just wondering if anyone might have experimented with using LimeSDR for testing frequency response of DIY filters besides using Gqrx panadapter/waterfall for this purpose, which does help but not good enough to allow fine-tuning. In my location FM stations are very strong and they ‘leak’ into HF bands. So, I built FM stop-band filter (80-110 Mhz) and just wanted to figure out if LimeSDR (or perhaps LimeSuite?) can be used to fine-tune it similar to how it’s done using VNAs.
Did you try to use search engine? Things are described there pretty well. Spoiler: it is possible, though you’ll need additional hardware (directional coupler or RF bridge, possibly calibrated noise source as well).
I don’t know, I can see a simple configuration to test basic filters, which would require minimal additional hardware:
First sweep: LimeSDR_TX->20dB attenuator->LimeSDR_RX
reconfigure hardware for second sweep: LimeSDR_TX->20dB attenuator->Filter under test->LimeSDR_RX
A CW signal transmitted by LimeSDR_TX sweeping over the required frequency range in sync with the LimeSDR_RX tuning over the same frequency range.
If you took the difference between the two set of measurements you would be measuring the relative attenuation and cancel out, or at least minimise, any gain or phase nonlinearities in the TX and RX path.
@mzs – yes, that’s how I would approach it. @ccsh – thanks for the rtl-dongle article - this is exactly what I’d like to accomplish with LimeSDR board but ideally without any additional hardware.
Both options involve using additional hardware pieces, which is fine. But before going there, I’d really like to explore if it is possible to do this using just the software and LimeSDR board and no additional h/w pieces.
If you don’t want to use any additional equipment, than I think the only way to go will be making sure that TX gain is lower by 20-30 dB than maximum value, and doing it as @mzs have described, but without external attenuator.
@ccsh - yes, that’s what I had in mind too. In theory it should be possible if LimeSDR non-linearities are negligible for the width of the sweep (70 - 120 Mhz in my case of FM band stop should do). I’ll do a lil more digging on this.
I added the 20dB attenuator for safety, because at initial powerup there could be a TX spike from the board initializing. I thought that you may possibly get away with 10dB because the maximum TX power is about +18 dBm depending on frequency and the maximum RX power level before physical damage is about +12 dBm. But that felt too close so 20-30dB does sound reasonable, me personally I would not connect any output to an input without forcing the signal level to always be below the level required for damage, that way you are reducing the chance of binning or damaging hardware to zero.
Makes good sense. With some unknowns about the board behaviour it’s good practice to err on the side of caution. If I had to just do it, I would not think of trying it w/o ATT. Unless Lime Micro folks who I assume do know this board inside out speak up and confirm it’s possible to do it just like with any other VNA. Which would be mighty cool!
But I guess it all depends of measurement accuracy you need 