Tx dead or wot!

wolverine_349 · Joined: 07 Dec 2005 Posts: 6

My project involves RS-232 communiaction between pic16f689 and the PC.I m using H11L1 driver IC for providing schmiitt trigg.signal to the pic(Rx) and another H11L1 for converting TTL from the pic to RS-232 signal(Tx).The pic is able to receive data entered via terminal.But I haven't been able to receive
data from the pic.I have already tested if the pic is receiving data by putting data onto the LCD and it works.Wish the Tx was working too.i read the TXREG reg. on the LCD and it shows the 1st character that had to be transmitted,and the TSR is shown to be empty in the transmit status reg TXSTA. Is ther anyway i could trouble shoot the transmission from the pic. Would luv to have any kinda help on this.

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

Post a small (but complete) program that shows how you are
transmitting data to the PIC. Also post more details about
the circuit you are using with the H11-L1 chip, such as what
Vcc voltage are you using, and are you using a series resistor
with the diode, etc. If you do this, more people will be able
to help you. Link to device data sheet:
http://www.fairchildsemi.com/ds/H1/H11L1-M.pdf

Humberto · Posted: Wed Dec 07, 2005 2:29 pm

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

Humberto · Posted: Wed Dec 07, 2005 2:54 pm

The problem he declare is that he can�t receive any data sent from PIC >>> to PC,
that�s why I think the problem is in the PC Comm RS232 levels.
I would not assume that the PC has a Maxim chip inside.

Humberto

wolverine_349 · Joined: 07 Dec 2005 Posts: 6

[quote="Humberto"][quote]

The problem is that RS232 data is bipolar
+3 to +12 Volts for "ON or 0-state (SPACE) condition" and
-3 to -12 Volts indicates an "OFF" 1-state (MARK) condition.

Using the H11L1 optocoupler the output always swing between 0 to +5V and this way the PC Comm Port never gets the OFF state.

Humberto u r right about 0-5V swing,but what i didnt mention is that the o/p
of the optocoupler is connected to a npn transistor having VCC tied to a 12V rail. The tx of the pic is connected to the cathode(on H11L1) through 1K res. and anode is 5V.the

***************-----------12v
***************** | |
**************560<***<10K
*****************<***<
*******220********|*** |
<-----^^^--- -------- |****|
to pc rx*********** \****|
**************npn***|----------------< o/p from H11L1(pin 4)
*****************< /
*******************|
*******************|
*******************:gnd p.s.--->IGNORE *

I tried using the INVERT option for #use rs232 becoz of the npn trans ,but to no avail.dis is drivin me real crazy.more so becoz m reading the txreg onto the lcd and its showing chars that i am transmitting.so,must be the driver not being able to produce a compatible rs232 signal.

i will put the code next time around ,although m pretty sure it works.i have tried printf,putc,putchar ....everything.does ne1 know a way out of this?

treitmey · Joined: 23 Jan 2004 Posts: 1094 Location: Appleton,WI USA

Are you trying to use the invert option on a hardware usart pin?

This is not possible.

Humberto · Posted: Thu Dec 08, 2005 2:35 pm

Ttelmah · Guest

It is worth looking at how other things handle this. You do not need much
-ve voltage. So you could use a 7660 capacitor charge pump to get the
-ve rail, or steal some voltage from one of the handshaking lines (this is a common trick' technique), or even charge a -ve capacitor from the PC's serial output line. If you want to save space, and can get them (Maxim, can at times be a 'pig' as a source), you could try the DS275, which combines two of these techniques, to give aout -4v, and is adequate for many applications. +/-3v, is what is really required to 'meet' the spec. The extra voltage is to guarantee this signal level over longer wires. Also the load impedance is quite low (nominally 3KR), so it may simply be that your circuit (which is illegible - use the 'code' control to prevent the text from being rejustified), does not actually get the signal even to 0v...

Best Wishes

Guest · Posted: Fri Dec 09, 2005 11:14 am

hi all u guys, guess wot,tx wasnt dead afterall.i changed the H11L1 Vcc to 12v instead of 5V and voila!.now i finally have a full duplex communication Smile

.but its @ a price.by now i have changed 2 of these H11L1 ICs cause they seem to be going dead n all this in 1 day.dat too my application requires round the clock communication.am burning some serious cash on these ICs.dunno wots causing it ,since it can take Vcc upto 16V.
The 560ohm resistor,i s\had shown in the last post, is heating up beyond normal operation ,so am getting my fingers getting burnt too along with all the cash.could ther be any specific reason?forgot to mention,its been happening even when h11L1 vcc was @ 5V.
even though am happy all z working ,i would like to fix these bugs b4 they fix me.btw thanks u guys for pitching in with all that support.

wolverine

Humberto · Posted: Fri Dec 09, 2005 4:51 pm

Ttelmah · Guest

I'd guess that with the 12v supply, he was actually getting something like +2 to +12v 'out', and with the lower current drawn on the 5v supply, he is now getting perhaps +0.8 to 5v 'out', which is _just_ dropping the line low enough to now be seen by the PC.
As a comment to him, if you want this to be _reliable_ on anything beyond a very simple prototype system, stop 'fiddling around', and start generating signals closer to the proper RS232 levels. Otherwise it working, is basically 'luck'....

Best Wishes

wolverine_349 · Joined: 07 Dec 2005 Posts: 6

[quote="Humberto"]

Humberto · Posted: Sat Dec 10, 2005 3:29 pm

neil · Joined: 08 Sep 2003 Posts: 128

A MAX232 Chip is comparable in cost to a schmitt and an NPN. The only solution is to do it properly and not to marvel at the fact that it works when it really shouldn't!!!!

How about looking here: http://www.ccsinfo.com/forum/viewtopic.php?t=25018

and the last few posts of this thread: http://www.ccsinfo.com/forum/viewtopic.php?t=25010

You can see from here why we get frustrated. Using PICs requires electronics knowledge as well as programming knowledge!

Neil.