Issue with Ping driver?

Rayman

Code for robot waits for ping reading to be less than 12 inches and then do something.

But, a loop that only calls ping driver does not work...

repeat
    j:=ping.Inches(PING_Pin)
    if j<12
       (do something)

The Ping LED doesn't even turn on if you do this...

Adding a delay to the loop seems to fix the issue.

Maybe that's a "feature", but I'm tempted to call that a bug...

Mike Green

The library PING))) driver is very simple and written completely in Spin (about a dozen lines). How do you have the PING))) wired? It should be running off 5V and the signal line should have a 3.3K resistor between it and the Propeller I/O pin. What's your system clock?

Rayman

I'm using the P17 header on the WX Activity Board.
The jumper is set to 5V.

But, come to think of it, I didn't have this issue running the same code on an old Activity Board...
Strange...

Mike Green

The servo headers already have 3.9K series resistors. You shouldn't have additional resistance in series.

Rayman

Not using any series resistors... Counting on the ones already part of the Activity Board...

There must be some difference between old Activity Board and Activity Board WX.
Or, maybe I have a bad WX board...
Tried two different Ping modules, so I don't think there's a problem there...

I tried both P16 and P17 headers on the WX board and they both have this issue.
This isn't horrible, I just insert a delay and then it seems to work.

Still, left me scratching my head...

frida

If your ping is SPIN code then there are 11 code lines in it.
In your loop there are only 3 code lines, so it's much faster.
What happens when you call ping over and over before it's done with the first result?
Taken from:
http://obex.parallax.com/sites/all/modules/pubdlcnt/pubdlcnt.php?file=http://obex.parallax.com/sites/default/files/obj/id/454/Ping_Demo_- _Archive__Date_2011.06.13__Time_14.44.zip & nid = 454

Phil Pilgrim (PhiPi)

The Ping))) requires a minimum 200uS delay between the end of one measurement and the beginning of the next. Most program examples show a 100ms delay.

-Phil

Rayman

Adding a 50 us delay in the loop makes it work.

But, I'm 99% sure my old Activity Board worked without this delay.
(Wish I hadn't just gave it away now).

I suppose it's possible that I remember wrong and had some debug code in there that I removed and thereby broke the code...

If that's the case, then I'm back to thinking there should be a, say, 100 us delay added to the driver.
Or, some note that a delay is needed...

Phil Pilgrim (PhiPi)

I would recommend an even longer delay, say the 100ms that example programs use. The reason is that you want secondary echoes to die out before taking a new sample, as they could register as primary echoes in a sample requested too soon.

-Phil