Products > Test Equipment
Siglent SNA5000A two and four port VNA's
G0HZU:
I suppose another reason could be that something went wrong during the full 4 port calibration. Doing a full 4 port cal using just a two port Ecal would be quite a tedious process so maybe something happened during this procedure to corrupt the 0-1850 MHz part of the span. Maybe a port cable was swapped too quickly before the analyser had finished a sweep for example. However, the sporadic upward blips at 29:11 and 29:12 look suspicious to me. There's also a permanent downward dip in the trace at 1850MHz when the blips disappear. This doesn't look right either.
tautech:
--- Quote from: G0HZU on August 31, 2022, 09:02:22 pm ---I suppose another reason could be that something went wrong during the full 4 port calibration. Doing a full 4 port cal using just a two port Ecal would be quite a tedious process so maybe something happened during this procedure to corrupt the 0-1850 MHz part of the span. Maybe a port cable was swapped too quickly before the analyser had finished a sweep for example. However, the sporadic upward blips at 29:11 and 29:12 look suspicious to me. There's also a permanent downward dip in the trace here either side of the upward blips. This doesn't look right either.
--- End quote ---
Later when some time is available I’ll do a 2 port Cal with ~1.85 GHz set to CF to examine this more closely.
We have the 4.5 GHz Siglent SMA Cal kit which puts that near the middle.
Poke me if I get distracted please. :)
G0HZU:
Thanks. Can you do it at -30dBm source power and 5kHz BW?
It's a while since I've been inside one of the old Agilent E5071B VNAs but the 5071B uses a similar octave tuning VCO that also covers (I think) 2 to 4GHz.
The 5071B VNAs can be prone to synth locking issues at the range change points. From memory this happens to the E5071B at approximately 2.1GHz and 4.2GHz and it also affects 8.3GHz. When the VNA becomes poorly it produces erratic blips at these points of the sweep because the synth can't reacquire lock quickly enough at a range change point. Sometimes it just loses lock and I think this is what happens at 8.3GHz. This initially causes noise on the synth and as the fault gets gradually worse it cause significant tracking errors. It tends to get worse as the VNA warms up. Eventually the VNA detects a lock problem and starts beeping out error messages. On a typical E5071B it takes a few weeks to go from a minor tracking problem to losing lock for significant parts of the sweep. I've seen this happen several times as I've used a lot of E5071B VNAs over the years. I think it may be a common fault as the VNA gets older. It requires a part replacement in the synthesiser module to fix it.
I'm not sure what is happening at and below 1850MHz on the Siglent VNA. It could be caused by a tracking error or maybe it is just a calibration procedure issue. The sporadic blips at 1850 MHz at 29:08 through 29:13 in the video look very similar to what I've seen on a poorly E5071B when it changes range during the sweep.
tautech:
--- Quote from: G0HZU on August 31, 2022, 09:39:16 pm ---Thanks. Can you do it at -30dBm source power and 5kHz BW?
I'm not sure what is happening at and below 1850MHz on the Siglent VNA. It could be caused by a tracking error or maybe it is just a calibration procedure issue. The sporadic blips at 1850 MHz at 29:08 through 29:13 in the video look very similar to what I've seen on a poorly E5071B when it changes range during the sweep.
--- End quote ---
Please excuse my pitiful efforts Port 1>2 via Siglent Utility kit N-SMA adapters, 2x 6" cheap SMA cables and Cal kit Thru for a 1dB/div S21 measurement at -30dB and 5kHz BW.
I really need spend more time with this. :palm:
G0HZU:
Your plot doesn't look right so something has gone wrong with the calibration. Ideally, you should see a flat s21 trace with just a bit of noise on it if cal'd at -30dBm. Also, can you show s11? See below for a through measurement after a two port cal at -30dBm using a homebrew N cal kit up to 6GHz and an E5071B VNA. The s21 trace is at 1dB/div and the phase is at 5deg/div. The s11 shows >50dB return loss.
Also see a 1 port measurement of a 54.5 ohm chip resistor after a -30dBm calibration. You can see the return loss through HF and VHF is as expected at about 27dB with minimal noise with a 5kHz BW. There's no averaging or smoothing used in these plots.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version