18LF2620 serial port issue

BLL · Joined: 11 Nov 2006 Posts: 181 Location: Birmingham, UK

Hi, I am being driven mad by a serial port!
I have an 18LF2620 running on 3.3V, using a 10MHz crystal. The port is on RC6 and 7. It is set at 9600Bd. Connected to this port is a pl2303 serial to USB converter, which goes to my PC. I use a resistive divider on the rcv line, so the 5V of the pl2303's TX line is reduced to 3.3V for the PIC.
I have used the very simple test program from the manual
printf("start");
while(true)
putc(getc()+1);

From the terminal program, if I type characters at more than about 2 per second, I see the correct response. If I slowly type, nothing comes back until I type faster, when it catches up! The power line is stable, the PIC has a 0.1 at its supply pins.

What is likely to be causing this strange behaviour please?

Any help much appreciated.

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

post a schematic -
your COMPLETE minimum program that shows the problem
and compiler version.

too much to guess at otherwise.

BTW: why are you using brand-X instead of nice reliable FTDI FT-232 chip ??

asmallri · Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia

asmallri · Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia

temtronic · Posted: Thu Sep 10, 2015 5:20 am

couple of test options...

1) disconnect PIC and just loopback the USBTTL adapter and see what happens.....

2) Run the PIC at 5 volts and see what happens. This assuming it's rated for 5 volts, I didn't check...too many PICs, too little time !

BTW I've never had any issues with the $2 'knockoff' USBTTL modules.
Jay

ezflyr · Joined: 25 Oct 2010 Posts: 1019 Location: Tewksbury, MA

BLL · Joined: 11 Nov 2006 Posts: 181 Location: Birmingham, UK

hanks for the reply. The simple test code was in my first post. I would include an image of the circuit if I could see how to do it!! It's a jpeg but the insert image doesn't appear right. Having not included as image before, I am lost!

BLL · Joined: 11 Nov 2006 Posts: 181 Location: Birmingham, UK

To the gentleman suggesting loopback of the usb/serial adaptor, this works fine. I can't run the pic at 5V as there is an smt chip on the board which will fail if run over 3.3V, so that's not an option!!

ezflyr · Joined: 25 Oct 2010 Posts: 1019 Location: Tewksbury, MA

Hi,

Sorry for adding a bit of 'chafe' (above) to your thread Very Happy

!

To show us your schematic, you need to post the image to a free image hosting site (for example), and then post the link to the image here. Chances are you have a level incompatibility issue that we will be able to identify and help you to resolve once we see your schematic.

Example image hosting site: http://postimage.org/
_________________
John

If it's worth doing, it's worth doing in real hardware!

temtronic · Posted: Thu Sep 10, 2015 2:47 pm

Well good news is you can eliminate the PC terminal program and the USBTTL module as 'loopback' functions fine...

So...

Using resistors to lower the 5V USB rcv signal to the 3V PIC will work to allow it to receive data ..

But, and there's always a 'but'....

Your 3V PIC might not be outputting a high enough '1' for the USBTTL module to see as a proper '1'. If it needs 80% of it's VCC of say 5 volts, that's 4 volt the PIC would have to produce and that can't happen runnin git on a VDD of 3 volts.

You might also post your test program, could be a simple thing like not having 'errors' in the USE RS232(....options...).

Jay

asmallri · Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia

deleted because my mind was out of phase with reality.
_________________
Regards, Andrew

http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!!

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

do you connect like this?

https://commons.wikimedia.org/wiki/File%3A3_volt_interface_to_5_v_controller.pdf

if not, you might try it.

ezflyr · Joined: 25 Oct 2010 Posts: 1019 Location: Tewksbury, MA

asmallri · Joined: 12 Aug 2004 Posts: 1635 Location: Perth, Australia

Sorry ezflyr it is me suffering from dementia :-)

I mistakenly thought you posted this comment earlier: "BTW: why are you using brand-X instead of nice reliable FTDI FT-232 chip ??"
_________________
Regards, Andrew

http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!!

asmboy · Joined: 20 Nov 2007 Posts: 2128 Location: albany ny

my comment was not meant to disrespect Prolific USB bridge chips but in my direct experience - the PL 2303 is not the best part for the application.

1- the part has no meaningful distribution in north america - just a lonely sales office in Canada.

AND
2- has been plagued in the past with issues of using an unsigned
driver for windows.

On the other hand my consultancy and a manufacturing partner have been shipping a 3 port bridge design "universal bridge card" that is an any-port to all ports hub :: with TTL, RS-232 +/-8v signalling, and USB courtesy of an FT-232 chip. we've shipped thousands of these as an OEM vendor for use in very high end research instruments , deployed round the globe.

And in the nearly 4 years of shipping the product have had near zero failures and no issues with driver support EVER. including during the period when FTDI were ( misguidedly IMHO) trying to protect their I.P. through suspect technical means.

I was not pleased with the driver debacle FTDI initiated , but i did not have a dog in the hunt due to no problems for me or my customers.

And it was in nobody's interest for me to engage in "knee-jerk" re engineering fixes for a part that was causing no grief to my bottom line or my customers.

Perhaps it is possible that they are not the nicest company in their tactics, but i have never had an issue with the parts being genuine , matching the spec sheet and being of high reliability in the field. When you sell to customers who include your $30 product in their $50K instrument - you don't fix what ain't broke..... Very Happy