ColdFire/68K Microcontrollers and ProcessorsのトピックRe: Can a noise hang a ColdFire forever...?

Can a noise hang a ColdFire forever...?

Ricardo_RauppV — Fri, 07 Mar 2008 07:23:39 GMT

Hi people

I´m using a 5213 target and it hangs up after a turn off a close 12V/35mA relay wich doesn´t have freewheling diode, that is, generating a noise.

I´m using watchdog and others interrupts in order to make it resets, but "sometimes" it is not enought, the cpu

simply stops and only by a hw reset it turns to work back.

I´m not concerned about the reset issue,but about the hang forever detail....reset after noise is acceptable e expectaed, but hang up ...never... ??

My question:

- Is ColdFire able to "allways" reset after a disturb (conducted or induced)?or should I use a external watchdog?Some special care to assure reset allways?

Thanks

Ricardo Raupp

Re: Can a noise hang a ColdFire forever...?

murat — Thu, 03 Apr 2008 14:43:36 GMT

.Hi Ricardo,

I have a board design with MCF52235 and encountered the same problem.My processor hangs up when we conduct suge/burst and ESD tests on the board.(You can see the same effect by touching to BKPT pin with a charged material) But after a long period of searching the problem I realized that a spike in BKPT pin causes the processor to enter in halt mode and the only way to exit this mode is to reset the processor.I think this is a shortcoming that is found on all Coldfire family.and designer of this processor should take care of this problem.

To overcome this problem you should have a very good protection against surge/bursy and ESD..

Best regards...

Murat Kurus

ENTES ELECTRONIC R&D

Re: Can a noise hang a ColdFire forever...?

sjmelnikoff — Mon, 07 Apr 2008 15:38:41 GMT

It would also be a good idea to use an external watchdog chip, such as the Texas TPS3813. This will reset the processor if it hangs, and also acts as a voltage monitor.

Steve.

Re: Can a noise hang a ColdFire forever...?

Ricardo_RauppV — Mon, 28 Apr 2008 03:50:17 GMT

Hi guys
Sorry for delay to thanks your help..but I was in troubles ...
Well...
I connected the bdm pins direct to power connections in order to dosable thses
pins to be a noise conductor or interference.
I coould see a significant improvement..but not enough to avoid ColdFire
hangup.
The only way to bring ColdFire back was through reset pin, wich lead me to
conclude the chip was not latched up...
After I have made lot of tests, I decide to do 3 things
- Add a hw watchdog in any CF design of mine
- Install a metalic cover shield to ColdFire mcu for little more hard
environment
- Gaveup to believe internal mcu controls can save the chip against noise

Thanks all of you !!!
Ricardo Raupp

Re: Can a noise hang a ColdFire forever...?

DrSeuss — Fri, 06 Feb 2009 05:54:13 GMT

ersonal; mso-style-noshow:yes; mso-ansi-font-size:10.0pt; mso-bidi-font-size:10.0pt; font-family:Arial; mso-ascii-font-family:Arial; mso-hansi-font-family:Arial; mso-bidi-font-family:Arial; color:navy;} @page Section1 {size:8.5in 11.0in; margin:1.0in 1.25in 1.0in 1.25in; mso-header-margin:.5in; mso-footer-margin:.5in; mso-paper-source:0;} div.Section1 {pageection1;} /* List Definitions */ @list l0 {mso-list-id:1264613104; mso-list-type:hybrid; mso-list-template-ids:2049337404 67698705 67698713 67698715 67698703 67698713 67698715 67698703 67698713 67698715;} @list l0:level1 {mso-level-text:"%1\)"; mso-level-tab-stop:.5in; mso-level-number-position:left; text-indent:-.25in;} ol {margin-bottom:0in;} ul {margin-bottom:0in;} -->

I can think of a few things that might cause the CPU to appear to lock up. (ESD is not one).

1) Make sure that the Bus Timeout timer is set correctly (should be from reset).

2) Make sure that there is a service routine that either resets the devices or pulls the stack frame from the stack.

3) Turn off BDM mode. This will eliminate the BKPT as an issue.

4) Make Very sure that the test pin is grounded! This is the only thing that will kill the CPU.

5) Check for good strong up/downs on CLKMOD pins.

6) Implement a watch dog timer.

7) Fill all unused memory with a trapf instruction and of course have the vector and recovery routine. This will regain control if the CPU fast than an illegal op code.

Message Edited by DrSeuss on 2009-02-05 09:54 PM