Forum Replies Created
- Posts
In that case, you can look at the phase of the I/Q samples (using atan2). Convert phase to frequency for the instantaneous frequency. You can reduce noise by low pass filtering the frequency results.
If you are using the API, we provide streaming I/Q data at 40 MSPS, which you could have centered at 3 GHz.
For amplitude, use I*I+Q*Q, and for phase use the atan2 function.
By relative phase, I mean 2 things:
1) If your generator time base does not exactly match the BB60C, the phase angle in the I/Q data will rotate at the frequency difference.
2) If you lock time bases together, there will be a fixed phase offset between the generator and the BB60C. In most cases, there will be a *new* fixed phase offset each time you start I/Q streaming, but exact multiples of 20 MHz may avoid this.
You can download the API manual from our website for more information.
Using the API, you can measure amplitude and phase 40 million times per second. By adjusting sample rate and signal bandwidth, and adding some averaging, you can make more accurate, but slower, amplitude and phase measurements.The BB60C would be the only current option for this. It has 27 MHz of instantaneous bandwidth.
In zero span mode, video triggered, you could trigger on the rising edge of the pulse, and observe e.g. 1 microsecond before rising edge and 19 microseconds after, and make measurements on the instantaneous amplitude (including pulse width at 25 nanosecond resolution), frequency, and relative phase of the pulse.Tishers,
Yes, the BB60C can be connected to the TG44A using the TG Sync signal to enable scalar network analysis, which provides about 600 frequency/amplitude signals per second.
When doing scalar network analysis, there is no advantage to having the 10 MHz time base connected (the TG reference is uncorrected 1 ppm, the BB60C is corrected, typically < 0.1 ppm initially). But if you need frequency lock between the BB60C and the TG44A, you can connect the TG44A 10 MHz out to the BB60C 10 MHz in. It may be 1 ppm off, but the frequency will be exactly locked (typically when testing phase shift using streaming I/Q data when 0 ppm *relative* frequency error is required).Sure, I understand. You may be interested in the SM200A when we come out with it (going in to production next month). The required filters for this are built in, and the dynamic range is much better than the BB60C, so you should see minimal increase in the noise floor when you connect an antenna. Or, if they are completely satisfied with the DRT price and performance, there’s really no compelling reason to switch.
Petr,
Internally, this pulse could be generated, but it would only be used internal to the VSG25A to modulate a carrier frequency above 100 MHz. We do not have user accessible baseband signals.It looks like the BB60C went from -133 to -112.5, and the DRT went from -128 to -121. This probably reflects how much out-of-band energy is hitting the amplifier, and that amplifier’s linearity. My guess is the DRT uses either a more linear amp, or a smaller passband.
One thing you could try, if you need a lower noise floor in this band, is to add a band pass filter. This should remove some out-of-band energy before it hits the amplifier and bring the noise floor down.In saConfigSweepCoupling, there is a parameter that gets passed in, bool reject. If you set this to true, and decrease your VBW to maybe 1/10 of your RBW, you should get the expected noise reading and shape.
The reason for this is the USB-SA44B does not have hardware image rejection, and instead relies on a software algorithm. When this is not enabled, an image response shows up 21.4 MHz away from the actual input signal.
For broadband signal applications, the BB60C would be an ideal choice, since it has excellent hardware image rejection. But the SA44B can be configured to work well for signals up to 20 MHz wide, using the image reject flag appropriately.
Generally, if the signal is continuous, the image reject flag should be true. If the signal is pulsed or frequency hopping, you must set the flag to false to catch the signal, and the image response is unavoidable on the SA44B.
Hopefully this helps.Your VBW was less than RBW, so you were low pass filtering the power response. Did you try setting VBW to auto, or = RBW?
I don’t think we ever made an Allan Deviation tool. The closest we have is the frequency difference meter, where you can watch the frequency error drift over time.
And when you change the detector to min/max, ensuring spur reject off, you see no change at all in Spike?
Mikeh,
Have you tried attenuating the signal to maybe -10 dBm (well into linear region of SA-series), ensuring detector is min/max, spur reject is off, gaussian filter. What kind of reading do you get?The last picture had detector set to average. I’m just trying to figure out what the difference may be.
Generally, customers have reported that this works, but there has been some additional troubleshooting from time to time.
The frequency difference meter would be good for CW signals. I don’t know how it’ll perform with 4FSK. You might try our modulation analyzer in zero span for that. I think I’ve seen around 1 Hz typical error on readings there with time bases locked.
For a CW, using the frequency difference meter, a typical accuracy would be 0.01 ppb plus time base error, so in your case, +/- 1.01 ppb. At 440 MHz, +/- 0.4444 Hz.
Adding modulation complicates things a lot.The actual frequency error for sweeping is (reference error + 1 sample), so if you have a sweep with a “bin size” of 10 Hz (around 40 Hz RBW), and a 1 ppb timebase, your marker should read within 10.44 Hz of the actual frequency.
You can calculate bin size as (span / number of points).
If you are using the frequency difference meter or the modulation analysis tool center frequency on a CW, the frequency error should be very small, as long as your signal to noise level is good. With the frequency difference meter, I typically see something like 0.01 ppb error.Dragan,
1. I believe AUTO RBW/VBW is a function of Spike. If you like the RBW that Spike selects for your measurement, simply use this RBW, and set VBW = RBW.
2. In the same saConfigSweepCoupling function where you set RBW, set Reject = true for spurious reject
3. saConfigAcquisition is where you select average or min/maxThe BB60C has a lot of advantages: sweep speed, signal purity, instantaneous bandwidth, etc. But the SA124B is perfectly capable of making good measurements on many types of signals above 6 GHz. Crowded spectrum, signals wider than 40 MHz, or intermittent signals are difficult or impossible to measure with the SA124B.
You could downconvert to any convenient frequency, but there are a lot of off-the-shelf converters for satellite systems that target 1-2 GHz IF, and a lot of mixers that don’t allow an IF much higher than this.
There is a DC blocking capacitor, so you should be fine up to 16 volts DC.
You can use a signal generator and mixer to downconvert to 1-2 GHz and feed into the BB60C. The mixer will drop amplitude 6-8 dB typically