Forums

H8S 2674R All Module Clock Stop Mode??

Started by Dave Madsen March 30, 2005
We are using a Renesas H8S 2674R processor and cannot get the
specified current consumption numbers when entering the All Module
Clock Stop mode. Has anyone successfully used this mode in this family
of H8S processors?
On Thu, 31 Mar 2005 00:46:06 GMT, in article
     <424b4783.34581515@news.sisna.com> dmadsen@campbellsci.com
     "Dave Madsen" wrote:

>We are using a Renesas H8S 2674R processor and cannot get the >specified current consumption numbers when entering the All Module >Clock Stop mode. Has anyone successfully used this mode in this family >of H8S processors?
Due to other tasks [real life] I forgot to follow up to this before. You are of course remembering that most standby modes retain the I/O port status so a HIGH output causes NO reduction in current consumption for that pin. I suggest you check what the I/O pins are doing when you enter that standby mode. This is described in the manual on Power Down modes and is the first item in the Usage notes (see in conjunction with Table 22.1 Operating Modes for what each module actually does). -- Paul Carpenter | paul@pcserviceselectronics.co.uk <http://www.pcserviceselectronics.co.uk/> PC Services <http://www.gnuh8.org.uk/> GNU H8 & mailing list info <http://www.badweb.org.uk/> For those web sites you hate
Thanks for the reply.
I built up a board with only the processor and a FLASH chip, and I
sent the device to all module stop immediately after reset (i.e. all
pins were inputs). This still did not give the specified current
value.
I then configured all I/O pins as outputs and drove them low, and
modified the bare board to tie all input only pins low.
I then did extensive testing with all module clock stop vs. software
standby, setting and not setting the OPE bit in the sbycr register.
With the OPE bit set all output pins are maintained when going to
software standby as they are in all module clock stop. 
Our results showed that in software standby we got the expected
(specified) current consumption (<100uA for the whole board). When
using all module stop we got to 18mA (the spec is 125uA).
I have tried everything I can think of. I even changed the PLL from 2X
mode to 1X mode and enabled the divider to divide by 32. This dropped
the current down to < 2mA. This proved that indeed the current is
being consumed in the processor. The problem with this mode is wake up
is much slower because things are 64x slower, until the PLL can be set
to 2X and the divide to 1 (this takes about 50 usec). Also, I need to
feed the clock to an external device and it must be the non-divided
version of the clock.
The manual seems simple enough. No DMA or DTC or pending internal
interrupts then set the module stop bits for all modules and the all
module clock stop bit then execute the sleep instruction.
Have you or do you know if anyone has seen the specified current
values for the all module clock stop mode?? If so how? 

Thanks,
Dave

On Mon, 04 Apr 2005 22:15:41 +0100 (BST), paul$@pcserv.demon.co.uk
(Paul Carpenter) wrote:

>On Thu, 31 Mar 2005 00:46:06 GMT, in article > <424b4783.34581515@news.sisna.com> dmadsen@campbellsci.com > "Dave Madsen" wrote: > >>We are using a Renesas H8S 2674R processor and cannot get the >>specified current consumption numbers when entering the All Module >>Clock Stop mode. Has anyone successfully used this mode in this family >>of H8S processors? > >Due to other tasks [real life] I forgot to follow up to this before. > >You are of course remembering that most standby modes retain the I/O port >status so a HIGH output causes NO reduction in current consumption for that >pin. I suggest you check what the I/O pins are doing when you enter that >standby mode. > >This is described in the manual on Power Down modes and is the first item >in the Usage notes (see in conjunction with Table 22.1 Operating Modes >for what each module actually does). > >-- >Paul Carpenter | paul@pcserviceselectronics.co.uk ><http://www.pcserviceselectronics.co.uk/> PC Services ><http://www.gnuh8.org.uk/> GNU H8 & mailing list info ><http://www.badweb.org.uk/> For those web sites you hate >
On Monday, in article <425aca87.17027203@news.sisna.com>
     dmadsen@campbellsci.com "Dave Madsen" wrote:

>Thanks for the reply. >I built up a board with only the processor and a FLASH chip, and I >sent the device to all module stop immediately after reset (i.e. all >pins were inputs). This still did not give the specified current >value.
What else is connected (pull-ups, serial converters etc..)? Just a thought, because you must be getting the code in there somehow.
>I then configured all I/O pins as outputs and drove them low, and >modified the bare board to tie all input only pins low. >I then did extensive testing with all module clock stop vs. software >standby, setting and not setting the OPE bit in the sbycr register. >With the OPE bit set all output pins are maintained when going to >software standby as they are in all module clock stop. >Our results showed that in software standby we got the expected >(specified) current consumption (<100uA for the whole board). When >using all module stop we got to 18mA (the spec is 125uA). >I have tried everything I can think of. I even changed the PLL from 2X >mode to 1X mode and enabled the divider to divide by 32. This dropped >the current down to < 2mA. This proved that indeed the current is >being consumed in the processor. The problem with this mode is wake up >is much slower because things are 64x slower, until the PLL can be set >to 2X and the divide to 1 (this takes about 50 usec). Also, I need to >feed the clock to an external device and it must be the non-divided >version of the clock.
Done all the things I would expect to test. However I note that if you are having the clock output going, that circuit will draw current and depending on what it is driving could effect the current drawn. Have you tried this with the clock output not loaded or disabled? Have you tried this without the external flash?
>The manual seems simple enough. No DMA or DTC or pending internal >interrupts then set the module stop bits for all modules and the all >module clock stop bit then execute the sleep instruction.
I assume you are stopping the 8 bit timer and I assume you are not using the WDT as well.
>Have you or do you know if anyone has seen the specified current >values for the all module clock stop mode?? If so how?
I personally have not used that particular device but various H8 devices. I would have thought others here may have had something to offer on the subject. You might also get some assistance from others on the GNUH8 mailing list (see signature block). I would also suggest talking to your distributor to get from Renesas the test circuit they used to get the figures which in my datasheet have a note "The values are for reference" against the curent consumption figures. This seems a bit of specmanship to me which would need ratifying by Renesas to get what they mean by that.
>Thanks, >Dave > >On Mon, 04 Apr 2005 22:15:41 +0100 (BST), paul$@pcserv.demon.co.uk >(Paul Carpenter) wrote: > >>On Thu, 31 Mar 2005 00:46:06 GMT, in article >> <424b4783.34581515@news.sisna.com> dmadsen@campbellsci.com >> "Dave Madsen" wrote: >> >>>We are using a Renesas H8S 2674R processor and cannot get the >>>specified current consumption numbers when entering the All Module >>>Clock Stop mode. Has anyone successfully used this mode in this family >>>of H8S processors? >> >>Due to other tasks [real life] I forgot to follow up to this before. >> >>You are of course remembering that most standby modes retain the I/O port >>status so a HIGH output causes NO reduction in current consumption for that >>pin. I suggest you check what the I/O pins are doing when you enter that >>standby mode. >> >>This is described in the manual on Power Down modes and is the first item >>in the Usage notes (see in conjunction with Table 22.1 Operating Modes >>for what each module actually does).
-- Paul Carpenter | paul@pcserviceselectronics.co.uk <http://www.pcserviceselectronics.co.uk/> PC Services <http://www.gnuh8.org.uk/> GNU H8 & mailing list info <http://www.badweb.org.uk/> For those web sites you hate
On Wed, 13 Apr 2005 01:51:16 +0100 (BST), paul$@pcserv.demon.co.uk
(Paul Carpenter) wrote:

>On Monday, in article <425aca87.17027203@news.sisna.com> > dmadsen@campbellsci.com "Dave Madsen" wrote: > >>Thanks for the reply. >>I built up a board with only the processor and a FLASH chip, and I >>sent the device to all module stop immediately after reset (i.e. all >>pins were inputs). This still did not give the specified current >>value. > >What else is connected (pull-ups, serial converters etc..)? >Just a thought, because you must be getting the code in there somehow. >
The FLASH chip is in a socket, so we program it out of circuit. We do run Rx and Tx to another board with RS232 drivers, but these are disconnected and disabled for the current measurements. No pull-ups on the bare board.
>>I then configured all I/O pins as outputs and drove them low, and >>modified the bare board to tie all input only pins low. >>I then did extensive testing with all module clock stop vs. software >>standby, setting and not setting the OPE bit in the sbycr register. >>With the OPE bit set all output pins are maintained when going to >>software standby as they are in all module clock stop. >>Our results showed that in software standby we got the expected >>(specified) current consumption (<100uA for the whole board). When >>using all module stop we got to 18mA (the spec is 125uA). >>I have tried everything I can think of. I even changed the PLL from 2X >>mode to 1X mode and enabled the divider to divide by 32. This dropped >>the current down to < 2mA. This proved that indeed the current is >>being consumed in the processor. The problem with this mode is wake up >>is much slower because things are 64x slower, until the PLL can be set >>to 2X and the divide to 1 (this takes about 50 usec). Also, I need to >>feed the clock to an external device and it must be the non-divided >>version of the clock. > >Done all the things I would expect to test. >However I note that if you are having the clock output going, that circuit >will draw current and depending on what it is driving could effect the >current drawn. > >Have you tried this with the clock output not loaded or disabled? >Have you tried this without the external flash? >
I have set and cleared the PSTOP bit, which disables and enables the phi clock from being fed out. This makes very little difference on the current consumption of the bare board. The fact that the current is very low with software standby (regardless of the setting of OPE) and not with all module clock stop seems to eliminate the FLASH chip.
>>The manual seems simple enough. No DMA or DTC or pending internal >>interrupts then set the module stop bits for all modules and the all >>module clock stop bit then execute the sleep instruction. > >I assume you are stopping the 8 bit timer and I assume you are not using the >WDT as well. >
Correct, all of the peripheral modules are unused.
>>Have you or do you know if anyone has seen the specified current >>values for the all module clock stop mode?? If so how? > >I personally have not used that particular device but various H8 devices. >I would have thought others here may have had something to offer on the >subject. > >You might also get some assistance from others on the GNUH8 mailing list >(see signature block). I would also suggest talking to your distributor >to get from Renesas the test circuit they used to get the figures >which in my datasheet have a note "The values are for reference" against >the curent consumption figures. This seems a bit of specmanship to me >which would need ratifying by Renesas to get what they mean by that.
Thanks for your suggestions. We are working with our Renesas Rep and he has been in contact with Renesas, it is slow getting answers from them, but they say that the numbers are real.
> >>Thanks, >>Dave >> >>On Mon, 04 Apr 2005 22:15:41 +0100 (BST), paul$@pcserv.demon.co.uk >>(Paul Carpenter) wrote: >> >>>On Thu, 31 Mar 2005 00:46:06 GMT, in article >>> <424b4783.34581515@news.sisna.com> dmadsen@campbellsci.com >>> "Dave Madsen" wrote: >>> >>>>We are using a Renesas H8S 2674R processor and cannot get the >>>>specified current consumption numbers when entering the All Module >>>>Clock Stop mode. Has anyone successfully used this mode in this family >>>>of H8S processors? >>> >>>Due to other tasks [real life] I forgot to follow up to this before. >>> >>>You are of course remembering that most standby modes retain the I/O port >>>status so a HIGH output causes NO reduction in current consumption for that >>>pin. I suggest you check what the I/O pins are doing when you enter that >>>standby mode. >>> >>>This is described in the manual on Power Down modes and is the first item >>>in the Usage notes (see in conjunction with Table 22.1 Operating Modes >>>for what each module actually does). > >-- >Paul Carpenter | paul@pcserviceselectronics.co.uk ><http://www.pcserviceselectronics.co.uk/> PC Services ><http://www.gnuh8.org.uk/> GNU H8 & mailing list info ><http://www.badweb.org.uk/> For those web sites you hate >