Virtualbench oscilloscope timebase

[Bigwan2062]

Really liking the virtualbench unit I've been evaluating over the last few days, but I find the maximum timebase of the oscilloscope (1s/div) limiting for my applications. Any ways/plans to extend this in the future? (A totally unfair comparison, but I'm used to using an Agilent DSO7034B with max timebase of 50s/div and frequently use that timebase...)

[ZachHindes]

Right now, no, there is only a maximum of 1s/div.  You could use the LabVIEW API to acquire data from the scope and have more fine-tuned control over acquisition rate and time.

[Bigwan2062]

Hi Zach,

If I use Labview to drive the hardware of the virtualbench, and ignore the Virtualbench app (which would be what I would be doing - I'm using Labview to control the aforementioned Agilent scope) then I COULD acquire a longer trace than 1s/div in software? Could I, for example, capture 100 seconds of data from the MSO?

Regards

Ian

[WilliamEarle]

The acquisition time is limited by how long it takes to fill up the buffer (1MS). The maximum time the scope can acquire for is 1e6/(sample rate). So for example, at 1GS/s, you can acquire for ~1ms before filling up the buffer.

The slowest the scope can acquire is approx 15.259kHz. At this rate, you'll get the longest acquisition possible for VirtualBench, which is 1e6/15.259e3 = ~65 sec.

Regards,

William Earle

NI R&D

[Bigwan2062]

Thanks William,

That's REALLY helpful! Now I assume that only applies to a single channel - do you divide that down for every additional channel (analogue and digital)?

Regards

Ian

[WilliamEarle]

There's a 1MS buffer backing each channel, so, it's 1e6/(sample rate) regardless of how many channels you have on.

Note if you turn on both analog channels, the maximum sample rate is reduced from 1GS/s to 500MS/s.

My reply assumed you were only using analog channels. With digital channels, it's a little more complicated. The digital sampler only writes a sample to the buffer whenever something on the input pins changes. So, if you have a signal with very short bursts of activity and then long idles, the digital channels may be able to acquire much longer than 1ms at a 1GHz sample rate.

If you use both analog and digital channels, the acquisition is constrained by the analog channel limits.

The VirtualBench API asks for an acquisition time. For the analog channels, we convert this to a number of samples given the sample rate, and simply run the hardware until we've latched that many samples. For the digital channels, how long until it takes to fill up the buffer depends on your input signals, and is not known until runtime.

If your signal is idle, you may not fill up the buffer at all. In this case, we have a timer that stops the digital lines once the acquisition time you ask for though the API is met. This timer runs on a very fast clock, and as a result has a hard time limit of 4.29 sec for pre-trigger time and 4.29 sec for (acquisition time - pre-trigger time). So, the longest the digital channels can acquire for is ~8.58 seconds.

If you have an acquisition with both analog and digital channels enabled, and set it up to acquire for 65 sec, you'll get 65 sec of data from the analog channels and 8.58 seconds (at most) of data from the digital channels.

[Dr_Edward_D4H]

Was there any progress made on the 4.29s limit of the Logic analyzer?

I am currently using the Logic Analyzer to monitor a logic signal for periods of time much great than 4.29s. Dealing with this limit we have tried to call the Function in a loop with an acquisition time of 4s increments but reading the buffer takes 300-400ms. This data is then lost to the ether.

Is there a work around that anyone has found?