DSPRelated.com
Forums

DSP misfunction when short power-down

Started by the_believer_1976 September 22, 2003
Hi everyone,

I am experiencing problems with 5407 when the power supply is cut for
some miliseconds. When returning, the program is properly boot loaded
from the external memory, but executed on a wrong way.

With a longer cut of supply this does not happen.

Does anyone know about it?
Thanks very much,
Believer




Believer-

> I am experiencing problems with 5407 when the power supply is cut for
> some miliseconds. When returning, the program is properly boot loaded
> from the external memory, but executed on a wrong way.
>
> With a longer cut of supply this does not happen.

Did you figure this out yet? Was there something that got caught in partial
reset,
or a strange state? I'm curious because it sounds like something we might run
into
some day with our embedded systems.

-Jeff



Hi,
Believer ,
As far as the data sheets speaks , to reset the 5409 we have to give a
short low pulse of duration (4H+4) H =
0.5Tc0 Tc0 = 40nSec ie total 120 nsec. so few millisec of power cut is
reseting it. May be the memory is not getting corrupted ( please check the
datasheet of the memory) the refreshing cycle time.
I think this may be causing a problem, the PC and other initialization
is getting changed. Please check the register contents before and after
millisec power cut.

I am suggesting a cause. Eager to know the real cause. Please let us know
what is happening when there is a long power down.
Regds.
Soumit.

----- Original Message -----
From: <>
To: <>
Sent: Friday, September 26, 2003 12:57 AM
Subject: [c54x] Digest Number 634 > _____________________________________
> Note: If you do a simple "reply" with your email client, only the author
of this message will receive your answer. You need to do a "reply all" if
you want your answer to be distributed to the entire group.
>
> _____________________________________
> About this discussion group:
>
> To Join: Send an email to
>
> To Post: Send an email to
>
> To Leave: Send an email to
>
> Archives: http://www.yahoogroups.com/group/c54x
>
> Other Groups: http://www.dsprelated.com >
> There are 7 messages in this issue.
>
> Topics in this digest:
>
> 1. Re: DSP misfunction when short power-down
> From: Jeff Brower <>
> 2. Simulating HPI
> From: "juanchopolo99" <>
> 3. Re: 5402/DSK5416 Basics
> From: "Brian C. Lane" <>
> 4. Delay slot of FCALLD
> From: "kt_rhee" <>
> 5. Parallel Bootloading the 5402
> From: "Yaser Namazi" <>
> 6. problem of MCBSP on DSK
> From: "afatnonoafat" <>
> 7. Re: Parallel Bootloading the 5402
> From: mirzaalivandi mohammad < > ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 1
> Date: Wed, 24 Sep 2003 12:30:15 -0500
> From: Jeff Brower <>
> Subject: Re: DSP misfunction when short power-down
>
> Believer-
>
> > I am experiencing problems with 5407 when the power supply is cut for
> > some miliseconds. When returning, the program is properly boot loaded
> > from the external memory, but executed on a wrong way.
> >
> > With a longer cut of supply this does not happen.
>
> Did you figure this out yet? Was there something that got caught in
partial reset,
> or a strange state? I'm curious because it sounds like something we might
run into
> some day with our embedded systems.
>
> -Jeff > ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 2
> Date: Wed, 24 Sep 2003 18:36:25 -0000
> From: "juanchopolo99" <>
> Subject: Simulating HPI
>
> Hello
>
> I tried to simulate the HPI (Host Port Interface) with CCS 2.2 for
> C5000. I follow the guidelines for make a file that write the HPIA
> register and then write a data in the address point by HPIA, follow I
> connect with port connect, but when run the program always CCS tell
> me that I have a syntax error. I tried many formats but always the
> same.
> You could give me a real example?.
>
> Thanks.
>
> Juan Pablo Amaya >
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 3
> Date: Wed, 24 Sep 2003 11:53:55 -0700
> From: "Brian C. Lane" <>
> Subject: Re: 5402/DSK5416 Basics
>
> reebil wrote:
>
> > Hi,
> > 1. I would like to drive a standard 20x2 LCD in 4 bit mode using the
> > DSK (obviously as part of a greater project). I want to use the
> > expansion peripheral interface for this purpose, using some of the
> > MCBSP pins. As a test program, I am trying to turn the LEDs on and
> > off without using the board.h functions brd_led_*(). As I understand
> > it, I should be able to access the LEDs using the CNTL1 register.
> >
> > How can I set and clear each LED individually? That is, what C
> > function or ASM mnemonic can I use to achieve
> > "SETBIT" CNTL1, USERLED2 or
> > "CLRBIT" CNTL1, USERLED2
> >
>
> I don't use the DSK so I don't know what API calls to use. But to
> control the McBSP lines when the McBSP is disabled (of course) you use
> the bits in the PCR register associated with the McBSP.
>
> > 2. Can someone please explain what "wait states" are?
> >
>
> Wait states are delays that are inserted when accessing external memory
> like flash or external ram. The DSP runs so fast that no external flash
> can keep up with it running at 100MHz, so you setup the wait-state
> register to slow down external memory requests so that they don't exceed
> the speed rating of the flash (or other external memory mapped device).
>
> > I wouldn't dare ask these questions in comp.dsp, and I hope that you
> > guys will be tolerant of them. If these questions are answered in any
> > FAQ or other reference, please point me in the right direction.
> >
>
> Oh My! Be tolerant of questions!? Never! What kind of place do you think
> this is? Do you think we all just sit here waiting to jump at your beck
> and call? We certainly have better issues to discuss than your
> insignificant little problems; we have code to write, products to
> design, people to flame! Why I ought to just ...
>
> Oh, sorry, I forgot that this wasn't comp.dsp <G>.
>
> Welcome to the list, you'll find things a bit more relaxed here.
>
> Brian >
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 4
> Date: Thu, 25 Sep 2003 10:32:11 -0000
> From: "kt_rhee" <>
> Subject: Delay slot of FCALLD
>
> Hi,
>
> As you know, there can be one 2-word instruction or two one-word
> instructions in delay slot of "calld" and "fcalld." If I want to
> access the stack in the delay slot, I should consider the place
> of "PC+2." For example,
>
> calld _foo ; call the function foo
> stl A, *sp(arg+1)
> ld *sp(tmp+1),A ; delay slot
>
> But, in the case of "fcalld," I should consider XPC as well as PC+2.
> I found that the SP is different for each delay slot. So, I should
> write the above code like below. (in case of FAR mode)
>
> fcalld _foo ; call the function foo with FAR mode
> stl A, *sp(arg+1)
> ld *sp(tmp+2),A ; delay slot (should be *sp(tmp+2))
>
> Now, I'm confusing with the case that 2-word instruction follows the
> fcalld. An instruction like "mvdk", I should use it as
>
> fcalld _foo
> mvdk *(ar0), *sp(arg+2)
>
> Unfortunately, I should use like below in case of "st,"
>
> fcalld _foo
> st #145, *sp(arg+1) > I hope to know the kind of rule I should know to use the delay slots
> properly in FAR mode. Is there any rule about it?
>
> Best Regards,
> KT Lee. > ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 5
> Date: Thu, 25 Sep 2003 10:59:27 -0000
> From: "Yaser Namazi" <>
> Subject: Parallel Bootloading the 5402
>
> Hi
>
> I have a little problem in bootloading the 5402 in the parallel
> mode. I'm using a 8-bit wide flash memory for bootloading the DSP
> (5402), but I cannot get any result; could anyone help me with an
> example of a properly programmed flash memory cintents for
> bootloading in parallel mode.
>
> Many thanks,
> Yaser. >
>
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 6
> Date: Thu, 25 Sep 2003 14:04:48 -0000
> From: "afatnonoafat" <>
> Subject: problem of MCBSP on DSK
>
> hello :
> I have some problem about using MCBSP ON DSK.
>
> i've tried some MCBSP examples:
> DAA(tel) and CODEC(audio in out),
> and sure that MCBSP0/1 are all at work. > but now i got some problem:
>
> i've tried to use CNTL2(CPLD register) register
> to arrange the mcbsp 0/1 to
> the DSK expansion peripheral interface,
> but i can't detect any signal.
>
> is there only CNTL2(CPLD register) register i need to care ?
> should i setup any other registers?
>
> thanks for help !! >
> ________________________________________________________________________
> ________________________________________________________________________
>
> Message: 7
> Date: Thu, 25 Sep 2003 05:47:44 -0700 (PDT)
> From: mirzaalivandi mohammad <>
> Subject: Re: Parallel Bootloading the 5402
>
> dear friend
> hi
> as you know @ start up MP/MC flag is samled with it's
> hardware pin ,& it is so important to define your boot
> loader algorithm plz check it first.then check
> sequence of your booting,for this point that parallel
> bootloading is not the first method in the sequential
> check.(for example HPI method is the first).
> regards
> mohammad alivandi > --- Yaser Namazi <> wrote:
> > Hi
> >
> > I have a little problem in bootloading the 5402 in
> > the parallel
> > mode. I'm using a 8-bit wide flash memory for
> > bootloading the DSP
> > (5402), but I cannot get any result; could anyone
> > help me with an
> > example of a properly programmed flash memory
> > cintents for
> > bootloading in parallel mode.
> >
> > Many thanks,
> > Yaser.
> >
> >
> >
> > _____________________________________
> > Note: If you do a simple "reply" with your email
> > client, only the author of this message will receive
> > your answer. You need to do a "reply all" if you
> > want your answer to be distributed to the entire
> > group.
> >
> > _____________________________________
> > About this discussion group:
> >
> > To Join: Send an email to
> >
> >
> > To Post: Send an email to
> >
> > To Leave: Send an email to
> >
> >
> > Archives: http://www.yahoogroups.com/group/c54x
> >
> > Other Groups: http://www.dsprelated.com
> >
> >
> > ">http://docs.yahoo.com/info/terms/
> >
> > __________________________________ >
> ________________________________________________________________________
> ________________________________________________________________________ >
> ">http://docs.yahoo.com/info/terms/



DISCLAIMER: The information contained in this message is intended only and
solely for the addressed individual or entity indicated in this message and for
the exclusive use of the said addressed individual or entity indicated in this
message (or responsible for delivery of the message to such person) and may
contain legally privileged and confidential information belonging to Tata
Consultancy Services. It must not be printed, read, copied, disclosed,
forwarded, distributed or used (in whatsoever manner) by any person other than
the addressee. Unauthorized use, disclosure or copying is strictly prohibited
and may constitute unlawful act and can possibly attract legal action, civil
and/or criminal. The contents of this message need not necessarily reflect or
endorse the views of Tata Consultancy Services on any subject matter. Any action
taken or omitted to be taken based on this message is entirely at your risk and
neither the originator of this message nor Tata Consultancy Services takes any
responsibility or liability towards the same. Opinions, conclusions and any
other information contained in this message that do not relate to the official
business of Tata Consultancy Services shall be understood as neither given nor
endorsed by Tata Consultancy Services or any affiliate of Tata Consultancy
Services. If you have received this message in error, you should destroy this
message and may please notify the sender by e-mail. Thank you.