Hi everyone,
    I'm trying a new approach now, with the same results.  I've 
programmed a basic stamp to pulse high for 10 miliseconds out of every 
second.  There is about 4.7 volts between ground (connected to pin 5) 
and the high pulse (connected to pin 1).  I still cat get any 
interrupts.  I know this must be something simple/stupid since from what 
I've read people can usually get this working.  If anyone would like to 
point out the obvious please do.


Thanks,
Brent Bartels

>     I'm trying a new approach now, with the same results.  I've
> programmed a basic stamp to pulse high for 10 miliseconds out of every
> second.  There is about 4.7 volts between ground (connected to pin 5)
> and the high pulse (connected to pin 1).

I am unable to get PPS working at all under Linux either, in a similar
situation.  No help from anyone here yet, but I'm working on it.

In your case, you might try using an op-amp or a TTL->RS232 converter to
get your voltage levels higher.

> interrupts.  I know this must be something simple/stupid since from what
> I've read people can usually get this working.  If anyone would like to
> point out the obvious please do.

I've written a polling program that finds the pulse okay, yet my PPS still
won't work.  I might refine the program to use interrupts and if so I'll
let you know.

-Chris

>     I'm trying a new approach now, with the same results.  I've
> programmed a basic stamp to pulse high for 10 miliseconds out of every
> second.  There is about 4.7 volts between ground (connected to pin 5)
> and the high pulse (connected to pin 1).

I am unable to get PPS working at all under Linux either, in a similar
situation.  No help from anyone here yet, but I'm working on it.

In your case, you might try using an op-amp or a TTL->RS232 converter to
get your voltage levels higher.

> interrupts.  I know this must be something simple/stupid since from what
> I've read people can usually get this working.  If anyone would like to
> point out the obvious please do.

I've written a polling program that finds the pulse okay, yet my PPS still
won't work.  I might refine the program to use interrupts and if so I'll
let you know.

-Chris

Christopher Strong wrote:

>>    I'm trying a new approach now, with the same results.  I've
>>programmed a basic stamp to pulse high for 10 miliseconds out of every
>>second.  There is about 4.7 volts between ground (connected to pin 5)
>>and the high pulse (connected to pin 1).

(...)

> I've written a polling program that finds the pulse okay, yet my PPS still
> won't work.  I might refine the program to use interrupts and if so I'll
> let you know.

Hi Folks--

Take a look at

http://www.wraith.sf.ca.us/ntp/

....down where the pps shm driver docs are. This shm_splc.c utility works with
linux 2.6.3(+) and linux 2.4.2x, as a shm driver with no linux patches,
and also with the ppslite or old (2.4.x) ppskit linux. I've put some
hardware debugging hints up on the page that explain how to use the
driver to verify your hardware setup.

A pure ttl signal can be run into your parallel port (pin 10, ground on 18),
or an RS-232 signal can be wired to DCD (pin 8 on a db-25), and the program can
use either source.

Once you have your hardware working this way it should also work fine
with your ppskit linux--on a uniprocessor. I'm not sure if ppskit linux
SMP works ok with ntp.

.../Steven