9. Choosing the Fastest Communication Interface

There are many factors that influence selecting a physical communication interface. In this chapter, we focus solely on one parameter – speed.

When considering the speed of any communication unit, you always have to split the general term Speed into two different parameters: Latency and Bandwidth. To explain their meaning, we use an analogy of a water pipe:

The length of the pipe defines its latency – the time required to transport small unit of water from point A to point B. Therefore, the latency is measured in seconds.

The diameter of the pipe defines its bandwidth – the amount of water per second that can flow through its cross section. Thus, the bandwidth is measured in volume/second.

From this analogy, you see that a small amount of water (small data) is transported faster with a short pipe (small latency). Its diameter (bandwidth) does not play a significant role. In contrast, a pipe with big diameter (high bandwidth) transports large volume of water (big data) faster, even if it is longer (high latency). Let us have a look at the Latency and Bandwidth of the most common communication interfaces:

From the above, we can conclude the general rule:

If your remote control application performs many small data exchanges between your control PC and the instrument, for example querying many oscilloscope amplitude measurements, the GPIB or the USB interface is the optimal choice due to their low latency.

If you transfer big data volumes, for example entire oscilloscope waveforms with tens of thousands of samples, you achieve the shortest transfer times with 1GB LAN or the USB interface due to their high bandwidth.

Keep in mind that the overall performance of you remote control application is affected by many other factors besides the speed of the communication interfaces. In practice, very often delays caused by other factors that outweigh delays caused by any communication interface (see Chapter 8: Speed Optimization). The purpose of this chapter is just to give a general idea on how different interfaces behave in terms of speed under different data volume conditions.

Notes

*Note1 - Latency values:

The latency mentioned in the table above may vary significantly depending on instrument's internal architecture. The numbers mentioned in the table are pure interface latency values. The only way to find out the latency of your measurement setup is to perform an evaluation using for example the VISA Trace Tool. One good latency measurement procedure is querying instrument's *IDN? string, because of the command's quick processing time.

LAN VXI-11 vs HiSLIP:

HiSLIP protocol provides smaller latency times compared to VXI-11. Always prefer HiSLIP before VXI-11 if your instrument supports it. The detailed elaboration and list of Rohde & Schwarz instruments supporting HiSLIP you can find here: 1MA208: Fast Remote Instrument Control with HiSLIP