| Products > Test Equipment |
| Bode plot up to a few 100 MHz |
| << < (9/11) > >> |
| joeqsmith:
I've been playing around a bit with rectangular waveguides in the 8-12GHz range. An open with waveguide is a bit meaningless when it comes to calibration. Instead, we mechanically offset the short to shift the phase. Here I walk through the calibration for another member and compare the results with my old Agilent PNA: https://www.eevblog.com/forum/rf-microwave/experimenting-with-waveguides-using-the-litevna/msg4801919/#msg4801919 |
| mawyatt:
--- Quote from: alm on October 06, 2023, 12:51:05 pm --- I'm skeptical about the ability to get useful measurements at hundreds of MHz using a high impedance environment, since there is no such thing as high impedance at higher frequencies. Similar to using 10x scope probes for hundreds of MHz. They work, assuming you are have a very well behaved low impedance source. The impedance fluctuations over the bandwidth would be huge, likely affecting the measurement. --- End quote --- Here's some interesting links to making relatively high impedance measurements at higher frequencies, but not necessarily Bode Plot type measurements (especially "Closed Loop"), but more of a relative type measurement. Dr Cripps simple yet highly effective "SMUI" an unterminated semirigid cable with SMA has proven quite useful in our past, allowing effective but not too intrusive voltage measurements. We also used some of these voltage probes with micro-coax and wafer probe fixtures for finer measurements with our custom broadband chips and hybrids. Of course these are for relative measurements in response to the near E-field, but our experience was they did not perturb the DUT too much and proved quite useful. Another interesting use of such was the near field microwave microscopy :-+ Anyway, thought some might find this useful!! Best, https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwiv6OrmxuGBAxXSM0QIHWfTBMgQFnoECCIQAQ&url=https%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F6668%2F4753970%2F04753992.pdf&usg=AOvVaw2RuytMvnomfmQi2IvpcRSw&opi=89978449 https://www.academia.edu/28719448/Cripps_Steve_C_RF_power_amplifiers_for_wireless_communications_ourdev_672470C259C |
| Martinn:
--- Quote from: tszaboo on October 06, 2023, 07:48:07 am ---I see an 1 GHz white noise source for about 30 USD (don't know how good quality is) so OP really needs to consider whether paying a lot more for phase information is worth it. Even if the white noise source is not flat to 1 GHz it's possible to measure it and compensate the results with the source magnitude difference. --- End quote --- Like this one https://www.aliexpress.com/item/4001032161351.html What about dynamic range/linearity requirement vs. swept sine? Regarding "Bode" phase or just amplitude: I guess for verifying a 100 MHz scope frontend amplitude only would be fine. Not sure how much extra information there would be compared to looking at a 3 dB roll-off only. Maybe if there's some peaking inbetween might be more obvious in the phase plot. |
| Martinn:
--- Quote from: joeqsmith on October 06, 2023, 12:26:19 pm ---Without all four S parameters, how do you measure the amplifiers stability? It's another shortcoming of the low cost VNAs out of China, 2-port 1-path. My work around was to add a transfer relay. --- End quote --- I'm not a RF guy so bear with me. Why would I need anything else than in-out (S21) for verifying if a DC-100 MHz low noise amplifier would be stable? You could either look at the transient response (excessive ringing) or the bode amplitude (peaking) to find out its stability margin. --- Quote from: alm on October 06, 2023, 12:51:05 pm ---The VNWA 3SE (latest version) also incorporates a transfer switch so it can measure all four S parameters without changing connections. --- End quote --- Again, why would I need this (besides S21)? --- Quote from: alm on October 06, 2023, 12:51:05 pm ---I'm skeptical about the ability to get useful measurements at hundreds of MHz using a high impedance environment, since there is no such thing as high impedance at higher frequencies. Similar to using 10x scope probes for hundreds of MHz. They work, assuming you are have a very well behaved low impedance source. The impedance fluctuations over the bandwidth would be huge, likely affecting the measurement. --- End quote --- Yes and no. Yes above some MHz you have to be very careful with layout and wiring. No - high Z scope inputs with 500 MHz BW exist, so do probes - so why not? If I have any high Z amplifier, just solder a 50 ohms terminator at its input, place a SMA connector across it and feed it with whatever source. This should be OK for 100 MHz easily (again, I am not talking about GHz measurements, I guess above 300 MHz things get complicated). As a matter of fact, one thing I'd like to try is a classical DC servo JFET analog scope frontend (which most likely even my RTB2004 has). For measuring, I'd terminate the input with 50 ohms; output would be somewhat lower impedance anyway or I'd insert a buffer (normally, next stage would be PGA/VGA driving a differential ADC input, which I don't want to mess with. |
| alm:
--- Quote from: Martinn on October 06, 2023, 03:51:40 pm ---No - high Z scope inputs with 500 MHz BW exist, so do probes - so why not? --- End quote --- What is the impedance of such scope input or probe at 500 MHz based on the capacitive reactance? Would you really call that high impedance? |
| Navigation |
| Message Index |
| Next page |
| Previous page |