Model 2400 SourceMeter Technical Note: 

Understanding the 2400 Memory Buffer



Technical Summary:

The Model 2400 has 5000 points of data storage. The 5000 point data memory is 

divided into two distinct 2500 point data buffers with quite different uses 

and quite different features. For some applications over the GPIB bus, it is 

possible to combine the 2 buffers to store up to 5000 readings.



Technical Details:

The specifications for the 2400 indicate the following for the Memory Buffer: 

"5000 readings @ 5 1/2 digits (two 2,500 point buffers). Includes selected 

measured value(s) and time stamp. Lithium battery backup." To fully 

understand the memory buffer architecture, the following details are given.



There are two 2,500 point data buffers in the Model 2400. One of the buffers 

is used to store data into the internal data (TRACE) buffer. This is the 

buffer that is used when you wish to store readings from either the front 

panel by hitting the STORE button or using the TRACE subsystem SCPI commands 

over the GPIB. This buffer is battery backed using a Lithium battery, 

allowing you to get your data back now or two years from now. 



When you create a sweep from the front panel using the 2400, the measured 

data is stored in the TRACE buffer automatically. The number of readings 

stored is based on the total number of points used in the sweep, up to a 

maximum of 2500 points. The sweep can not be extended past 2500 points by 

using the second buffer.



The second buffer, often called the Trigger Count Buffer, is used by the 

READ? or FETCH? SCPI commands over the bus. A 2500 point data buffer has 

been allocated for these commands. It is possible using one READ? or FETCH? 

command to take up to 2500 readings and return them over the GPIB bus. This 

buffer is not battery backed. If you turn off the 2400 power, you  lose the 

data. This buffer can not be combined with the TRACE buffer to get a total of 

5000 from the front panel. A unique feature of this buffer is that it is 

pipelined. In other words, while the 2400 is taking measurements, it can also 

be transferring data out of the READ buffer at the same time.



So how do I get 5000 readings? The technique can only be used over the GPIB 

bus. The method requires using both buffers. The user could program the 2400 

as follows:



*RST

TRACE:POINTS 2500

TRIGGER:COUNT 2500

TRACE:FEED:CONTROL NEXT

OUTPUT ON

INIT;:INIT

TRACE:DATA?;:FETCH?



The first INIT command in the sixth line stores 2500 readings into the Trace 

Buffer. The second INIT in the sixth line stores 2500 readings into the FETCH 

buffer. The last line of the program retrieves the 2500 readings from both 

buffers, giving you a total of 5000 readings. This is how you get a total of 

5000 readings across the bus.



There are two points to keep in mind when attempting to do this. First, a 

semicolon will separate the data returned from the TRACE:DATA? command and 

the FETCH? command. Data returned by a TRACE:DATA? command is separated by 

commas as well as data returned by the FETCH? command. A semicolon will be 

sent by the 2400 to separate the data from the two query commands. Your 

program must be smart enough to detect the semicolon as well as the comma 

when parsing the data.



Second, this method may not be appropriate in time critical applications. 

You could end up with a short period of time that separates the TRACE:DATA? 

operation and the FETCH? operation. A time critical applications may not be 

able to tolerate any time delays.



Please weight these issues carefully when developing your application.