1 us

ccmcca · Joined: 08 Oct 2005 Posts: 13

Hello i am working with a GPS module that have 1 pps and the duration of it pulse is 1 MICRO SEG (1us).
The question is ....
Can i capture this little pulse with the RB0 of my 18f252, with an external interrupt as #INT_EXT???
How long must be the pulse, for capture on #INT_EXT of my pic???
THANKS
Excuse me but i speak spanish....

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

This information is in the PIC data sheet. Look near the end, in the
section on "Electrical Characteristics". In Table 22-6, it shows the Tinp
parameter: "INT pin high or low time". The minimum value is Tcy.

Tcy is 1/4 of the oscillator frequency. So for a 1 us pulse, you need
a minimum PIC oscillator frequency of 4 MHz. It's probably safer if
you ran the PIC at a higher frequency, because maybe your input
pulse is not always at 1 us. What if it's 0.9 us, sometimes ?

It's safer if you run the PIC at 8 MHz or more.

ccmcca · Joined: 08 Oct 2005 Posts: 13

I have a 10 Mhz osc with 4xx - 40Mhz
But i cant detect this pulse.
I dont know whath's going on...

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

What are the voltage levels of the pulse ? (high and low levels).

ccmcca · Joined: 08 Oct 2005 Posts: 13

0v LOW
2.85V HIGH

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

What's the Vdd of your PIC ?

If it's +5v, the minimum Vih voltage for the INT_EXT pin is 4.0 volts.

ccmcca · Joined: 08 Oct 2005 Posts: 13

Yes it is 5v.
What i have to do ....put some pullup or a transistor....?
THANK!!!!!!!!!

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

To convert a TTL level to a 5v CMOS level, you could use a 74HCT125
(with a Vcc voltage of +5v for the 74HCT125).

See this thread and the links in it.
http://www.ccsinfo.com/forum/viewtopic.php?t=30433&highlight=level+conversion

ccmcca · Joined: 08 Oct 2005 Posts: 13

MUCHAS GRACIAS AMIGO

THANK A LOT!!!!!!!!!!!!!!!!!!!!!!!!1

RossJ · Joined: 25 Aug 2004 Posts: 66

Another solution you might consider is to simply use the port B interrupt on change capability of RB4-RB7. They have the same timing capability as INTx but are TTL input buffered instead of Schmitt Trigger. As a result, the minimum logic high level is only 2.0v instead of the 4.0v of a ST input (assuming Vcc is 5.0v).

The 2.85v you mention should be more than enough to directly drive RBx. The only thing to consider is whether the 2.85v output can ever be lower than 2.0v whilst driving high. Hopefully the GPS module datasheet can tell you this.

ccmcca · Joined: 08 Oct 2005 Posts: 13

Ok I wil tried with the RB4....
Because less components is much better.
Thanks

ckielstra · Joined: 18 Mar 2004 Posts: 3680 Location: The Netherlands

RossJ · Joined: 25 Aug 2004 Posts: 66

You're not missing anything... that's exactly the point. Because ext0/1/2 pins are Schmitt triggered they are incompatible. Only pins with TTL inputs are suitable.

I have this problem when connecting 3.3v radio comms modules to the PIC because the USART RXD pin is always Schmitt Trigger buffered. So I need a level converter. When I use a 'software UART', I can use TTL pins and no level converter is required.

/Ross.

RossJ · Joined: 25 Aug 2004 Posts: 66

Also, your point about the RB4-7 all triggering interrupt on change events is a good one. The best solution is to configure the remaining 3 pins as outputs (whether you use them or not). This ensures they have no impact on the GPS module pulse detection.

One more thing... since RB4-7 is 'on change' and not rising/falling edge triggered, you may get an interrupt for both edges. This is unlikely to be the case with a pulse of only 1us, even at 40MHz system clock. But it could be an issue if the pulse can be longer (or you poll the flag and clear it immediately). A simple solution to ensure this is never a problem is to wait for the pulse to finish before clearing the interrupt flag. This can be done regardless of whether you use polling or interrupts.

/Ross.