LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Serial Communication delay

I am experiencing a delay when I run my commands through LV7.
The following is my pc settings:
Baud rate:57600
bits: 8
Parity: 1
stop: 1
no flow control

However when I run the exact same command on hyperterminal I get no delay.
I am writing 15 bytes at a time. Is there anything I can try to fix this issue.
0 Kudos
Message 1 of 19
(4,817 Views)
Are you configuring the port and then closing it each time in the loop? This would cause extra delay. If not then I don't have a clue.
- tbob

Inventor of the WORM Global
Message 2 of 19
(4,808 Views)
No I am not. I only open it at the beginning of my program and only close it at the end (Exit is pressed).
0 Kudos
Message 3 of 19
(4,796 Views)
Hi Preitano,

How do you do the reading? Is it in a loop? If so, is there a delay or other functions (vi's) in that loop?

If I'm not in the right direction, can you provide details to how you implemented the read. For instance, you could post a small screen capture of that section of code.

Ray
Message 4 of 19
(4,792 Views)
The following is my vi for writing and reading.
0 Kudos
Message 5 of 19
(4,790 Views)
What about the timeout? I can't remember what the default value is (no LV on tis PC).

I will try to describe what I mean...

When you do a serial read, you need to find out the number of bytes available at the buffer to be read. If for some reason, your value to be read is higher than the actual number of bytes available to be read, then the vi simply waits until one of the two conditions:
1. the amound of data reaches the number of bytes to be read, or
2. it times-out and completes the read with whatever data was available in the buffer.

However, let's find out how you implemented your vi to find out how to reduce / eliminate unwanted delays.

Ray
🙂
Message 6 of 19
(4,687 Views)


@PReitano wrote:
The following is my vi for writing and reading.




Thanks Preitano.

Unfortunately, I do not have LV on this machine, can you post a screen capture.

Sorry...

Ray
Message 7 of 19
(4,789 Views)
Hi,

I posted the screen capture in a word document.
0 Kudos
Message 8 of 19
(4,776 Views)
Hi Preitano,

I had not considered that implementation.

Let's walk through the code:

1. Write to the serial port (send command, etc).
2. Wait 200 ms
3. Either wait for timeout or have data at the serial buffer.
4. Read until serial buffer is empty.

I assume there are no delays inside the True case, simply a read serial port and an append to string.
And you use the value from "Bytes at Port" to set how many bytes to read. Seems okay.

Have you ever noticed loss of data? ie: not all messages are received?
It is true that you are reading at 57600. I am used to much slower speeds, such as 9600.

I don't see why you would have noticeable delays as compared with HyperTerminal. The only concern I would have is if some of the messages are lost due to not having a delay.

The thing to consider is that Hyperterminal (or other such sw) will read the port continuously and never stop.
In LV, you read until -- something --... then stop reading to process the data (make decision, etc.). And there is a bit of overhead because of it. It is this overhead that adds up with every iteration of the loop. Maybe someone can correct me, but I seem to remember that it is approx 4ms per action. So every iteration of your loop probably takes approx 20ms. Which is not that much... but it adds up quickly.


In terms of delays, what is the difference? Is it consistent? Does it slow down?

To trap the delay, you may need to be clever. The "Highlight Execution" is not your friend in this case 😉

However, if for some reason or other, the number of bytes reported by the buffer is not accurate, the "timeout" of the VISA Read may kick in and cause excessive delay. The only way to confirm this would be to compare the two values (number from "Bytes at Port" and actual bytes read).

Hope this helps..

Ray
Message 9 of 19
(4,776 Views)
Hi JoeLabView,

I am currently not experiencing any loss of data. The program seems to be working quite well, except for the delay.
The delay is consistent and visible on the scope when a write is taking place. In my program the write and read sections of the "serial read.vi" is split because I don't always need to read after I write.
0 Kudos
Message 10 of 19
(4,768 Views)