All Forums
 Hardware
 General Hardware, ADSmartIO, Digital I/O and A/Ds
 AGX CPLD EXT_IRQ
 Forum Locked
 Send Topic to a Friend
 Printer Friendly
Author Topic  

lperletz

6 Posts

Posted - 28 Apr 2008 :  15:47:08  Show Profile  Email Poster
I would like to monitor the EXT_IRQ signal on the AGX board. Could you tell me the relevant CPLD pin numbers for this signal?
Thank you

akidder

1519 Posts

Posted - 28 Apr 2008 :  18:31:27  Show Profile  Email Poster
It looks like we didn't map EXT_IRQ in our Windows CE DIO driver, but you should still be able to map to it using the interrupt class topic 936.
Go to Top of Page

akidder

1519 Posts

Posted - 29 Apr 2008 :  09:53:39  Show Profile  Email Poster
My colleague rightfully points out that I didn't actually answer your question

The EXT_IRQ starts and J8 pin 24. It gets pulled up (10k to 3.3V), is ESD protected and goes through a series resistor. It ultimately ends up at pin 63 of the CPLD.

I'm not sure why you need to know the CPLD pin. Only the pull-up resistor should have any effect on the input signal. Let us know if you're seeing a problem with the signal.
Go to Top of Page

lperletz

6 Posts

Posted - 29 Apr 2008 :  12:12:57  Show Profile  Email Poster
Thanks for the info.

I have a case where two AGX boards respond differently to a 1PPS pulse applied to J8 pin 24. The ISR for the EXT_IRQ simply implements a running counter (keeping track of time). When a common 1PPS signal is supplied to each AGX, one the two boards shows a running count greater than the other board (at a rate of about 1 spurious count every two weeks).

As part of my investigation, I want to monitor the 1PPS from J8 pin 24 up to its input to the PXA 255. So, I would also like to know the CPLD pin that outputs the interrupt signal to the PXA 255

I have to assume that my software could be at the bottom of this problem, but in addition to reviewing the code, I wanted to take a look at the 1PPS signal as it travels to the processor.

Thank for for your help.
Go to Top of Page

akidder

1519 Posts

Posted - 29 Apr 2008 :  12:39:55  Show Profile  Email Poster
Understood. I guess your application can't tolerate that 1 ppm inaccuracy?

Drop us a line and we will get you the added detail you need to track this signal further.
Go to Top of Page

lperletz

6 Posts

Posted - 08 May 2008 :  09:53:16  Show Profile  Email Poster
After reviewing my code, I have found that my interrupt handler for the EXT_IRQ returns void whereas the the example code provided by ADS returns IRQ_HANDLED for LINUX versions at 2.4.23 and above. (I am at 2.4.27).
Is this a potential problem?

Also, I would like to repeat my request for some more help with my efforts to monitor the interrupt signals on the AGX:

I noticed that interrupt pulses appear at pin 60 in addition to those coincident with my 1PPS signal. I infer from a kernel header file that some other sources for the CPLD EXT_IRQ are: SSP, CAN, ETHERNET, EXT (my 1PPS), and UARTS 1 - 4. I note a lot more interrupts with the Ethernet cable attached to the AGX, but some non - 1PPS interrupts remain after the Ethernet is removed. I do not use CAN or UARTS 1 - 4

Could you tell me all the sources for the CPLD IRQ (pin 60) and their input CPLD pins? By the way, what does SSP stand for?

Thank you for your help.
Go to Top of Page

ctacke

877 Posts

Posted - 08 May 2008 :  09:56:50  Show Profile  Email Poster
SSP stands for Synchronous Serial Port
Go to Top of Page

lperletz

6 Posts

Posted - 08 May 2008 :  16:34:33  Show Profile  Email Poster
Thanks for decoding SSP (I had thought it was a reference to Synchonous Serial Port, except there is no mention of it in the AGX manual).

I would still like some help with the other two questions in my previous post:

1. What are all the sources that can produce a CPLD IRQ (pin 60) and what are their CPLD input pins?

2. Is there a possible problem with my not returnsing an "IRQ_HANDLED" in my EXT_IRQ interrupt handler?

Thank you
Go to Top of Page

rwhaley

628 Posts

Posted - 08 May 2008 :  16:48:55  Show Profile  Email Poster
Hi Larry,

1. Did you follow up on akidders earlier offer to work with you individually on the details of the CPLD irq signal components? If not, I would recommend that you take him up on this offer.

2. This is how IRQ_HANDLED is defined in 2.4 kernels:

#define IRQ_HANDLED

So, in 2.4 kernels there is no difference between these 2 lines:

return;

return IRQ_HANDLED;

Apparently IRQ_HANDLED is present just so 2.4 interrupt handlers can be similar to 2.6 interrupt handlers (where IRQ_HANDLED not null).

Go to Top of Page

lperletz

6 Posts

Posted - 09 May 2008 :  11:01:03  Show Profile  Email Poster
Thanks for the clarification of IRQ_HANDLED.

I have tried to follow up with akidder using support.us@eurotech.com but I have not received any replies after my first email.

Go to Top of Page

cbadr

154 Posts

Posted - 09 May 2008 :  11:44:54  Show Profile  Email Poster
Our records show that our support team did reply to your email on April 29. We have just received a new inquiry from you and we are working on getting you the answers you need. We try our best to reply to all inquiries within one or two days from the date we receive them.
Go to Top of Page

lperletz

6 Posts

Posted - 09 May 2008 :  12:00:49  Show Profile  Email Poster
Agreed. I did receive a very helpful reply on April 29.

I was spoiled by that very prompt reply.

I just wanted to be sure that my subsequent emails were received.

Thanks for your support
Go to Top of Page
  Topic  
 Forum Locked
 Send Topic to a Friend
 Printer Friendly
Jump To:
Eurotech Support Forums © Eurotech Inc. Go To Top Of Page
This page was generated in 0.05 seconds. Snitz Forums 2000