| View previous topic :: View next topic |
| Author |
Message |
Mario
Guest
|
 Posted: Tue Nov 18, 2008 10:46 pm Post subject: Overclocking a Freescale 9S12DP256C MCU |
 |
|
Hi,
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
Thanks |
|
| |
|
 Back to top |
Vladimir Vassilevsky
Guest
|
| Posted: Wed Nov 19, 2008 5:05 am Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Mario wrote:
| Quote: | Hi,
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
|
It appears that modern HCS12 can be overclocked by as much 3 x rated
clock. The clock is configured by the PLL register settings.
Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com |
|
| |
|
| Back to top |
Oliver Betz
Guest
|
| Posted: Wed Nov 19, 2008 9:58 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Vladimir Vassilevsky wrote:
| Quote: | I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
It appears that modern HCS12 can be overclocked by as much 3 x rated
clock. The clock is configured by the PLL register settings.
|
although I'm also rather sure that they can run somewhat faster than
24MHz, do you have any sources, measurements, test suites proving your
"x 3" claim?
Oliver
--
Oliver Betz, Munich
despammed.com might be broken, use Reply-To: |
|
| |
|
| Back to top |
Vladimir Vassilevsky
Guest
|
| Posted: Wed Nov 19, 2008 10:52 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Oliver Betz wrote:
| Quote: | Vladimir Vassilevsky wrote:
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
It appears that modern HCS12 can be overclocked by as much 3 x rated
clock. The clock is configured by the PLL register settings.
although I'm also rather sure that they can run somewhat faster than
24MHz, do you have any sources, measurements, test suites proving your
"x 3" claim?
|
Why can't you just try and see for yourself?
I have a device with S12D64. I overclocked it just for curiosity. It
appeared to be operable up to ~70MHz. Probably the silicon can go even
higher; 70MHz is where the internal PLL got stuck. Of course, I didn't
check every fine detail of the operation however it seemed to work normally.
Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com |
|
| |
|
| Back to top |
Oliver Betz
Guest
|
| Posted: Thu Nov 20, 2008 9:07 am Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Vladimir Vassilevsky wrote:
| Quote: | Oliver Betz wrote:
Vladimir Vassilevsky wrote:
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
It appears that modern HCS12 can be overclocked by as much 3 x rated
clock. The clock is configured by the PLL register settings.
although I'm also rather sure that they can run somewhat faster than
24MHz, do you have any sources, measurements, test suites proving your
"x 3" claim?
Why can't you just try and see for yourself?
|
I _did_ overclock S12 devices when I checked the PLL, but made no
thorough tests whether I can _safely use_ the higher frequency. Just
wanted to know whether you have better information.
| Quote: | I have a device with S12D64. I overclocked it just for curiosity. It
appeared to be operable up to ~70MHz. Probably the silicon can go even
higher; 70MHz is where the internal PLL got stuck. Of course, I didn't
check every fine detail of the operation however it seemed to work normally.
|
This "test" doesn't IMNSHO justify your statement in your previous
posting. You shouldn't state "It appears that modern HCS12 can be
overclocked by as much 3 x" if you didn't reasonable tests.
Oliver
--
Oliver Betz, Munich
despammed.com might be broken, use Reply-To: |
|
| |
|
| Back to top |
Paul Keinanen
Guest
|
| Posted: Thu Nov 20, 2008 4:31 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
On Tue, 18 Nov 2008 14:46:18 -0800 (PST), Mario <mallivy@yahoo.com>
wrote:
| Quote: | I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
|
Why do you want to do this ?
Does your software consume so much resources that the CPU is not
capable of handling it at the nominal maximum clock frequency ?
While overclocking might work in a one off (such as a student)
project, you are not going to be able to use this approach in real
world professional design jobs. In those situations, you either have
to improve your software or use a more efficient processor.
Chip manufacturers have to be conservative when specifying the maximum
clock frequency so that the chip will work with any operating voltage
within the specification and within the whole temperature range for
any chip produced at various fabs with slightly different processing.
Typically the limiting factor will be the minimum supply voltage and
the maximum ambient temperature. Running the chip at the maximum
specified operational voltage will typically allow running at a higher
clock frequency. However, this will typically increase the power
dissipation proportional to the clock frequency and the supply voltage
squared, so the junction temperature will also increase.
Apart for the direct meltdown, the increase of the junction
temperature will decrease the MTBF (Mean Time Before Failure) quite
rapidly. To avoid this, you have to limit the maximum allowed ambient
temperature or increase the cooling by a large heatsink or forced air
cooling.
If these are acceptable limitations and especially if you are able to
select the best chips from different production lots, you certainly
would be able to overclock quite a lot, but you would typically not be
able to design a commercially successful product with these
limitations.
Paul |
|
| |
|
| Back to top |
Didi
Guest
|
| Posted: Thu Nov 20, 2008 4:55 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
On Nov 19, 12:46 am, Mario <mall...@yahoo.com> wrote:
| Quote: | Hi,
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
Thanks
|
The maximum clock frequency specified is at highest temperature
and lowest supply voltage.
Freescale are in that busyness not since yesterday so I would be
very surprised if there is a lot of reserve to be had over
temperature, voltage and of course batch variations.
But if you can sacrifice some range of the above - say, this
is just a prototype you want to have running faster and not
put in production - just try to raise the frequency until it stops
working, then settle for say 90 or 95% that and there you are.
Didi
------------------------------------------------------
Dimiter Popoff Transgalactic Instruments
http://www.tgi-sci.com
------------------------------------------------------
http://www.flickr.com/photos/didi_tgi/sets/72157600228621276/ |
|
| |
|
| Back to top |
jw
Guest
|
| Posted: Thu Nov 20, 2008 6:45 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
On Nov 18, 5:46 pm, Mario <mall...@yahoo.com> wrote:
| Quote: | Hi,
I'm currently using an Adapt9S12DP256 module for my final year
project. I'd like to know if it's possible to overclock it and if it
is, what are the details in doing so?
Thanks
|
Yes it can be. But one thing to watch would the A/D Convertors. Once
you start overclocking them, they lose linearity.
And as others have asked, why? Do you need more speed? Do you have
tasks not completing within their allowed time? If not, there is no
real reason to crank up the clock.
Probably not an issue, but in real world EMI/EMC is an issue, and
higher clock speeds cause more problems.
JW |
|
| |
|
| Back to top |
Vladimir Vassilevsky
Guest
|
| Posted: Thu Nov 20, 2008 11:46 pm Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Didi wrote:
| Quote: | The maximum clock frequency specified is at highest temperature
and lowest supply voltage.
Freescale are in that busyness not since yesterday so I would be
very surprised if there is a lot of reserve to be had over
temperature, voltage and of course batch variations.
|
There are also non-technical reasons for the vendors to under rate their
chips: marketing, legacy compatibility, export regulations.
Motorola is one of the oldest players and this is why modern HCS12s have
that much of potential reserve (3 times of speed). The techology
improved many times since HCS12 was introduced; there is no advantage in
the manufacturing of the initial slow variant.
| Quote: | But if you can sacrifice some range of the above - say, this
is just a prototype you want to have running faster and not
put in production - just try to raise the frequency until it stops
working, then settle for say 90 or 95% that and there you are.
|
Fie, how boring.
First of all, overclocking is sport. It is fun :-)
Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com |
|
| |
|
| Back to top |
Oliver Betz
Guest
|
| Posted: Fri Nov 21, 2008 9:17 am Post subject: Re: Overclocking a Freescale 9S12DP256C MCU |
|
|
Vladimir Vassilevsky wrote:
| Quote: | Didi wrote:
The maximum clock frequency specified is at highest temperature
and lowest supply voltage.
Freescale are in that busyness not since yesterday so I would be
very surprised if there is a lot of reserve to be had over
temperature, voltage and of course batch variations.
There are also non-technical reasons for the vendors to under rate their
chips: marketing, legacy compatibility, export regulations.
Motorola is one of the oldest players and this is why modern HCS12s have
that much of potential reserve (3 times of speed). The techology
improved many times since HCS12 was introduced; there is no advantage in
|
tell me more - what exactly did they change?
Oliver
--
Oliver Betz, Munich
despammed.com might be broken, use Reply-To: |
|
| |
|
| Back to top |
|
