Searching \ for '[OT] PIC Oomph' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'PIC Oomph'.

Exact match. Not showing close matches.
PICList Thread
'[OT] PIC Oomph'
2000\03\30@200357 by David E. Olson

flavicon
face
My next project is to build data collection and monitoring hardware for a
dynamometer. We have a dyno that provides rudimentary data on HP and
torque - not enough for remapping our engine management system (Porsche
racing if you care).

Since this isn't far off of engine management systems, which are more
commonly controlled by Motorola chips, I'm wondering if a PIC(s) could hack
it. I heard once before (maybe on this list) that the Harvard architecture
may not be conducive to a high speed, real time, multiple decision making
process. Although the 18-series PIC have pretty respectable stats. Clearly,
there will be a PC involved in the analysis, so I'm not looking to build the
complete system with PICs. Just curious on your interpretation of a PIC
stepping up to the plate. I'd rather not jump over to Motorola - too much
invested in PICs.

My other options can be data collection boards in a PC. My gut says I can
capture and do more with purposed PICs rather than generic acquisition and
more code on the PC.

I am very familiar with the Bosch Motronic systems which are Motorola based.

BTW - gone through all the DIY FI data on their site. Cool stuff if you're
interested in engine management systems.

-DO

2000\03\30@205829 by Andrew Warren

face
flavicon
face
David E. Olson <spam_OUTPICLISTTakeThisOuTspamMITVMA.MIT.EDU> wrote:

> My next project is to build data collection and monitoring
> hardware for a dynamometer. .... I'm wondering if a PIC(s) could
> hack it. I heard once before (maybe on this list) that the Harvard
> architecture may not be conducive to a high speed, real time,
> multiple decision making process.

   David:

   The Harvard architecture neither helps nor hurts; PICs are just
   as capable of performing those tasks as any other micros with
   comparable speed and memory.

> Just curious on your interpretation of a PIC stepping up to the plate.
> I'd rather not jump over to Motorola

   Well, if you're just doing data-acqisition and some simple
   monitoring/analysis -- presumably in the dyno, not in the car --
   then any one of a number of PICs would probably work just fine.

> My other options can be data collection boards in a PC. My gut says
> I can capture and do more with purposed PICs rather than generic
> acquisition and more code on the PC.

   Maybe, maybe not.  I'm not sure what advantage ANY embedded
   controller would have over a PC with decent data-acquisition
   hardware.

> I am very familiar with the Bosch Motronic systems which are
> Motorola based.

   Yeah, but that doesn't have much to do with THIS project, does
   it?

   Maybe I'm misunderstanding... ARE you proposing to incorporate
   the PICs into the car's engine-management system somehow?

   -Andy


=== Andrew Warren - .....fastfwdKILLspamspam@spam@ix.netcom.com
=== Fast Forward Engineering - San Diego, California
=== http://www.geocities.com/SiliconValley/2499

2000\03\30@205838 by David VanHorn

flavicon
face
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

An AVR might be a mid-term step for you.

I did a project lasy year, which had a fair bit going on in several
interrupts, including one at 300k ints/sec, and I still had enough
horsepower left for a background task..

The $200 ice didn't hurt either :)

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.2 for non-commercial use <http://www.pgp.com>

iQA/AwUBOOQu6YFlGDz1l6VWEQJPlwCcCmxLh+z1d9s2m6ru3Z+kRmLhi84AoLB2
6E5nCzkYYqcWdtmA1nToOYGR
=y5D3
-----END PGP SIGNATURE-----

2000\03\30@214047 by Spehro Pefhany

picon face
At 07:59 PM 3/30/00 -0800, you wrote:

>Since this isn't far off of engine management systems, which are more
>commonly controlled by Motorola chips, I'm wondering if a PIC(s) could hack
>it. I heard once before (maybe on this list) that the Harvard architecture
>may not be conducive to a high speed, real time, multiple decision making
>process.

The Motorola chips in question usually have the TPU, essentially a
dedicated I/O co-processor that can do multiple timing and other operations
along side the main processor. It's probably the lack of this "peripheral"
that will hurt more than the choice of whatever core processor (though
the CPU32 is also quite speedy and has a very nice orthogonal instruction
set). There are 16 timer channels and two 16 bit timebases in the 68332 TPU.
Each channel has a pin. (which isn't many out of the 132 or 144 pins).
Some engine controllers are now using even more powerful PowerPC chips.

If you are interested, you can have a look at:

199.104.132.208/brdata/PDFDB/MICROCONTROLLERS/32-BIT/M951447481497col
lateral.pdf

On the plus side, it does have BDM, so an emulator may not be required.

Best regards,

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speffspamKILLspaminterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com
Contributions invited->The AVR-gcc FAQ is at: http://www.bluecollarlinux.com
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

2000\03\30@220219 by David E. Olson

flavicon
face
>     David:
>
>     The Harvard architecture neither helps nor hurts; PICs are just
>     as capable of performing those tasks as any other micros with
>     comparable speed and memory.
>

Which I believe - just thought someone raised the point before...

> > Just curious on your interpretation of a PIC stepping up to the plate.
> > I'd rather not jump over to Motorola
>
>     Well, if you're just doing data-acquisition and some simple
>     monitoring/analysis -- presumably in the dyno, not in the car --
>     then any one of a number of PICs would probably work just fine.

This is dyno-based. Our engine management in the car is handled by a company
called MOTEC. The purpose behind this project is to give us more data to
adjust the various mappings in their system. It's not a question of how much
horsepower and torque - it's when and where. We adjust the ECU for different
tracks, engines and races.

> > My other options can be data collection boards in a PC. My gut says
> > I can capture and do more with purposed PICs rather than generic
> > acquisition and more code on the PC.
>
>     Maybe, maybe not.  I'm not sure what advantage ANY embedded
>     controller would have over a PC with decent data-acquisition
>     hardware.

I sorta agree here. It's also a bit of core competency too. Been doing PICs
for stuff like this and never used PC acquisition hardware.

> > I am very familiar with the Bosch Motronic systems which are
> > Motorola based.
>
>     Yeah, but that doesn't have much to do with THIS project, does
>     it?

OK. Too much useless information...thought it would help frame the issue.

>     Maybe I'm misunderstanding... ARE you proposing to incorporate
>     the PICs into the car's engine-management system somehow?
>

Not right now. It could lead to that if what we're able to do is
complimentary or better than what we have. A lot of this is R&D. This post
is squarely in the "R" side. 1/4 second per lap advantage means a lot when
you're running Daytona.

-DO

2000\03\30@231500 by Andrew Warren

face
flavicon
face
David E. Olson <.....PICLISTKILLspamspam.....MITVMA.MIT.EDU> wrote:

> > I'm not sure what advantage ANY embedded controller would have
> > over a PC with decent data-acquisition hardware.
>
> I sorta agree here. It's also a bit of core competency too. Been
> doing PICs for stuff like this and never used PC acquisition
> hardware.

   David:

   There are enough people running Motec-controlled Porsches that a
   lot of independent (i.e., not affiliated with any teams you'll
   be competing with) shops have acquired a WEALTH of information
   about them.  I know it'll probably be fun to go through the
   whole process yourself, but mightn't it be easier to just hire
   one of those guys to do the mapping for you?  Offhand, I'd
   recommend either Steve Weiner at Rennsport or Bob Norwood.

> 1/4 second per lap advantage means a lot when you're running
> Daytona.

   No offense, but if you can gain a quarter-second through ECU
   mapping, your current map must be SERIOUSLY wrong.

   -Andy


=== Andrew Warren - EraseMEfastfwdspam_OUTspamTakeThisOuTix.netcom.com
=== Fast Forward Engineering - San Diego, California
=== http://www.geocities.com/SiliconValley/2499

2000\03\30@234428 by Plunkett, Dennis

flavicon
face
{Quote hidden}

       I think that your barking up the wrong tree here, just remember that
not all laps are at full speed. The 1/4 second equates to around 3Hp at full
pace during a poll lap or around (Around .35%). Some ECUs are mapped to
efficency, then rempaaed to give power, then remapped to give speed this is
very handy for when the pace car come out.

       Just remember that Daytona is not won by the fastest car!


       Dennis

2000\03\31@010900 by Andrew Warren

face
flavicon
face
Plunkett, Dennis <RemoveMEPICLISTTakeThisOuTspamMITVMA.MIT.EDU> wrote:

> > No offense, but if you can gain a quarter-second [per lap at
> > Daytona] through ECU mapping, your current map must be
> > SERIOUSLY wrong.
>
> I think that your barking up the wrong tree here,....

   I don't think so.

> just remember that not all laps are at full speed.

   True.  And?

> The 1/4 second equates to around 3Hp at full pace during a poll lap
> or around (Around .35%).

   An additional three horsepower (whatever THAT means... 3 more
   horsepower at the power peak?  A power curve that's 3 hp higher
   across the whole band?) will NOT give anyone a quarter-second
   per lap.

   And... Are you also saying that 3 hp is 0.35% of the power that
   a Daytona-spec 911 makes?  3/0.0035 works out to about 850
   horsepower, which is more than a little high.

> Just remember that Daytona is not won by the fastest car!

   I'll try to keep that in mind.

   -Andy


=== Andrew Warren - spamBeGonefastfwdspamBeGonespamix.netcom.com
=== Fast Forward Engineering - San Diego, California
=== http://www.geocities.com/SiliconValley/2499

2000\03\31@094732 by David E. Olson

flavicon
face
>     There are enough people running Motec-controlled Porsches that a
>     lot of independent (i.e., not affiliated with any teams you'll
>     be competing with) shops have acquired a WEALTH of information
>     about them.  I know it'll probably be fun to go through the
>     whole process yourself, but mightn't it be easier to just hire
>     one of those guys to do the mapping for you?  Offhand, I'd
>     recommend either Steve Weiner at Rennsport or Bob Norwood.

We've used Jerry Woods. Our biggest concern is that the expertise is not
usually accessible or available when we need it. We've had schedules pushed
back because we can't get the dyno time or the "experts" to make it happen.
The feasibility of shipping engines around the US starts to get ridiculous
after a while. We're not averse to using the experts - they clearly (at this
time) know more about the system. We'd like to consider taking some control
of our systems.

> > 1/4 second per lap advantage means a lot when you're running
> > Daytona.
>
>     No offense, but if you can gain a quarter-second through ECU
>     mapping, your current map must be SERIOUSLY wrong.

As it can be. No one map is consistent for each engine. Working with a
SERIOUSLY wrong map makes you lose or blow up. And at this time, we have
some incorrect maps for a new batch of engines. We can't get time with the
local expert and were hesitant to go out and test the car. It's frustrating.
Hence, the initial premise of the post - could a PIC(s) do it - maybe not
for this season but, we've got the facilities we're doing the research to
see if we can.

-DO

2000\03\31@100643 by David E. Olson

flavicon
face
Friends,

Let's put this post to rest. All I wanted was some reaction from the crowd
on a PIC (or series of them) to perform real-time analysis for a dyno. I
appreciate the private and public replies. If you want to pursue it further,
let's take it privately. There are a bunch of other factors involved that
don't relate to PICs. Questioning our ability to generate 3HP somewhere in
the powerband is getting way off the subject.

-DO

More... (looser matching)
- Last day of these posts
- In 2000 , 2001 only
- Today
- New search...