Transmission works with wires but not always with wireless

[jyi1]

I am using the Laipac 315 MHz RF modules and I serout only once and it works with wires 100% of the time. When I try wireless it works once in a while. On the other hand, when I serout 5 times the wireless works most of the time but does not work once in a while. I have added 4 $55s for a preamble and manchester encoded everything.

[skimask]

I am using the Laipac 315 modules and I serout only once and it works with wires 100% of the time. When I try wireless it works once in a while. On the other hand, when I serout 5 times the wireless works most of the time but does not work once in a while. I have add 4 $55 for a preamble and manchester encoded everything.

But yet we don't see any code to base an opinion on what could possibly be wrong!

[Darrel Taylor]

skimask,

Not all solutions require the persons code to solve the problem.

If you've used a "Laipac 315" before, then maybe you've noticed something that might apply.

He's made a very specific request, and shouldn't expect an answer for quite some time. Hopefully from someone that's tried that device before.

You're indignant response only 1 minute after his post, does not help.
<br>

[skimask]

skimask, Not all solutions require the persons code to solve the problem. If you've used a "Laipac 315" before, then maybe you've noticed something that might apply. He's made a very specific request, and shouldn't expect an answer for quite some time. Hopefully from someone that's tried that device before. You're indignant response only 1 minute after his post, does not help. <br>

True...but if the guy has the code in the first place, why not hasten the process and post it straight away!

And tell me that this thread/code and this one
http://www.picbasic.co.uk/forum/showthread.php?t=6145
don't look very similar...too similar...not that there's anything wrong with that. I mean, at least the guy (whether it be the same guy as the other thread or actually a new guy) searched the forum's and found some usable code instead asking for somebody else to write it all.
And no, I don't think everything is a conspiracy...or is it?

[jyi1]

Transceiver

Include "Modedefs.bas"

ct55 var byte: encoded1 var byte: p var byte: counter var byte: e var byte: action var byte:
encoded var byte: encoded2 var byte: action2 var byte: B0 var bit: B1 var bit: B0=1: B1=0:
p=0: e=0: encoded1=0: counter = 0: action=0: action2=0: encoded=0: encoded2=0:

Main:

Clr:
p=0: e=0: encoded1=0: action=0: action2=0: encoded=0: encoded2=0

Maina:
ct55=0 'counter for $55 is zero

Wait55:
serin PORTB.0, n2400, encoded1
If encoded1 = $55 Then
ct55 = ct55 + 1
If ct55 = 4 Then goto Waitaa
Else
goto Maina
Endif
goto Wait55

Waitaa:
serin PORTB.0, n2400, encoded1
If encoded1 <> $aa Then goto Maina
serin PORTB.0, n2400, encoded1
write 0, encoded1

For counter = 0 to 3 : p.0[counter] = encoded1.0[counter*2]:Next counter
write 1,p

e = p:
If ((e.3=B0) and (e.2=B1)) Then
goto L3
Else
write 2, p
goto Clr
Endif

L3:
If((p.1=0) and (p.0=0)) Then
action=%1010
write 3, p
For counter = 0 to 3
If action.0[counter]=0 Then
encoded.0[counter*2]=0 : encoded.0[counter*2+1]=1
Else
encoded.0[counter*2]=1 : encoded.0[counter*2+1]=0
Endif
Next counter:
write 8, encoded
For counter = 0 to 4
serout PORTB.7, n2400, [$55,$55,$55,$55,$aa, encoded]
Next counter
goto Clr


Endif

If ((p.1=0) and (p.0=1)) Then
action2 = %1011
write 4, p
For counter = 0 to 3
If action2.0[counter]=0 Then
encoded2.0[counter*2] = 0: encoded2.0[counter*2+1] = 1
Else
encoded2.0[counter*2] = 1: encoded2.0[counter*2+1] = 0
Endif
Next counter:
write 9, encoded2
For counter = 0 to 4
serout PORTB.7, n2400, [$55,$55, $55, $55,$aa, encoded2]
Next counter
goto Clr

Endif
End

Receiver

Include "Modedefs.bas"
cot55 var byte: encoded1 var byte: p var byte: counter var byte: e var byte:
p = 0: e = 0: encoded1 = 0: counter = 0:

Main:


Mainb:
cot55=0 'counter for $55 is zero

Wait55:
serin PORTB.0, n2400, encoded1
If encoded1 = $55 Then cot55 = cot55 + 1
If cot55 = 4 Then
goto Waitaa
Else
goto Wait55
Endif
'goto Wait55

Waitaa:
serin PORTB.0, n2400, encoded1
If encoded1 <> $aa Then goto Maina
serin PORTB.0, n2400, encoded1
write 0, encoded1

For counter = 0 to 3 : p.0[counter] = encoded1.0[counter*2]: Next counter
write 1,p

If (p.1=1) Then
goto B4
Else
write 2,p
High 2
Pause 1000
Low 2
goto Mainb
Endif

B4:
e = p
write 3, e
High 1
Pause 1000
Low 1
High 5
serout PORTB.3, n2400, [e.3]
serout PORTB.4, n2400, [e.2]
goto Mainb

End

[Jerson]

I am using the Laipac 315 MHz RF modules and I serout only once and it works with wires 100% of the time. When I try wireless it works once in a while. On the other hand, when I serout 5 times the wireless works most of the time but does not work once in a while. I have added 4 $55s for a preamble and manchester encoded everything.

JYI1:

Firstly, I feel a name is more suited to talking in a forum. Anyway, after seeing your code, I do feel you can look for the 55,aa alone as the marker to your data. You may not see the 4x55s followed by the aa considering that the Laipac receiver module is ASK(correct me if I am wrong).

I have used similar modules with PIC and implemented a wireless joystick which works like you say almost 100% of the time. There is an occasional hiccup due to which I would not get 100%. To achieve this, I know that RF shielding of the receiver is priority #1. I did not Manchester code my data, that could be the reason for the almost 100% accuracy. Also, the modules I have used (Summitek) specify the data rate in the region of almost 1200 baud and not 2400 baud as you use. Perhaps you need to look into this as well.

Jerson

[jyi1]

Yes Laipac uses ASK and has 2400 bps. I do not understand why the $55,$55,$55,$55,$aa is not always picked up by the receiver when I send it once. It seems like I need to send it the same thing 5 times for it to pick up reasonably well.

[Jerson]

The module needs time to get above the noise floor before it can decode the $55 stream correctly. That is why I suggested to just look for the $55,$AA in your code; you still have to transmit the $55s before you send the $AA. Those should be considered as padding and ignored in the receiver. You may increase it beyond the 4 that you have now to improve performance if you will.

Transmit $55,$55,$55,$55,$55,$AA,"data"
Receive .................."ah there it is","data" You ignore the first 4x$55


Jerson

[jyi1]

So I should basically send a couple of $55 and then just wait for $55, $aa only?
I'm thinking of coding it like: serin2 PORTB.0, 16780, [WAIT($55, $aa) ,encoded1], would this be good enough?

[Jerson]

I think you got the point. Still, tell us your name. JYI1 sounds like a name from the new StarTrek series

Jerson

[languer]

I am using the Laipac 315 MHz RF modules and I serout only once and it works with wires 100% of the time. When I try wireless it works once in a while. On the other hand, when I serout 5 times the wireless works most of the time but does not work once in a while. I have added 4 $55s for a preamble and manchester encoded everything.

I think skimask has a very good point. You started with the statement above and are now deeper than you could ever imagine. There is a subtle truth to the KISS principle, it works and it is simple.

• What are your expectations?
• Do you have something that works 4-out-of-5 times? I mean you send the message 5 times and you receive it 4.
• Where you expecting the link to work 100% of the time? If so, do you know what your link budget is? Do you now how much margin you have? What distance you need to cover?
• I do not know of any comm system that works 100% of the time. The best systems (very expensive) use redundancy, error correction, complex modulation schemes, etc. Normal systems (i.e. remotes, sensors, telemetry) usually just encode and send several times.
• You pick what you think you need in terms of reliability and no more usually. If you have way more, you over-designed it (your wasting battery power, causing unintended radiation, or something else).

I am not trying to persuade you from making something better. But seeing how complicated it is getting, do you really need it this complex? Or can you keep your manchester coded signal (which apparently is already working) send it a few times and only use the first one that passes a CRC check or something like that?

[skimask]

I think skimask has a very good point. You started with the statement above and are now deeper than you could ever imagine. There is a subtle truth to the KISS principle, it works and it is simple.

And one other point I might make...
The 432-433Mhz modules seems to work, the 315 doesn't.
Why not use 2 432-433Mhz modules, one master, one slave. Sure you can't use them both at the same time, bidirectionally. But for instance, maybe the slave will ONLY answer after the master has spoken. Use a PICs pin to provide power to the modules, turn off the master's receiver while it's transmitting, turn off the transmitter while it's receiving, and so on with the slave, only the other way around in reverse...
I think I get what the end result is supposed to be, with changing all the bits around and sending the modified code back. But why not let the PIC handle the bit changing after receiving a certain code or something along those lines.

OK. You're crazy!
__________________
www.davehouston.net

YES!!!