Perfect timing! I was just looking into this. The derating is quite significant for high voltages, so I’m using a combination of MLCC and film capacitors. Thanks for sharing your knowledge! 😁
Excellent discussion! I've been lucky enough to have projects that allowed 'oversized' and higher voltage parts. Nice to know more precisely what those advantages are.
Excellent video. I'm skimming through it but I'm sure it will be another one of your videos which I reference countless times. I certainly use more capacitors than any other component in my designs. I work on a lot of DSP/amplifiers/etc so decoupling/coupling is quite important for my projects
@@PhilsLab Thank you very much! I will probably spend several additional hours soon looking into that exact topic. I'm about to do a final design revision on a class AB amplifier (~25-50W). Most of the entire final design revision has to due with improving the electrolytic capacitors on the driver and linear power supply, so I have an immediate excuse and interest to delve into this topic. I really appreciate you bringing this topic to my attention, its not something I had thought about since school
Murata is characterizing their CCaps using 10 mVAC and 500 mVAC bias (at least in Simsurfing models). How in your setup -6dbm of power frequency sweep translates to constant voltage bias? My understanding that DUT is not connected via fixed impedance transmission line, apart from calibrated out parasitics that power is dominated by DUT ESR. So, setting constant power output AC bias would produce different VAC bias at each frequency.
dBm power here is with reference to a 50 Ohm resistive load at the output (and the 50R output series resistance of the Bode 100). Aside from effects of calibration of course, the unit does produce a constant voltage at its output, which then is attenuated/altered after the 50R series output resistance due to the load. You can check out the 'Bode Analyzer User Manual' (PDF) for more details (e.g. section 8.4).
@@PhilsLab Looking at the pdf now. "You can choose Vrms which is the root mean square of the output voltage at 50 Ω load. Again the real output voltage depends on the load you connect to the output. The internal source voltage is 2 times higher than the displayed value. " That's what I expect, instrument is calibrated for 50R loads, has 50R source impedance, so you probably set output to Vrms/2 of your expected bias for DUT connected directly and not via transmission line.
![Seungmin "그렇게, 천천히, 우리(As we are)" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/kAzmhLHePqU/mqdefault.jpg)
A top quality video as usual
Thank you!
Very eager to watch the pi filter video you promised ❤ thanks a lot, Phil.
Everybody is talking about capacitor DC bias derating, but nobody talks about ferrite bead current derating...
Topic for a separate video (to come) :)
Wow, I never thought the derating was so much -50% at 3/4 of the rated voltage. Video is very timely for a current project. Thanks Phil.
Thanks, Craig! Yeah, and there a lot of MLCCs where the derating is a lot worse!
Perfect timing! I was just looking into this. The derating is quite significant for high voltages, so I’m using a combination of MLCC and film capacitors. Thanks for sharing your knowledge! 😁
Awesome, glad to hear the timing was right! :)
Nice Video Phil Thank you so much! 😊
Thanks for watching!
Excellent discussion! I've been lucky enough to have projects that allowed 'oversized' and higher voltage parts. Nice to know more precisely what those advantages are.
Thank you!
Excellent video. I'm skimming through it but I'm sure it will be another one of your videos which I reference countless times. I certainly use more capacitors than any other component in my designs. I work on a lot of DSP/amplifiers/etc so decoupling/coupling is quite important for my projects
Thank you! Keeping in mind component derating is definitely very important for all projects.
@@PhilsLab Thank you very much! I will probably spend several additional hours soon looking into that exact topic. I'm about to do a final design revision on a class AB amplifier (~25-50W). Most of the entire final design revision has to due with improving the electrolytic capacitors on the driver and linear power supply, so I have an immediate excuse and interest to delve into this topic. I really appreciate you bringing this topic to my attention, its not something I had thought about since school
What are your opinions on feedthrough capacitors?
Murata is characterizing their CCaps using 10 mVAC and 500 mVAC bias (at least in Simsurfing models).
How in your setup -6dbm of power frequency sweep translates to constant voltage bias?
My understanding that DUT is not connected via fixed impedance transmission line, apart from calibrated out parasitics that power is dominated by DUT ESR.
So, setting constant power output AC bias would produce different VAC bias at each frequency.
dBm power here is with reference to a 50 Ohm resistive load at the output (and the 50R output series resistance of the Bode 100). Aside from effects of calibration of course, the unit does produce a constant voltage at its output, which then is attenuated/altered after the 50R series output resistance due to the load. You can check out the 'Bode Analyzer User Manual' (PDF) for more details (e.g. section 8.4).
@@PhilsLab Looking at the pdf now.
"You can choose Vrms which is the root mean square of the output voltage at 50 Ω load. Again the real
output voltage depends on the load you connect to the output. The internal source voltage is 2 times
higher than the displayed value. "
That's what I expect, instrument is calibrated for 50R loads, has 50R source impedance, so you probably set output to Vrms/2 of your expected bias for DUT connected directly and not via transmission line.
Nice voice
Thanks!