Thanks for the request Andrew.

Hi Andrew,

I’ll look into this, see if there is a way to get a fix into the next release. It looks like a bug maybe introduced when we updated the spectrogram plot? I’ll let you know what I find.

Thanks for the heads up.

Regards,
Andrew

What does the CPU usage look like when real-time is running? Is there an increase/jump in cpu usage when you move your mouse across the screen?

Are you running your monitors off the integrated GPU or do you have a dedicated GPU?

If you have time and the ability, I would be interested if it helps to run just a single monitor setup with a ~1080p resolution. I’m ultimately curious if it is related to your specific video setup. I can’t reproduce any issues here on a i7-6700 running 2 1080p displays.

Do you have any anti-virus software running other than windows defender? If you have any other heavyweight software running it might be interesting to see if closing it out helps any. This might be a long shot, but setting the PC to the high performance power plan might help, go to the “power options” menu in the control panel and set your power plan to “high performance”.

Regards,

Will you try something for me? It looks like you have a high resolution monitor? Maybe a 3K/4K monitor? Does it help if you try this same test after changing your PC resolution to 1080p? (or any resolution below 1080p)

Regards

c3dr,

What kind of PC are you running the unit on? I’ve seen this type of behavior before on a low power laptop. Usually indicative of the PC not able to keep up with the processing and creating erroneous results like this. For the real-time spectrum analyzer modes we recommend quad core i5/i7 processors to keep up with the processing requirements.

I look forward to your response.

Regards,
Andrew

Hi Steve,

Yes, the center frequency you select in the bbConfigureCenterSpan is the 0Hz/baseband frequency of the IQ samples and your IF bandwidth (27MHz) is centered around this frequency.

Regards,
Andrew

Hi stevecrm,

So you may indeed be pulling the I/Q data and plotting it correctly, but unfortunately it will still not produce the constellation plot you see in Spike. The reason for this is that Spike is doing additional processing between acquiring I/Q data from the API and plotting it in the constellation plot.

The main steps the Spike software is performing is resampling, symbol timing recovery, carrier recovery, and filtering. This can be quite a complex subject and not one I can cover in a forum post. You can find more information in a digital communications textbook, by searching online for search terms like “PSK demodulation”, or exploring software packages like GNURadio which has software blocks for the major components of a PSK receiver.

Keysight has an excellent block diagram of the various components in a spectrum analyzer when making PSK modulation measurements. See the following link on page 282.

http://literature.cdn.keysight.com/litweb/pdf/E4440-90620.pdf

I hope this helps, let me know if you have follow up questions.
Regards

Noeldi,

Good question, and it’s a bit tricky to answer. Our waterfall plot currently is only available in the sweep mode. Each sweep is plotted on the waterfall and generally each sweep is going to take 1ms+ depending on the analyzer and settings. So that would be your minimum resolution.

To get 10us resolution in a waterfall plot, we would need to introduce the waterfall plot into the zero-span mode (or some other I/Q mode) and perform overlapping FFT’s on the IQ data. We don’t have anything like this in Spike right now.

As Justin mentioned, our max bandwidth IQ device is 40MHz right now with the SM200A, or 27 MHz with the BB60C. Both are programmable and this type of resolution is possible if you were willing to program and work with the I/Q data yourself using our API.

We are working to increase the I/Q bandwidth of the SM200A to 160MHz for up to 1 second captures. This is still in development.

Regards,
Andrew

You can change the max file size allowed in the preference menu. It is limited to 1GB when running the 32-bit version of Spike.

There are several ways you could reduce file size depending on your needs, you can increase RBW, add a sweep interval, or take advantage of the decimation in time/freq on the recording toolbar.

If you need even more flexibility than what Spike provides, you could consider programming to the API directly, then you full control over the acquisition and recording process.

Regards

You can change the max file size allowed in the preference menu. It is limited to 1GB when running the 32-bit version of Spike.

If you need even more flexibility than what Spike provides, you could consider programming to the API directly, then you full control over the acquisition and recording process.

Regards

Hi Paul,

Are you referring to the spectrum plot in zero-span mode? or the time plot?

In both plots, the input power setting will play a large factor in the noise floor. The BB60C has maximum sensitivity at -50dBm input power/reference level. You want to try to set your input power setting to ~5dB above your expected input level to achieve the best dynamic range.

In the spectrum plot, the noise floor is also affected by the RBW. You can lower the RBW in that plot, but you might have to increase the length of the captures to achieve lower RBWs.

In the time domain plot, the noise of the AM samples will be affected by the IF bandwidth setting. You can expect the time domain samples to decrease by 3dB every time you halve the IF bandwidth.

Let me know if you have additional questions.

Regards

[Andrew]

Hi Dan0,

The spectral window in zero-span mode is plotted using a width equal to the sample rate. The rolloff you see is the software filter we apply to the data. The filter cutoff (6dB point) is selectable in the zero-span controls. The filter is a standard windowed sinc filter using the blackman window. The size of the filter does vary based on decimation and there can be cascaded filters at higher decimations.

Let me know if you have follow up questions.

Regards

• This reply was modified 6 years, 9 months ago by Andrew.
• This reply was modified 6 years, 9 months ago by Andrew.

Thanks for reporting this.

Go ahead and run the MSI directly and let me know if it installs.

Regards

Jon,

The 3D waterfall plot was removed in version 3.1.8, so I suspect 3.1.7 is the last version with it. I will link it below.

The plot is gone for now, we don’t have any plans to introduce it again.

I don’t think you can install two different versions at the same time. But to get around this, I would install 3.1.7 and then simply copy the whole Spike directory at C:/Program Files/Signal Hound/Spike/ to somewhere else. That Spike directory is fully standalone and is not dependent on any other files on the system. Then simply install the latest version of Spike. You can run the latest version of Spike from the normal location and when you need the 3D plot, run it from the directory you saved.

https://signalhound.com/sigdownloads/Spike/Spike(x64)_3_1_7.zip

Let me know if you have questions.

Regards,
Andrew

[Andrew]

Catalin,

Thanks for the heads up. I’ll fix this as soon as I can. You can also just send me a quick email if you see something like this.

Regards,
Andrew

• This reply was modified 6 years, 10 months ago by Andrew.
• This reply was modified 6 years, 10 months ago by Andrew.

Hello Andrew,

The calling convention for the BB60 functions is Cdecl.

I don’t know if I’ve heard of someone using Excel VBA, but I have seen C#, VB, VB.NET. I would imagine it’s possible.

Hopefully someone here can help.

Regards,

kullboys,

In the latest version of Spike, the waterfall plot has cursors and a timescale. In the process though, we removed the 3D plot so we can focus more on the 2D plot features. Additionally we added a configurable sweep depth and scroll bar. You can have a sweep history up to 50k sweeps.

You can find the latest version of Spike here
http://www.signalhound.com/Spike

In regards to your USB problems, what spectrum analyzer are you using? What kind of PC are you running it on? Do you have a make/model number for your PC and CPU?

Regards,
Andrew

Based on the screenshot, the all white background in the center of the application indicates there is a GPU issue/incompatibility.

Are you able to try updating the GPU drivers for your system? You will have to find them on the HP website. We may be using newer OpenGL features than you have available on the system. From the HP website, this PC supports OpenGL 2.0 which is the bare minimum required by our software. If none of the GPU driver updates fix the issue, I suggest looking for a newer PC to run the software.

Regards,
Andrew

HHonza,

Since each BB60C has different calibration data which controls the sensitivity of the receiver based on the reference level, it is possible for the same reference level to provide different gain/attenuation settings. A good test would be to set the reference level to -50dBm, match RBW/VBW/center freq and measure the level of the noise floor (trace averaging can help). You can also manually control the gain/atten by enabling the manual gain/atten settings (file menu).

Feel free to email justin@signalhound.com with any more information or pictures.

Regards,
Andrew

[Andrew]

Hello Nazar,

I’m looking at this function right now and there is only one location where this error value (-3) can be returned. The code for this is below.

if(device < 0 || device >= TG_MAX_DEVICES) {
return tgDeviceInvalidErr;
}

where TG_MAX_DEVICES = 4.

So if the device integer provided is less than 0 or greater then 3 then this error (-3) is returned. I would look into the interop between your C# application and this function. This integer argument is not being passed properly. You are using DllImport and setting the calling convention to Cdecl?

Regards,
Andrew

• This reply was modified 6 years, 11 months ago by Andrew.

[Author]

Posts