Forum Replies Created
- Posts
You can turn image/spur rejection off. This will force it to stay in the higher noise state, but also lets spurious signals through. Note that with the SA44B there is a 3 dB correction for broadband noise, as the software image rejection cannot filter out noise at the image frequency.
The BB60C can perform this measurement without the correction.
I would probably start with a BB60C for EMC it already has >50 dB spurious rejection, which is plenty for precompliance testing of unintentional radiators. For intentional radiators, I might just buy a simple high pass filter for looking at harmonics, and a low pass filter for looking at frequencies below the transmit frequency.
Preselection is a pain, and expensive, but when you need it there really is no good substitute. External high pass and low pass filters e.g. from Mini-Circuits, will work for many applications, but they do require manual setup.
Agung,
If the problem is that the USB-SA44B has some LO feedthru which is causing interference, there are several ways to mitigate this.
1) Use a reference level of -30 dBm. This will turn on the RF preamplifier, which also blocks LO leakage. Your leakage should be well below -50 dBm in this state.
2) Add an external 20 dB pad to the input of the SA44B and increase the sensitivity by lowering the reference level. With a 20 dB pad, and a reference level of -40 dBm, your LO leakage should be around -70 dBm.
3) If this is not enough, switching to the BB60C is recommended. Its LO feedthrough is below -80 dBm in all modes, and the LO frequency is >2.4 GHz higher than the signal, so it is very unlikely to interfere.Peter,
Yes, for crystals and narrowband filters, we needed a separate mode with much longer settling time. We chose a specific span, 100 kHz, as the switchover point, assuming that people who want to look at the pass band of a device with less than 10 kHz bandwidth would do it with a narrow span. For wideband rejection, where maybe a parasitic capacitance dominates, the faster sweep mode should work fine.Our web admin fixed the .vi issue. Hopefully this works now.
I sent you the .vi file. I’ll resend it. Hopefully it gets through…
Email support at signalhound dot com and I can send it to you…
Last try…
Not permitted for security reasons?!? Rename it to .vi
I thought I attached it… here it is.
I made a LabView program that opened, set frequency, amplitude, CW mode, waited 3 seconds, then closed the device. The only time I ran into a problem was when I did not close the device before trying to run it. If the device does not close, then when you try to open it, it will give you error -4.
Here it is. Maybe it will help?
sgModulationType is an enum, generally equivalent to an “int”. If you cannot include the header file with the enum (e.g. LabView), you can just use an int.
HerbertN,
Thank you for reporting this. I will take a look at this in a few days and see if I can duplicate the problem. Labview places several restrictions on what a DLL can and can’t do. There may be an incompatibility we need to address.+13 dBm reference is fine for the SA44B (it will tolerate up to about +20 dBm). The circuitry prefers square for lower phase noise, but +13 dBm has a pretty fast slew rate, so it probably doesn’t matter much.
If frequency doesn’t have to be exact (+/- 0.1 ppb typical), you might be able to find a used LPRO Rubidium standard for $150 or so.
Have you tried closing the device and re-opening it? If you are using the Spike API, this will likely work. With the older API, I believe there is a SHAPI_CyclePort or similar that should accomplish the same thing.
Ivan,
The BB60C spur reject option will generally remove input-related spurious (signals generated from image frequencies, harmonic mixing products, etc.). What it cannot remove is residual signals from internal clock multiples. It also cannot remove intermodulation products from the input itself (e.g. strong 101 MHz and 103 MHz produce low-level 99 and 105 MHz IM3).
If you see the signal with no input, it is probably a residual clock multiple (typically 10 MHz intervals, sometimes stronger 80 MHz multiples), which unfortunately cannot be removed. These should be below -106 dBm at maximum sensitivity, and are typically around -120 dBm except for a couple of spots.
If you suspect intermodulation products, increase the reference level. This will decrease artifacts from intermodulation. November 16, 2015 at 9:23 am in reply to: BB60c + TG44 – SNA – very strange behavior in 9k-10MHz range Thank you for bringing this to our attention. It looks like there is a software bug we will need to address. I’ll let you know what we find.
You could do this manually, by setting the TG frequency and amplitude directly, rather than SNA mode, or even write a small application to automate this.
I would love to add a suite of tools to the SNA for doublers, triplers, mixers, etc. but we do not currently have these features available in Spike.
The VSG’s odd-order harmonics at 100 MHz distort the time domain waveform, making it look like a high-passed square wave, but if you use an external 150-200 MHz low pass filter to remove the harmonics, you can easily clean it up and get a waveform that looks good in the time domain. In the manual we talk about applications that are sensitive to harmonics will require a low pass filter. This is one of those applications.
The TG124A also produces significant harmonics above 16 MHz, but the SA124B removes the harmonics, giving you a good measurement.
A good filter for applications from 100 MHz to 200 MHz requiring low harmonics is http://www.minicircuits.com/pdfs/BLP-200+.pdf
Any application where the output of the VSG goes to a receiver or analyzer will generally not be sensitive to harmonics. In fact, for clocking applications, the harmonics may actually be desirable in many cases to improve slew rate.
I’m glad you find it useful. I wish the 5 MHz RBW had better performance on the SA44B (DANL / spurious / amplitude accuracy), but I suppose for a quick, low res sweep it gets the job done.
The 200 MHz cutoff is necessary because of extremely poor spurious performance at lower frequencies.