- filippomarchese November 6, 2019 at 3:14 am
We have developed a C# application that uses the SA device API.
We see that the same application, connected to the same device, has very different performance on 2 PC with the same hardware (Intel Core I7).
On the first PC the saGetSweep_64f call takes 200ms to return the data.
On the second PC the saGetSweep_64f call takes 400ms to return the data.
Is there any reason for this behavior?
AndrewModeratorAndrew November 6, 2019 at 6:39 am
Some other things you could check,
– Are they both running the same OS?
– Does one have active anti-virus?
– Are both using the high performance power plan configuration in the power options control panel menu?
– I’m assuming you have verified that the sweep configuration hasn’t changed? (same RBW/VBW/spur reject/etc)
– Is one i7 a laptop and the other a desktop? Laptops usually have slower clocks and might account for some of the difference.
– Is one application running in debug mode vs release mode?
An additional 200ms is a long time for the same sweep. The major things that contribute to sweep time is the device sweep speed limitations, USB transfer time, and processing time. Both USB and processing time should be small compared to the sweep speed limitations of the SA.
You could also compare your sweep time to the sweep time in Spike with the same configuration. If Spike sweeps substantially faster, it might be a configuration issue or something related to the C# FFI.
Andrewfilippomarchese November 6, 2019 at 7:08 am
thank you for the fast reply.
We have just checked using Spike with the same sweep configuration, and we see a similar behavior (200ms vs 400ms).
We have the same OS (Windows 10 Pro 64bit), the same active antivirus (Symantec), we have tried to change the power plan on both PCs with no effect, both the PC are desktop.
Any other thing that we can check?filippomarchese November 10, 2019 at 11:54 pm
Any idea about this issue?
Justin CrooksModeratorJustin Crooks November 11, 2019 at 8:48 am
Do you mind if I ask what the center frequency, span, RBW and VBW settings are? What is the model of the i7 processors? Is spur reject on for both cases?
It is possible for a sweep to take twice as long if the PC is significantly slower, and smaller variations are fairly common, but sweeping twice as fast on a similar PC seems unusual.filippomarchese November 11, 2019 at 11:59 pm
We are using the following settings:
Center 869.525000 MHz
Span 70.000000 kHz
RBW 100.000000 Hz
VBW 100.000000 Hz
Both the PC have a I7-6700T @ 2.80GHz processor.
Please have attached images containing the Spike screenshot and system data.
Some additional useful (I hope) information:
– disabling spur reject the sweep time decrease to ~50ms on both the PC
– the sweep time on the PC 2 (the one with the higher time) changes systematically opening and closing Skype (!!): 400ms with Skype open, 375ms with Skype closed
– the sweep time on the PC 2, SOMETIME, decrease to 200ms (like the PC 1) for a few seconds, then goes up again to 400ms for minutes
– we have checked the sweep time on other 2 PC with similar hardware, and the sweep time is ~400ms on them too: we have just 1 “luky” PC with a small sweep time
Attachments:You must be logged in to view attached files.
Justin Crooks November 13, 2019 at 10:31 am
- This reply was modified 3 years, 10 months ago by Justin Crooks.
For what it’s worth, I get 375 ms on my PC for the same sweep. Unfortunately, sweep time, especially on narrow spans, is not deterministic. There are settings applied, and then acknowledged, several times per sweep. So the timing will vary based on your USB read/write latency, which can vary for a lot of reasons. Spans of several MHz will generally be more consistent since the sweep is controlled more in firmware than software.filippomarchese November 14, 2019 at 12:03 am
Thank you for the feedback. Our bad luck was that the only “fast” PC is the one that we used for the development of the application.
We need a 70KHz span for our application. Does it make sense to use a wider span and then ignore points that we don’t need? Could we have resolution troubles?
What is the minimum span value that allows us to have a stable sweep time?
You must be logged in to reply to this topic.