c# SerialPort problem, write() interrupted

[arjan321]

Hello all,

I have a strange problem with the .NET serialPort class.

When I write some data to the serialport, not all the data is
immediately send: every once in a while, only ~50 characters of the 64
characters I'm sending gets send.

Then, when I do a write() again, the data gets send. This is
unacceptable behavior for my program.

With only a console version, the problem occurs once every 100 times.
But with a UI with some events, this rate goes up to ~10%.

Are there any explanations for this?

The console-code:

static internal SerialPort Serial;
static void Main(string[] args)
{

Console.WriteLi ne("Welcome to this testapplication !");

Serial = new SerialPort("COM 4");

Serial.BaudRate = 9600;
Serial.Handshak e = Handshake.Reque stToSend;
Serial.Parity = Parity.None;
Serial.StopBits = StopBits.One;
//Serial.RtsEnabl e = true;
Serial.Encoding = System.Text.Enc oding.Default;
Serial.DataBits = 8;
Serial.NewLine = "\r";
Serial.Received BytesThreshold = 20;

Serial.Open();

byte[] values = new byte[256];

for (int j = 0; j < values.Length; j++)
{
values[j] = Convert.ToByte( j);
}

while (true)
{
for (int i = 0; i <= 192; i += 64)
{

send(values, i, 64);
Thread.Sleep(75 0);
Console.WriteLi ne("Another 64 bytes send..");
}
Console.WriteLi ne("----256 bytes send..");
Thread.Sleep(10 00);
}

}

static internal int send(byte[] p, int o, int c)
{

if ((o + c) p.Length)
{
c = p.Length - o;
}
try
{
Serial.Write(p, o, c);
}
catch (Exception e)
{
Console.WriteLi ne(e.Message);
}

return c;

}
}

The equipment I'm writing to uses CTS/RTS handshaking..

Kind Regards,

Arjan

[Doug Semler]

On Sep 18, 11:10 am, arjan...@gmail. com wrote:

Hello all,

I have a strange problem with the .NET serialPort class.

When I write some data to the serialport, not all the data is
immediately send: every once in a while, only ~50 characters of the 64
characters I'm sending gets send.

Then, when I do a write() again, the data gets send. This is
unacceptable behavior for my program.

With only a console version, the problem occurs once every 100 times.
But with a UI with some events, this rate goes up to ~10%.

Did you try Serial.BaseStre am.Flush() (don't know if it will work,
just a suggestion).

[arjan321]

On Sep 18, 8:56 pm, Doug Semler <dougsem...@gma il.comwrote:

On Sep 18, 11:10 am, arjan...@gmail. com wrote:

Hello all,

I have a strange problem with the .NET serialPort class.

When I write some data to the serialport, not all the data is
immediately send: every once in a while, only ~50 characters of the 64
characters I'm sending gets send.

Then, when I do a write() again, the data gets send. This is
unacceptable behavior for my program.

With only a console version, the problem occurs once every 100 times.
But with a UI with some events, this rate goes up to ~10%.

Did you try Serial.BaseStre am.Flush() (don't know if it will work,
just a suggestion).

Calling the Flush() function has the same effect as calling
writeLine(""): it'll transmit the buffers, most of the time..

Still, both ways are a work around for a problem that shouldn't be
there in the first place. Or do I have too high expections of serial
communication?

[arjan321]

On Sep 19, 9:13 am, arjan...@gmail. com wrote:

On Sep 18, 8:56 pm, Doug Semler <dougsem...@gma il.comwrote:

On Sep 18, 11:10 am, arjan...@gmail. com wrote:

Hello all,

I have a strange problem with the .NET serialPort class.

When I write some data to the serialport, not all the data is
immediately send: every once in a while, only ~50 characters of the 64
characters I'm sending gets send.

Then, when I do a write() again, the data gets send. This is
unacceptable behavior for my program.

With only a console version, the problem occurs once every 100 times.
But with a UI with some events, this rate goes up to ~10%.

Did you try Serial.BaseStre am.Flush() (don't know if it will work,
just a suggestion).

Calling the Flush() function has the same effect as calling
writeLine(""): it'll transmit the buffers, most of the time..

Still, both ways are a work around for a problem that shouldn't be
there in the first place. Or do I have too high expections of serial
communication?

Problem solved.. the hardware was holding data back.. Fixed the
hardware so the problem is gone :)