# What units are kpts?

I am parsing data from a Siglent SDS2000X oscilloscope. The first chunk (0x00 - 0x003) is for wave_length. The data sheet states that for arbitrary waveform generation, the wave length is "16 Kpts". What unit is Kpts? Kilo-Points? How would I convert that to a wavelength unit such as meters? Is a value of 3622260667 reasonable?

• You'd think that K wasn't kilo, but it probably is. – Chu Jul 2 '19 at 15:34
• @Chu Right, K is Kelvin. – Dmitry Grigoryev Jul 3 '19 at 11:44

K for "thousand"

pts for "points".

The unit is thousands of points.

How would I convert that to a wavelength unit such as meters?

You can't. This is the maximum number of samples that can be used by the waveform generator function of the instrument.

It isn't directly related to the wavelength of the signals the generator can produce (assuming the sample rate is adjustable, which isn't clear).

It does give relationship between the lowest and highest frequency components in the signal generated. This limit would be about 8000:1. This would also limit the minimum duty cycle if you were using it to produce low duty cycle pulses.

It doesn't say anything at all about the actual oscilloscope waveform measuring function, or the data produced when measuring a signal with the oscilloscope.

Is a value of 3622260667 reasonable?

3622260667 what?

If that's the frequency of your signal in Hz, then no, it isn't reasonable to measure this with the scope you linked to. The scope has a bandwidth of about 300 MHz. It won't be useful to capture a signal with frequency 10x that.

• k is for thousand. K is for Kelvin. But you are correct non the less. – winny Jul 3 '19 at 8:28
• @winny yes, in SI. But often in datasheets k is for decimal thousand, K is for multiples of 1024. I would be surprised if it didn't mean 16384 points. – Pete Kirkham Jul 3 '19 at 15:13
• @winny, and also it's just common for 'K' to be misused to mean 1000. – The Photon Jul 3 '19 at 21:39

It's not a wave length as in an actual wavelength. It's probably a contraction of waveform length! The amount of storage space the Arb generator (ha, another contraction!) has. That is you can store up to 16 k points (or individual samples) into its memory.

Although I have to admit, using typographic points (1 inch/72) to represent the wavelength could have some funny uses for millimeter- and sub-millimeter wave RF stuff; i.e. calling bands by the traditional point-size names.

Brevier band, anyone?

The oscilloscope cannot measure a wave length in a physical magnitude other than time. "Kpts" might be a thousands-of-samples count (which, given a sampling frequency, relates to time). If you're, e.g., measuring sound waves, you'd need to mix in the velocity of sound waves in the corresponding medium in order to convert it into metres.

To follow up with my own findings, pts indeed seems to mean points. What is a point? It seems to be a point on the oscillocope's display. It also seems that the K is for kilo, despite being capitalized. This can be seen in how the value changes from K to M on the scope itself. One would guess that M is for "mega". Finally, the value of 3622260667 would be in Kpts if there was data in that position. It turns out that when binary data is output, if no value exists, the oscope outputs garbage rather than something meaningful.