buzz

I've ordered more ID-20s recently and it seems the new batch (with "LV" prefix) from sparkfun exhibits weird issues.  You can read some of them in the ID-20 thread.  I've found that if you connect the reader to the 3v3 instead of 5v, it's almost 'normal'.  I just dont know if this is a good idea moving forward or how consistent it'll be.  I also saw another method using a resistor.

Finally got my ID-20 working well.  A big THANK YOU to everyone here who contributed...this stuff is truly fascinating.

I plan to run this RFID scanner in production with ~50 employees on it.  So after much stress testing, I'd like to contribute a few changes to the datareceived method to ensure maximum reliability.  Particularly with ID-20 reading multiple cards in quick succession.  Basically the code can't crash since no one may be around reboot it.

Changes I made:

- hopefully almost crashproof

- moved rawdata to local scope

- removed args.ReadTime, args.RFID output is now int, renamed class

- better error state output using args.Success (ie: to display error led's to user)

- better handles multiple/quick reads...basically when BytesToRead is not a perfect 16

- found that hexConvert was not as efficient as direct parsing of the paired hex

- removed cardPort.Write - this seems to echo the input...not sure why this was necessary...it seems to run fine w/o it and it speeds up the code quite a bit

private void _rfidCardDataReceived( object _cardPort, SerialDataReceivedEventArgs unused ){	// Do nothing if the reader is not supposed to be running.	if ( !_readerRunning ) return;	// Capture the port that the data was received on.	SerialPort cardPort = (SerialPort) _cardPort;	// Buffers for incoming card data.	byte[] rawData = new byte[16]; // Incoming raw bytes from the serial connection.	RfidEventArgs args = new RfidEventArgs();	string cardID = String.Empty;	int bytesToRead = cardPort.BytesToRead;	if ( bytesToRead == 16 )	{		cardPort.Read( rawData, 0, rawData.Length );		int[] chk = new int[6];		chk[0] = HexAsciiToInt( rawData, 1 );		chk[1] = HexAsciiToInt( rawData, 3 );		chk[2] = HexAsciiToInt( rawData, 5 );		chk[3] = HexAsciiToInt( rawData, 7 );		chk[4] = HexAsciiToInt( rawData, 9 );		chk[5] = HexAsciiToInt( rawData, 11 );		if ( ( chk[0] ^ chk[1] ^ chk[2] ^ chk[3] ^ chk[4] ) == chk[5] )		{			args.Success = true;			args.RFID = ( ( ( chk[1] << 8 | chk[2] ) << 8 | chk[3] ) << 8 | chk[4] );			OnRfidEvent( args );		}		else		{			args.Success = false;			args.ErrorMsg = "Checksum mismatch on RFID: " + ( chk[0] ^ chk[1] ^ chk[2] ^ chk[3] ^ chk[4] ) + " vs " + chk[5];			OnRfidEvent( args );		}	}	else if ( bytesToRead > 16 ) // bytesToRead exceeds 16 (ie: too many cards read at same time, nuke everything)	{		// clear buffer		cardPort.DiscardInBuffer();		args.Success = false;		args.ErrorMsg = "Serial port buffer exceeds 16 bytes: " + bytesToRead;		OnRfidEvent( args );	}}public int HexAsciiToInt( byte[] data, int offset, int length = 2 ){	int ret = 0;	for ( int i = 0; i < length; i++ )	{		ret = ret * (int) System.Math.Pow( 16, i ); // each digit is another power of 16		if ( data[offset + i] >= 48 && data[offset + i] <= 57 )			// number			ret += data[offset + i] - 48;		else			// letter A - F			ret += data[offset + i] - 55;	}	return ret;}

The full RFIDReader.cs...usage is similar to program code posted earlier in the thread except the class is renamed:

 RFIDReader.cs   4.28KB   9 downloads