S12 / MagniV Microcontrollers中的主题 Speed issues with 56F8013

Speed issues with 56F8013

liquidsquid — Mon, 16 Jul 2007 02:32:53 GMT

Hi all,
    I finally have CodeWarrior up and running with my obscure printer port... I am evaluating the 56F8013EVM for a home project I hope to turn into a product, and I am seemingly having some difficulties getting any performace out of the core.
    I have set up a processor expert project only turing on the bare essentials; The Cpu where I have the PLL clock set at 32MHz and using the On-chip relaxation oscillator, and two of the LEDs since I am using the EVM board for this test.
    I then simply call LED1_Toggle() in the for(; block of code in the main User Modules secion and observe the output on the pin.
    I was hoping if the core was running at 32MHz to see an output between 4-8MHz, maybe faster. Instead I see an output of 1.6MHz on a bare bones project, and a whopping 276kHz with the A/D module turned on in continuous mode.

First off, with the JTAG cable attached, does it slow things down when running?

I get no PLL error interrupts, however when I poll the locked flag of the PLLSR, it always says unlocked, though I am unsure how this could pass the initialization stage if so as the CPU initialization wont exit unless a lock ocurs.

Currently this project is pretty darned slim, so I cannot imagine much being wrong at this point, other than the device's PLL not cooperating or ProcessorExpert generating a crummy piece of code someplace which causes the PLL to unlock.

Thanks for any insight if anyone has run into this one before.

And quickly: Can I expect to be able to get a 32kHz sample rate on two channels at once? Even better, can I expect 32kHz sample rate staggered on four? I have been unable to confirm this on the demo board thanks to this clocking issue. The datasheet says so.

-Mark

Re: Speed issues with 56F8013

liquidsquid — Tue, 17 Jul 2007 07:12:47 GMT

Ok, it was the debugger slowing things up. I was having trouble getting the pod to run stand-alone to see any signalling, but once I did, the device truly flies. More then enough HP for what I need.

Now on to the next thing:
Getting the Beans in CodeWarrior setup to create the 32kHz sampling rate all interrupt driven. I want to write my code as an event-driven interface. However I am really struggling with these timer modules and the PWM, which I cannot seem to get enabled to give me a pulse train on a single output. I have some things to try still, but this is tough to do between dinner, bath for Joe, and bed...

This is certainly not the HCS08, which I have a lot of experience on. (Using the BDM, that is why the speed confusion).

BTW is there a more active forum sponsored or otherwise for newb questions like this? This one seems pretty dead compared to the old Yahoo groups.

-Mark

Re: Speed issues with 56F8013

liquidsquid — Tue, 17 Jul 2007 09:54:35 GMT

Ok, turns out it runs at full speed even in debug. Some setting somewhere during the process of my testing prevented the PLL from coming up, among other things. I created a spanking-new project to test just the PWM, and it fired right up. Then I imported my older project's beans into it, and it ran fine.

Gotta love problems like that when you are firing up something for the first time.