| Products > Test Equipment |
| Equipment to measure RC filter attenuation |
| (1/5) > >> |
| slugrustle:
I often use 1-stage and 2-stage RC filters for various purposes, such as * Filter AC noise from voltage divider intended to measure DC * Turn PWM signal into adjustable analog DC signal * Filter power supply to sensitive circuits with low current draw (opamps, references, etc.) I like to measure the filters to make sure they are operating close to expectations. I typically measure rising and falling input voltage step response (sig gen / knife switch input, scope on input and output) and attenuation. Attenuation measurement is giving me trouble. Usually, I use a signal generator with an injection transformer and a lab supply so I can see if DC bias affects the filter response (I sometimes use X7R capacitors in filters—sue me). I typically use 10x probes on filter input and output to avoid loading the output too much, AC couple the probes, enable 20MHz bandwidth limit, and turn on my scope's averaging statistics. The trouble is that I often run into the noise floor of the scope at the frequency of interest, even with a large input signal to the filter. The other trouble is a non-negligible DC offset on the 5mV/div range when AC coupled and the lack of an AC RMS measurement on my scope (Siglent SDS2504X+) so I use pk-pk measurement, but this is artificially inflated by noise. At this moment, I'm trying two Rigol DM3068 DMMs measuring AC voltage on input and output. Frequency limitation is 300kHz, and I'm having trouble with the low voltage AC measurement taking a long time to settle on the trend chart. What's a good setup to measure filter attenuation? I don't really care about phase. Some ideas come to mind: * Frequency response analyzer (Omicron Labs Bode 100, Venable 6305, etc.) * Picoscope 4262 in a sig gen + scope setup * Spectrum analyzer for input / output amplitude measurement instead of scope, still use sig gen + transformer + DC supply as stimulus. Can this handle DC offset? I'm guessing it is always or can be AC coupled. * Vector network analyzer, although they tend to have high minimum frequencies (9kHz, etc.). Can this handle DC offset? Any tips on which option above is best? Are there other or better ways to do this? |
| tautech:
--- Quote from: slugrustle on September 13, 2022, 05:32:44 pm ---What's a good setup to measure filter attenuation? --- End quote --- To 50 MHz, your scope Siglent SDS2504X+ using the inbuilt AWG and the Bode plot function. You can push higher with a standalone AWG but the limit is 120 MHz. Some guidance here in Siglents App note where they use a SDS1104X-E that hasn't 50 Ohm inputs therefore external feedthrough terminations are used. https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/ There might also be some help and guidance in the following threads: https://www.eevblog.com/forum/testgear/bode-plots-useful/ Here rf-loop explores the limits of the Bode plot feature in 4ch X-E's: https://www.eevblog.com/forum/testgear/siglent-sds1x04x-e-bodeplot-ii-(sfra)-features-and-testing-(coming)/ Simple 2 stage RC pass filter linked in one of the threads above. Tables, markers and results = OFF to reduce clutter. |
| tszaboo:
If you have a programmable signal generator/AWG and one or two multimeters, that will give you a versatile setup to measure this. Just automate the setup with a script, and you are fine all the way to your DMMs AC range, and drink a coffee while it works. Other setup I could use in the lab is a Keysight 3104T, which has the generator, and Bode plot function up to 20MHz. This is easier to setup, one instrument, and it totally goes to the frequencies where the cable already matters. |
| slugrustle:
To state it more succinctly, the specific problem I'm running into is the noise floor / minimum measurable voltage on both my Siglent SDS2504X+ scope and my Rigol DM3068 DMMs, even when using a 20Vpk-pk sine wave as the stimulus. I hit this amplitude measurement limit often before getting to the frequency of interest, where the filter is often attenuating by 80dB or so. The frequency of interest is often below 1MHz. Is there a good method that would allow measurement of filter attenuation down to say 100dB while allowing DC offset in the stimulus? Perhaps something in the list below or other options... * Frequency response analyzer (Omicron Labs Bode 100, Venable 6305, etc.) * Picoscope 4262 in a sig gen + scope setup * Spectrum analyzer for input / output amplitude measurement instead of scope, still use sig gen + transformer + DC supply as stimulus. Can this measure in dBVrms units? Can this handle DC offset? I'm guessing it is always or can be AC coupled. * Vector network analyzer, although they tend to have high minimum frequencies (9kHz, etc.). Can this handle DC offset? |
| tautech:
--- Quote from: slugrustle on September 14, 2022, 07:43:27 pm ---To state it more succinctly, the specific problem I'm running into is the noise floor / minimum measurable voltage on both my Siglent SDS2504X+ scope and my Rigol DM3068 DMMs, even when using a 20Vpk-pk sine wave as the stimulus. --- End quote --- And that stimulus source is from ? |
| Navigation |
| Message Index |
| Next page |