Oh yeah...and what about 2-byte? How does the logging code work? I think I remember JCBanks saying that the DMA code grabs the first 64 items in the MUT table so both bytes of the "2-byte" requests should be available...if thats how it works. Does this grab the data for the entire table(or 64 bytes), move it to RAM and then the ELM fetch the contents of the RAM location? I was trying to look through the uncommented DMA code posted on the SVN but it has been a LONG time since my old 1G DSM ASM days and even longer since 8051 ASM in school. Any DMA I did in college was in a DSP in C....so...

EDIT: Oops...I guess I should have read the first post more clearly. Sorry about that...

 

ahh so u are curtis

anyway new version is ready for testing... i mean release

gota update docs and will post it on website and will post here
-------------

i just pull the 1byte tephra load normally. that 64requests was only done cuz i believe JB didnt have time to create a ram mut table... we gota do some work and let the genuises do what they are good at

anyway in my program we run a 2nd mut table in ram. u can populate the ram table by hand (i include the xml to do that via ecuflash as a part of the patching) or if someone (like u) is running a registered version there is a small check box in the requests tab. this will setup the ram mut table for you, by pulling the addresses from the rom mut table and putting them in the ram table.

to get a 2byte of anything, first is to add the two memory addresses. say u have registerd version, in the settings tab, click "add" then select custom, give it a name say "RAM B1" and the address. then click "add" again give it a name like "RAM B2". i personally set the B2 so its not logged to disk or screen. now say "RAM B1" is at ID=9, and "RAM B2" is at ID=a". in evalulation formula for "RAM B1" type the formula (Which is also in the first post of this thread):
((256*x)+&ha)*1000/256


obviously discription makes it sound more then what it really is.

 

For sd2.0 it needs basically only relocation of VE map, then just extend rom2ram copy area to copy VE map to ram too and it should work if you just add VE map to your live xml. If I remember correctly I ported dma code to 8859 so that it should be "compatible" with sd2.0 changes ( I mean I did not use same rom area for dma code what is used with sd2.0 code). At the moment I'm busy with other project but 8859 sd2.0 really was in the scope originally but I just did not do that yet

Edit: And of course dma code itself needs to be ported to sd2.0 rom or vice versa...

 

An example of 2-byte:

(Obviously your IDs will need changing...)

<anyType xsi:type="Requests">
<isAltAxisForGraph>false</isAltAxisForGraph>
<graphMultiplier>0</graphMultiplier>
<ramAddress />
<requestID>12</requestID>
<MUTRequestID>04</MUTRequestID>
<requestDisc>2 Byte Airflow</requestDisc>
<unit>HZ</unit>
<evalFunction>((x*256)+&amp;h13)*6.29/64</evalFunction>
<rounding>0</rounding>
<isLoggedToDisk>true</isLoggedToDisk>
<isLoggedToScreen>true</isLoggedToScreen>
<isGraphed>false</isGraphed>
</anyType>


<anyType xsi:type="Requests">
<isAltAxisForGraph>false</isAltAxisForGraph>
<graphMultiplier>0</graphMultiplier>
<ramAddress />
<requestID>13</requestID>
<MUTRequestID>05</MUTRequestID>
<requestDisc>2 Byte Airflow</requestDisc>
<unit />
<evalFunction>x</evalFunction>
<rounding>0</rounding>
<isLoggedToDisk>false</isLoggedToDisk>
<isLoggedToScreen>false</isLoggedToScreen>
<isGraphed>false</isGraphed>
</anyType>


Long time no post for me, apologies for the lack of feedback Ziad, I'm having trouble finding the time to play as much these days

I'm still playing 'offline' though

 

guys new version is up

check the change log for updates

 

guys new version is up

check the change log for updates

 

hi Ziad

I believe there is an error in the car list def for the evo 8 5sp gsr as it have the same values as the evo8 6sp.

Also is it possible to adjust the boost scale on the VDR graph?

thanks
Andy

 

lol funny as no one noticed before. andy do u have a 5spd evo8?? or a 6spd?

its actually should be 6spd as jdm evos are 6spd... but i can change it to 5spd, need the weight of the car, and the gear ratios..... or should i steal from evoscan/virtual dyno.... i meant borrow

andy... u can either have the multiplied by 10 or not multiplied at all. so u can set "Scale map" to yes if u running psi/bar or no if u running KPA. i think it might slow the calculations down a bit (on the slower netbooks) if we have a full on eval (like the tables or requests).

 

Is the RAM MUT table updated by an E2 command or do I have to repatch and reflash the ROM?

I cant seem to get my new requests to work... It says it put them at 13 and 14 (after I told it to use Custom MUT 02 and 03 and selected the checkbox "Set Ram MUT Table From Rom"). I also deleted the two I added above and added the addresses directly with two custom requests. I attached some images.

I displayed the raw values and they stay at 128. I have two-byte working in evoscan so I know the addresses are correct.

Any ideas what I am doing wrong?












EDIT: Ha...I continued this from last night and didnt notice the replies. I will work on it some more and post back.

 

I was able to get the two bytes correctly. The RAM MUT table was not populated with the two addresses for the RPM 2-byte request 0x6B1E and 0x6B1F. I read the MUT table from RAM, added these to locations 0xE and 0xF respectively and it started working. However I have to change it back if I do a "DiskToRAM" or "InitializeRAM" which makes sense. I did put the formula in, like above, but it still doesn't display the calculated RPM. (In my case it is ((256*x)+&hF)*1000/256 if you cant see from the pictures). Is the RAM MUT only 16 items?

Oh yeah...and this is on the new version. It is also with my Tephra v7 patched by ELM. Could this be the source of some of my issues? I haven't had time to migrate my maps to donner's ROM...I want to need to give that a go to get the live map editing to work (if it doesnt with the ROM patched by new version).

This is what came back when I did the READ (green arrow one):


Changed the addresses for 0xE and 0xF and uploaded(yellow arrow one):


As you can see from the picture, to the right you can see the lowbyte is 196...which is 765.625RPM and close to the 781RPM of the 1-byte request. Seems the requests are working once I write the MUT table. Now getting the equation to print and I am in business. Nothing I cant do in Excel for now.

Formula settings for 2-Byte RPM:

 

Hi Ziad

Yeah my car is a 5spd I edited the car details so now it works with those values, i'll send them across later when I get home. and i'll also try the scale thing, does this effect the afr to?

Andy

 

thanks andy. might include a USDM8.. urs a usdm/SA/AU evo8? or did u go a 5spd conversion or a RS with a 5spd box?

dingus, after clicking the checkbox u need to start the logger, it will copy the values and populate the ram mut table. and yes looks like i was having a brain fart. the formula for 2 byte depends on two things, the 2byte formula and the request formula. so formula for ram 2 byte will be different from load 2 byte and 2 byte rpm.

so the formula should be
(256*B1)+&hB2) * whatever the request formula is

B1=is the first bye, and can be x.
B2=request address of 2nd byte.

ur formula looks correct mate, the 2byte rpm is slightly more accurate then 1 byte, obviously. hence they will not necessarily match. ur case the diff is 16rpm... seeing 1byte has resolution of 31rpm max to me it seems correct. so i would say u didnt start the logger hence it didnt change the ram mut table.

lastly if u click ram2disk... the new mut table is written to the file saved on hdd..... which is what its intension was. if it doesnt work u need to let me know.

 

I managed to get this working after work today. I was just being dense I guess. I did have to write the address of the second byte as &h0F instead of &hF in my 2-byte rpm equation.

Can I expand the size of the RAM MUT in the ECU definition? What is adjacent in RAM to the 32B 64B table?

 

Hi Ziad

My car is a European evo 8 gsr and comes with a 5 speed box as standard. I used the following values

3rd gear ratio 1.407
4th gear ratio 1.031
final drive 4.529

I also tried setting the scale to NO but it still didn't reduce the boost graph

thanks
Andy

 

andy i will have a look again for u.... can u email me a copy of ur evolivemap folder from ur documents, i think u know my email.


dingus the 8859 rom can go a bigger ram mut table, i will give instructions, but if they dont make sense email me ur xml file only.

anyway in ecuflash xml if u copied the table as per the program then u need to change the address it would currently have "49b00" as the address change it to "4a280". also change the static X elements from 16 to 32 and the static values keep on going.

in ELM, goto settings > ECU Defination > "MUT Requests". In address change "FFFFA000" to "FFFFA780" then click on "+" under Y Elements and add "2". then u can just change to a different table this will save the changes in the mut table.

then best is to goto requests tab, add more requests if necessary and tick the check box and run the logger once so the new mut table is written to the new location.