Forum Replies Created
- Posts
Essentially I’d just make a QPSK waveform, apply N rotations, probably in excel, maybe stagger them for a more realistic envelope, add them together and scale.
Our latest VSG60 software allows you to export waveforms. If you email me the details I might be able to throw a waveform together, as long as you’re ok with me rounding to easy offsets. justin at signalhound dot com
I’d need to know symbol rate, root raised cosine filtering unless otherwise specified, channel count and spacing
What exactly are you looking to do? Basically anything custom is going to require some DSP tools, like Matlab or our VSG60 software.
For the VSG25 you’ll need to choose your I/Q sample rate and pattern length so the sequence can repeat perfectly, e.g. 4 MSPS, upsample by 32 for a 128 MHz sample clock. Use a repeating pattern of e.g. 32 symbols for 1024 sample pattern length. Then you can shift the center frequency in units of 128/1024 MHz, repeat for each QPSK signal, add and scale.
You can generate a sequence twice as long, then use the middle portion, to remove any edge artifacts.Note that the VSG60 does not have the same limitations, and is generally the preferred tool for applications like this.
The challenge with the VSG25A is keeping your signal to <=2048 repeating I/Q samples. There's no reason you couldn't have 2 to 8 QPSK channels if they're ideally spaced and playing a short repeating sequence (Maybe 16-64 symbols would be a maximum). The VSG60 or VSG200 could handle much more complex (and arbitrarily longer) sequences, but 40 MHz of bandwidth is all you get.
It looks like a good processor. And I see Thunderbolt ports and a USB 3.2 type A. That should be everything you need to run the VSG200 and SM435B simultaneously.
That being said, we have seen occasional makes / models that should be able to run multiple devices simultaneously but still struggle, and it can be difficult to predict. But this laptop looks like a solid choice. February 4, 2026 at 3:56 pm in reply to: SM API: smGetDeviceList() & smNetworkConfigGetDeviceList() bug Thanks for pointing this out. I’ll make sure the software department sees this.
Yes, the wider bandwidth next generation VSG will include a trigger in/out.
Jay,
You can automate marker readouts using our SCPI library, but you cannot copy the text directly from the screen.Pedro,
We are actively gathering requirements for our next gen VSG line, with considerably more bandwidth. If you had to pick a bandwidth, what would you want, knowing that higher bandwidth generally increases cost? And how big of a waveform (in milliseconds) would you need? Single channel or multi-channel?CJobbers,
Hopefully our support team was able to help you. If not, reach out to support (at) signalhound (dot) com for more direct support.Damon,
You can download Spike for Linux (Ubuntu and Red Hat). The SA- products are Windows only, but BB- SM- and SP- products are supported under both.
https://signalhound.com/spike/Something like the BB60D would work well for this application. If you are still in your 30 day return window, reach out to support@signalhound.com
Nom,
Unfortunately the VSG60A requires the SIMD registers of the x86-64 architecture. We are adding limited AARCH64 support for some of our spectrum analyzers, but the VSG60A will probably not be ported.llevitt,
I’ve never tried this, but here are my initial thoughts. It seems to me that if you mapped your noise source’s ENR from RF to IF frequencies, then mapped the result back to RF frequencies, it might just work. It probably requires that image responses are sufficiently filtered, and that gain + ENR is in a reasonable range (15-35 dB?) to reduce the effect of the analyzer’s noise floor and compression.Garth,
Yes, if you wanted to do this programatically, you could use the sa_api and tg_api. Do 2 sweeps, one with your reference through, and one with your DUT. For each point within the sweep, tune the TG then do a small (100 kHz span or smaller) sweep around the signal and find the max value. Subtract the 2 sweeps for your insertion loss. We created a support ticket for this issue, but I don’t know where it is in the queue, so a “DIY” TG sweep might be a good solution in the short run.Devjack,
We are looking into this. It does look like we clamp to 1 kHz in this mode, which I can see does not make sense for your measurement, especially since the TG itself has 10 Hz resolution.
Two potential workarounds while we investigate:
1) You can use linear interpolation to get a good estimate of the 3 dB points
2) You can exit scalar analysis mode, and just do a sweep with max hold. Then step the TG in finer steps using the TG control.
I apologize for the inconvenience.I believe the base license (without cross-correlation) is $99 / year. If you don’t need the new features, you can keep using your 3.xx version for free.
https://signalhound.com/products/advanced-phase-noise-measurement-tool-kit/Amedina,
If you have an external 10 MHz source, I would use it to drive both instruments’ 10 MHz input. If not, you can use the 10 MHz output from the SM200 to drive the VNA400. This will discipline the VNA clocks and frequency output. I would do this before you calibrate the VNA.CH,
The software-based image rejection does limit the usefulness of the SA44B. For scans wider than 200 kHz, any time there is energy 21.4 MHz above, and 21.4 MHz below, the frequency being scanned, a spur will show up there. Narrower scans (200 kHz or below) have the same issue, but at different offset frequencies.
All of this is fixed in a product like the BB60D, with excellent hardware-based image rejection.MMark,
Below about 160 MHz, the SM-series is in its direct conversion band. But in order to squeeze this into our normal I/Q signal processing, we do an upfront digital tune for some combinations of center frequency and bandwidth. This will ultimately move the pass band, and will digitally filter out some frequencies. If you email our support team, we can go into more specifics for your use case, but this might be what you’re observing.