc...
what country you are?
Great stuff!, could you do one about ferrite beads?
Thanks for the suggestion! I will try to get to that topic in the near future!
@@FesZElectronics Thanks!
what is the AC 1on the first ground?
Its part of a voltage or current source; it sets the output of the signal source to be "1" in an AC type of simulation
Another way of get a proper noise reduction or decoupling is to use a little ferrite or resistor in series. I guess it's also a little relative. In the last case you showed there was an anti resonance, but overall it was better than the electrolitic capacitor except for that one resonance point.
So in the end you still improved the overall noise.
Well, you would add a ferrite to prevent noise going over the supply line, that won't prevent the noise at the source - its still a good idea to short circuit the noise source with components that have an effect on high frequency noise - like low value capacitors. Anyway, there are certain circuits (usually only high performance IC's like processors) that have a dedicate section in the datasheet describing the impedance their supply pin should see - for example below 0.1R for frequencies below 50MHz. This is needed to keep the supply stable enough for proper operation.
Now what about if i have a DC bias on these capacitors? With MLCC capacitors, there is a big decrease in capacity. That and other factors that are not ideal and not exactly known, make it kinda useless to simulate in my opinion.
You can either change the simulated capacitor value - based on the expected derating from DC biasing (ex use 6uF of capacitance instead of 10uF - the nominal value); or you can use DC-bias capacitor simulation models - some manufacturers do offer these, but the simulation will be quite slow since the capacitor modeling is highly complex.
Can this also work for example in lm3886 v+ and v- audio circuit,
The decoupling was chosen to be 4.7uF , 22uF and 1000uF.
www.circuitbasics.com/wp-content/uploads/2016/10/How-to-Design-a-Hi-Fi-Audio-Amplifier-With-an-LM3886-Circuit-Schematic-with-Diode-2.png
Te magyar vagy igaz?