Page 1 of 1

Selecting a MCU

Posted: Wed Oct 09, 2013 5:45 pm
by DreadyD
So my question is one that most viewing have and afraid once they see its complexity. Selecting a MCU based off several posts is rather difficult as the firmware would require some hacking to make work? Though it makes no difference to me, I would like to help sort the confusion and find a cheaper more versatile route then a MS1 kit. I spent 6 hours looking a up most of this info lol

The FreeEMS firmware was developed for the MC9S12XDP512 Microcontroller on the AD9S12XDP512M2 unit, but many user have found other units with a different controller all together and thought one would work in place of another. The MC9S12XDP512 is apart of the S12X family and often confused with the S12. The 512 is the k-bytes of Flash and my understanding the FreeEMS firmware requires about 350kb of flash space.(correct me) The extra space could be used at a latter date or used as a buffer.

Could use some help to make a list of usable MCU's and we can look for good distributors. With working links...

Working

AD9S12XDP512M2 -http://www.technologicalarts.ca/shop/st ... ation.html
They are also pre flashed with Freescale Serial Monitor

Funky

MC9S12XDP512 -http://www.wayengineer.com/index.php?ma ... cts_id=281 Needs to be flashed to be usable
Support/firmware- viewtopic.php?f=9&t=2141 - viewtopic.php?f=62&t=1515&start=70


Unsure

elmicro CardS12.XDP5 - http://elmicro.com/en/cards12.html would be a blank and need a flash

AD9S12XEP100M2 - http://www.technologicalarts.ca/shop/st ... ation.html
Faster all around and should be more accurate. Is the Non-multiplexed 8MB expanded memory bus used with FreeEMS? other then that could be a sweet upgrade for most.

Re: Selecting a MCU

Posted: Wed Oct 09, 2013 10:41 pm
by Fred
To your title: MC9S12XDP512 < FreeEMS runs on this, and nothing else. Period. In future this may change, anyone trying to change it prematurely will be castrated in a public place for disturbing and distracting the community. Additionally, after losing their balls, they'll receive no help or support in any way shape or form.

Nothing is a sweet upgrade! The MCU costs about 15USD at the right price, or 25USD at the wrong price, which pales in comparison with all of your other costs when doing an ECU setup, or even topping up your gas tank from half full. Don't be cheap. FreeEMS is about freedom, not price. There is no real world difference in performance to be had.

Taking a blank MCU requires a BDM to be used at all, as you've noticed. That adds 32+USD to the bill, plus the requirement to solder very fine pitch parts; not for everyone.

If you want to go FreeEMS right now, by FAR your best bet is to get in line to grab a Jaguar. By far.

Hope this helps.

Fred.

Re: Selecting a MCU

Posted: Wed Oct 09, 2013 11:17 pm
by DreadyD
Thank you Fred and I feel honored to have you reply first.

This is why I posted, so much needs to be put straight.

I still have my questions about the AD9S12XEP100M2 because of its speed. High rpm would require such logic?

Call me a back yard kid, but I would like to have a shot at hardware too and though the Jaguar is very nice. I would like to intergrade some of my own code for other projects related. Fuel mixing and barometric pressure as light insight. I also have most of the pcb components ready and stared my DIS. Thought about using 6 555 timers with heatsinks, but that's silly talk till I get a mcu.

I keep mostly to myself till I have related topics.

Re: Selecting a MCU

Posted: Thu Oct 10, 2013 1:29 am
by Fred
Depending on the decoder 32 thousand RPM is tested/possible... Two engines already spin 12k+ in real life. NO need for the XEP, whatsoever.

DIY builds are fun, but are fraught with dangers and problems. Ask any of the DIY builders here and they'll confirm. Do it and upgrade later.

Re: Selecting a MCU

Posted: Tue Oct 15, 2013 5:49 pm
by DreadyD
Thanks Fred, I still like that XeP for future or current prototyping.(even in non ecu prototypes) It could make a great trans tuner... I have bin watch the forums and following a little, I may end up getting a Jaguar or 2. I just wish I was around sooner to ask for a dev spot on the board, TH stye square box just for me lol but for the size of it should fit great in my 59 fiat.

The Crystal and the MC9S12XDP512 chip itself in all the current mcu or hardware would freeze here? They are only rated for -40c? http://en.wikipedia.org/wiki/Geography_ ... f_Winnipeg saved you reading and they missed a couple lol
Within the past ten years, the temperature has dipped below −40 °C (−40 °F) once on 5 February 2007 as well as both 29 and 30 January 2004. On 1 February 1996 the temperature was −41.8 °C (−43.2 °F) however, the wind chill was -57.1, making this the coldest windchill in the city's history.
An experiment I performed at school:
Two 1/4 watt resistors low 5% tolerance might do better then a single 1/2 watt resistor@20%, till one of the two fails or from expansion and contraction the total value is changed and the 20% could be used in your favor. If you don't want a resistance drop or raise just know the temperature you are operating at and test the resistor at that temperature. (this study was to make a more stable overclocked pc)

anyone test these effects? change's in resistance, also from different pressures from the mountains. Simple fix if you overclock or add a heater lmao but I would like to avoid this and sealing it in a vacuum. Already know electromagnetic or certain frequency's have a negative effect as everything should have a shield. (damn cellphones)

Also the sleep or hibernate modes the mcu have might help this problem? Does it stay warm?

Crazy canuck wants to go ice racing in the firefly(thought of that today)

Re: Selecting a MCU

Posted: Tue Oct 15, 2013 6:12 pm
by Fred
If you're silly enough to go out driving a modified car in below -40C (once every 3 years or so), I don't pity you :-)

Storage temperature range: –65 to 155 C - it's safe. It will likely continue to work outside of the official range, too, but no reliability can be guaranteed. Finding a clock to keep working is a bigger challenge.

You're not the first one to buy an XEP card, however if you do any work towards using it before we release 0.5.0, I'll consider it an act of sabotage and do whatever I can to crush you. Ditto any other porting efforts, to any other platform. Forks are evil. Together we stand, divided we fall. Don't be the wedge.

Most part values are chosen very optimally and will still perform well at quite wide tolerances.

Fred.

Re: Selecting a MCU

Posted: Tue Oct 15, 2013 8:27 pm
by DeuceEFI
DreadyD wrote:"... I may end up getting a Jaguar or 2. I just wish I was around sooner to ask for a dev spot on the board, TH stye square box just for me lol but for the size of it should fit great in my 59 fiat.
You can always join in one of the discussions in the Jaguar section:
viewforum.php?f=67
The Jaguar board originally was designed for my GM V6 with DIS ignition, but has morphed into something useful to a larger community. It fits nicely into most OEM ECU enclosures, I have installed a previous version in a factory GM ECU enclosure on my S10, others have used Toyota and Mazda factory enclosures to house their Jaguar boards.
DreadyD wrote:The Crystal and the MC9S12XDP512 chip itself in all the current mcu or hardware would freeze here? They are only rated for -40c? http://en.wikipedia.org/wiki/Geography_ ... f_Winnipeg saved you reading and they missed a couple lol
Within the past ten years, the temperature has dipped below -40 °C (-40 °F) once on 5 February 2007 as well as both 29 and 30 January 2004. On 1 February 1996 the temperature was -41.8 °C (-43.2 °F) however, the wind chill was -57.1, making this the coldest windchill in the city's history.
1. Windchill only applies to objects that contain water content, electronic components and metal can't "feel" the effects of windchill.
2. The regulators will warm up the specified enclosure enough to keep the ECU at above the -40C mark if supplied with 13.8-14.4v (battery/alternator voltage).
3. I would suggest that the ECU be placed in the (hopefully heated) operator cabin/cockpit if you plan on using it in sustained conditions below -40C. The MCU and crystal will may function below -40C but their accuracy isn't guaranteed. EDIT: I have also not tested a Jaguar PCB below -40C.
DreadyD wrote:Simple fix if you overclock or add a heater lmao but I would like to avoid this and sealing it in a vacuum. Already know electromagnetic or certain frequency's have a negative effect as everything should have a shield. (damn cellphones)
As stated above, the regulators will supply a small amount of heat to the specified aluminum alloy enclosure (or an OEM ECU enclosure) and it will act as a Faraday cage IF the enclosure is grounded separately from the PCB (as designed the PCB ground does not come in contact with the enclosure).

Also, overclocking an embedded MCU is not a good idea, you would be forcing it to work outside of the manufacturer's specifications which would mean that no measurement or calculation would be accurate. I come from a metrology (precision calibration laboratory) background and I don't recommend any practice that would make a precision measurement inaccurate on purpose. (I also do not recommend anyone "overclock" their PC if they depend on it for their "day" job. Just my $0.02
DreadyD wrote:Also the sleep or hibernate modes the mcu have might help this problem? Does it stay warm?
The MCU stays cool, no noticable heat is generated. I measured the MCU temperature at an ambient temperature of 18C the MCU was measured to be 18C during normal operation (controlling the 4 cylinder engine in my Chevrolet S10).

Since a regulator would be active (in other designs such as the RavAGE PCB) there would be a small amount of heat from the voltage regulator but I've not done any research into if it would be enough to keep an enclosure such as the one I specified for the Jaguar PCB warm enough at -40C (or below).

One last point I wanted to make was that the 9S12XEP MCU and the 9S12XDP MCU have the same operating and storage temperature limits, so there is no real reason to think that you have to go to the XEP MCU, it won't work any better or worse than the XDP at Artic temperatures.

Edit: Strike through "will" and added "may" in regards to the operation below -40C. Also added a clarification that I haven't tested a Jaguar below -40C.

Re: Selecting a MCU

Posted: Tue Oct 15, 2013 8:45 pm
by Fred
Re XEP at low temps, it could be worse (cooler), as it's manufactured with a finer process that uses less power.

If you load up your IO on the CPU it does put out a bit of heat (when on, in PLL, and driving outputs), but it's not very noticeable.

Re: Selecting a MCU

Posted: Wed Oct 16, 2013 12:28 am
by DreadyD
If you're silly enough to go out driving a modified car in below -40C (once every 3 years or so), I don't pity you :-)
"Crazy canuck" I had a 87 Mazda 323 LX (Lust Xtream) she had no problems in 2007 because her ecu was analog style lol. Only person that made it to work that day and I made 3x my wage.
Storage temperature range: –65 to 155 C - it's safe. It will likely continue to work outside of the official range, too, but no reliability can be guaranteed. Finding a clock to keep working is a bigger challenge.
What I was looking for and could not find thank you Fred, the firefly IC cracked cant even see the numbers and it a common problem with the turbo cars in canada it seems. Starts with a mas error and ends with over correction. I do have 2 other ecu's that suffer that same problem and spun 3 turbos because the mas closed at 4000 rpm boosting 9 psi. My firefly was never winter driven just not stored correctly.(My badish) Still trying to correct the problem, why ice racing and rolling it cross's my mind lmao
You're not the first one to buy an XEP card, however if you do any work towards using it before we release 0.5.0, I'll consider it an act of sabotage and do whatever I can to crush you. Ditto any other porting efforts, to any other platform. Forks are evil. Together we stand, divided we fall. Don't be the wedge.
I have yet to order any part of the ems, only have whats on hand and that's on a breadboard waiting a mcu. (needed something to do too clear my mind on a repair)
Seems like I'm not the only one that has looked at it and sounds like you have something planed? (I like surprises)
Re XEP at low temps, it could be worse (cooler), as it's manufactured with a finer process that uses less power.
As my thought concluded. Thank you Fred.


Thanks Deuce and great work, ill try to make my way to the discussion sooner or latter and I will be ordering 2 boards for sure. (1 extra to play around till a new release)
1. Windchill only applies to objects that contain water content, electronic components and metal can't "feel" the effects of windchill.
Not fully true, my research found resistors, lacker and most of the copper are effected at those temps, CPU pin are gold plated but at -40 or lower the plating can come off, still very rare It has happen.(could be from the electroplating) I know the conditions for failure might have to be perfect, but for consistent winters most electronics suffer and I toss out most of my winter repair jobs because of it.(Cost to high)
2. The regulators will warm up the specified enclosure enough to keep the ECU at above the -40C mark if supplied with 13.8-14.4v (battery/alternator voltage).
3. I would suggest that the ECU be placed in the (hopefully heated) operator cabin/cockpit if you plan on using it in sustained conditions below -40C. The MCU and crystal will may function below -40C but their accuracy isn't guaranteed. EDIT: I have also not tested a Jaguar PCB below -40C.
Good to know I might be the first too test at -40... you guys will help if I "hit a deer" right? I plan to have one of them 100% un-modded... (collection starter) any mods I do make might be to the case... ice racer = no windows
Also, overclocking an embedded MCU is not a good idea, you would be forcing it to work outside of the manufacturer's specifications which would mean that no measurement or calculation would be accurate. I come from a metrology (precision calibration laboratory) background and I don't recommend any practice that would make a precision measurement inaccurate on purpose. (I also do not recommend anyone "overclock" their PC if they depend on it for their "day" job. Just my $0.02
Not to butt heads with a dev. "Page.79" on the ds.
Always risks, its a risk to build it yourself too and that don't bother me. If manufacturer's specifications were kept we would not have advanced so far with electronics and be suck in the analog days. Its all in good learning fun, I do see the complications and know what it might do to the chip itself "if it works" and out of respect I'll wait. If I can help, I will and right now I'll just get caught up. Got over 1000 page's to read up on lol

As for the PC, it is my "day" job lol but my computer is also about 7 years old running compilations and extractions that gives my newer computer a hard time. Still battling 939 3800x2 3.49 ghz,1034mhz ddr 2200mhz system (scoped) and I don't overclock a newer computer till its needed, I just keep the option open. I also do hardmods and go beyond the bios if needed. Thermodynamics is key. (as far off topic ill get lol)

I have to thank both Fred and Deuce for contributing so quickly. Hope to get started right away and beta this winter.

Re: Selecting a MCU

Posted: Wed Oct 16, 2013 9:41 am
by Fred
Seems like I'm not the only one that has looked at it and sounds like you have something planed? (I like surprises)
You're not, and I get more and more grumpy every time. Anything that means a stable and complete FreeEMS (the surprise) is further away makes me angry. Premature xgateulation, premature XEP-dreaming, premature ARM-porting, you name it, if it's not working toward our common goal, then it IS working *against* it, which the cheerful people who tend to want to do these things are invariably slow to realise.
1. Windchill only applies to objects that contain water content, electronic components and metal can't "feel" the effects of windchill.
Not fully true
Yes, it absolutely is fully true! You can not have latent heat of vapourisation unless you're vapourising something. Wind only helps non-water based objects reach air temperature faster.
I have to thank both Fred and Deuce for contributing so quickly.
You're welcome.

Fred.