Problem with manchester code for RF

[Dhanushka]

Hi dhouston and all,
I'm using ASK RF module, which is similar to dhouston's "http://davehouston.net/modules.htm" list. its antenna 18cm long
wire for both transmitter and receiver. No any other transmitter or some thing operate nearby.

Thanks for your example thread, I'll try using that NEC protocol.

Dhanushka

[mackrackit]

The distance problem is most likely from the wrong antenna length.

315 MHz requires a 22.619 cm antenna. 1/4 wave length. And yes, a couple of cm will make a very big difference.

The 1/4 wave antenna needs to be worked against a ground plane.

[dhouston]

Hi dhouston and all,
I'm using ASK RF module, which is similar to dhouston's "http://davehouston.net/modules.htm" list. its antenna 18cm long
wire for both transmitter and receiver. No any other transmitter or some thing operate nearby.

ASK and OOK are essentially the same. The noise issue is illustrated here...

• http://davehouston.net/rf-noise.htm

It's easy to use your PC soundcard as a storage oscilloscope to view the receiver output. See...

• http://davehouston.net/learn.htm

As mackrackit has noted, something around 22-23cm antenna length will work best for 315MHz. The exact length is more critical for the transmitter than the receiver but cutting both to the appropriate length works best.

[Dave]

dhouston, How can you say: Manchester encoding is not immune from noise but does have built-in error detection. What is the "built-in error detection"? This is news to me as I write software for Tire Pressure Monitoring equipment which uses Manchester coding. Please explain.....

Dave Purola,
N8NTA

[dhouston]

dhouston, How can you say: Manchester encoding is not immune from noise but does have built-in error detection. What is the "built-in error detection"? This is news to me as I write software for Tire Pressure Monitoring equipment which uses Manchester coding. Please explain.....

If you don't see a change of state at the middle of a bit period, you've detected an error which is usually the result of noise if you are using it over a wireless link or a collision if you're using it over a wired link (e.g. 802.3, X-10). See the last paragraph at http://www.wildpackets.com/support/c...ignal_encoding.

[Dave]

dhouston, I call that type if "error detection" signal integrity, not error detection. You could have a noise source producing edge changes at the same rate as the baud rate you are trying to detect. Where is the error detection then?

Dave Purola,
N8NTA

[dhouston]

dhouston, I call that type if "error detection" signal integrity, not error detection. You could have a noise source producing edge changes at the same rate as the baud rate you are trying to detect. Where is the error detection then?

Call it what you like. I'll call it what it is.

[dhouston]

You could have a noise source producing edge changes at the same rate as the baud rate you are trying to detect. Where is the error detection then?

And where is the error if no low state gets changed to a high state?

[Dave]

dhouston, Try deciphering 1 individual transmission of 128 data bits (256 states) with 4 or more transmitters being keyed at the same time.... The data states at the receiver might look good but what about the actual data..... Like I said... Signal integrity

Dave Purola,
N8NTA