Screen capture on tps2000 series oscilloscopes

[abeaver]

Hi,

I have been been trying to solve this problem for the past few days. I have a Tektronixs TPS2024B oscilliscope that uses RS232 to communicate with Labview 8.0 and 2012 (both have same result). The goal is to obtain a screen capture using Tek's HARDCOPY commands. I can reicive a bitmap but it's only a portion of the bitmap and the rest is black. See picture and code below.

You can see the oscilliscope take the hardcopy and there are obviously bytes being written to the bitmap file from the read buffer. I have been trying to place delays all over the place to see if that was the issue but with no avail. 

Anyone have any suggestions? Thanks.

[Todd_Lesher · Accepted solution]

Instead of reading 1 million Bytes, read a few hundred at a time in a tight loop.

[abeaver]

And just concatenate the read buffer each time? How do you know when to stop?

[crossrulz]

You might want to check out this thread on performing a screen capture on the MSO4k series scope:http://forums.ni.com/t5/Instrument-Control-GPIB-Serial/Screen-capture-of-Tektronix-MSO4104B-using-La...

But since you are using serial, you need to read the data a lot more quickly to avoid the serial buffer from overflowing.  Send the Hardcopy Start command and then inside of a while loop just read 1MB data (or whatever your serial buffer size actually is).  Use a decently long timeout on the VISA Read (a little longer than it will take to move the 1MB of data over the serial port).  Take the string out of the VISA Read and write it directly to disk.  USE NO WAITS INSIDE OF THIS LOOP.  You will exit your loop when you have a timeout on the VISA read.

You will also want to open the file before starting the loop and close it after the loop.

[abeaver]

---

@crossrulz wrote:

You might want to check out this thread on performing a screen capture on the MSO4k series scope:http://forums.ni.com/t5/Instrument-Control-GPIB-Serial/Screen-capture-of-Tektronix-MSO4104B-using-La...

But since you are using serial, you need to read the data a lot more quickly to avoid the serial buffer from overflowing.  Send the Hardcopy Start command and then inside of a while loop just read 1MB data (or whatever your serial buffer size actually is).  Use a decently long timeout on the VISA Read (a little longer than it will take to move the 1MB of data over the serial port).  Take the string out of the VISA Read and write it directly to disk.  USE NO WAITS INSIDE OF THIS LOOP.  You will exit your loop when you have a timeout on the VISA read.

You will also want to open the file before starting the loop and close it after the loop.

---

[abeaver]

---

@Todd_Lesher wrote:

Instead of reading 1 million Bytes, read a few hundred at a time in a tight loop.

---

Here is how I did it. Unfortunately it takes about 25 seconds to transfer 20KB of data with a 9600 BR. Would anybody know why that is or how to speed it up?

Hopefully, this will save people loads of time in the future...

[abeaver]

I optimized my solution a little bit more to detect an error timeout rather than cacluating times. Errors are generated during the inner while loop because the VISA read reads exacly 150 bytes... that is more of a warning than an error. 

Also I "flush" the serial line by reading 1MB in the flushserial.vi before calling the tps2000screener.vi. 

[crossrulz]

I don't see the point of the sequence structure inside of your case structure.  There's nothing in there that needs sequenced.  Just trying to help clean stuff up.

[abeaver]

Yeah, not required. As you can see from the previous post it was left over from my timing structures. Good catch

[RavensFan]

---

@abeaver wrote:

. Unfortunately it takes about 25 seconds to transfer 20KB of data with a 9600 BR. Would anybody know why that is or how to speed it up?

---

That sounds about right.  9600 baud means 9600 bits/second.  There are about 10 bits per byte, so 960 bytes per second.  We'll round to a thousand.

20kB means it will take more than 20 seconds.