Skip to contentSkip to conversation

EP405 GPIO signal problem

DISCUSSION Toni Van Remortel 2 participants 4 messages Nov 20, 2003 – Nov 21, 2003

lore
Share & tools
lore
#0 lore raw

Hi,

I have an RTAI powered kernel module which outputs 32 bits data via the EP405 GPIO ports from the IBM 405GP CPU. These outputs are multiplexed with other signals.
Problem: when I output the 32 bits as one like this:
writel( buffer, (u32)gpio_base_addr+GPIO0_OR );
(where buffer is of u32 type)
I get some glitches after my DAC. The DAC has a 70ns settling time, but the glitches are 120µs. So I presume it's a CPU problem.

Is this a normal thing? I don't think so. It is strange behavior, that's all I know. How can I prevent those 'peaks' in my output? Peaks are generated by 0xFFFFFFFF output (that's what I can make out of the scope-signal).
--
Toni Van Remortel
Wetenschappelijk Medewerker - D-science lab
Real time Linux for embedded systems: http://linemb.d-sciencelab.com

              Tel: +32 3 205 61 72 - Fax: +32 3 205 61 95
                      E-mail: [email protected]

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#1 lore raw

Op do 20-11-2003, om 15:02 schreef Toni Van Remortel:

Quoted 2 lines

I get some glitches after my DAC. The DAC has a 70ns settling time, but the glitches are 120µs. So I presume it's a CPU problem.

Which isn't. Just checked the digital signals and those are perfect. So it is my DAC.
--
Toni Van Remortel
Wetenschappelijk Medewerker - D-science lab
Real time Linux for embedded systems: http://linemb.d-sciencelab.com

              Tel: +32 3 205 61 72 - Fax: +32 3 205 61 95
                      E-mail: [email protected]

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#2 lore raw
In reply to Toni Van Remortel
Quoted 6 lines · Toni Van Remortel

I have an RTAI powered kernel module which outputs 32 bits data via the EP405 GPIO ports from the IBM 405GP CPU. These outputs are multiplexed with other signals.
Problem: when I output the 32 bits as one like this:
writel( buffer, (u32)gpio_base_addr+GPIO0_OR );
(where buffer is of u32 type)

One thing which looks strange to me. Are you sure you can use 32 GPIO pins?

This is impossible because 405GP has only 23, and I doubt you can use even this number, because some of the pins are shared with IRQ and PerCSx lines.

Eugene

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#3 lore raw

Op do 20-11-2003, om 19:19 schreef Eugene Surovegin:

Quoted 10 lines · Toni Van Remortel
Quoted 6 lines

I have an RTAI powered kernel module which outputs 32 bits data via the EP405 GPIO ports from the IBM 405GP CPU. These outputs are multiplexed with other signals.
Problem: when I output the 32 bits as one like this:
writel( buffer, (u32)gpio_base_addr+GPIO0_OR );
(where buffer is of u32 type)

One thing which looks strange to me. Are you sure you can use 32 GPIO pins?

Ain't possible, I know.

Quoted 3 lines

This is impossible because 405GP has only 23, and I doubt you can use even this number, because some of the pins are shared with IRQ and PerCSx lines.

You can set all 23 GPIO's as output. I actually only use 17 of them (2 x 8bit position and 1 on/off).
Now I've taken a deeper look at the problem, it might be caused by the EP405 itself. My DAC settles at 70ns.
I still have to check if the signals are outputted all at the same time. Should go and fetch a logic analyzer somewhere ;)

--
Toni Van Remortel
Wetenschappelijk Medewerker - D-science lab
Real time Linux for embedded systems: http://linemb.d-sciencelab.com

              Tel: +32 3 205 61 72 - Fax: +32 3 205 61 95
                      E-mail: [email protected]

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/