Title: Need help with DQ250 reverse Post by: fknbrkn on September 21, 2024, 06:59:36 AM Sooo i have no background in DQ operations and also hw for reading ram so im intuitive here
I want to manage input torque limiter, as i read somewhere its limited at 450 in code and also 410nm diagnostic limit? Thats all input So i start with the ida/ghidra and a bit stuck in addressing mode Lets take 02E300051C as fully defined and mmr_MMaxGetriebeschutz_ko limiter as example @80031744 Ive got no xrefs so dealing with the pointers Code: copyMotormomentMapsToRam: ; CODE XREF: copyMapsToRam+C2↓p This code make pointer(?) to 8003164C (start of section) to D0005D2C with length of section = 0x110 Code: void FUN_80069648(void) Doesnt look like copy code so i think its pointer but duno why it uses array length 0x80031A3A(start of section) - 0x8003164c (map address) gives 0xF8 or [7C] index in array it brings me to this routine Code:
the psVar16[0x77] is the mmr_MVorhaltGetriebeschutz_ko if im not mistaken psVar16[0x7f] is the mmr_freigabe_ko_Befuellfunktion which is 0xEF in this file so i can translate this into: if ( !mmr_freigabe_ko_Befuellfunktion.15 || (mmr_MVorhaltGetriebeschutz_ko + inputTorque(?)) <= mmr_MMaxGetriebeschutz_ko) { if mmr_MVorhaltGetriebeschutz_ko + inputTorque(?) + 300 < mmr_MMaxGetriebeschutz_ko { uVar8 = DAT_d0005d50; //filtered input torque? DAT_d0005d50 = uVar8 & 0x7fff // limited to 32767? } } else { uVar8 = DAT_d0005d50; DAT_d0005d50 = uVar8 | 0x8000; // } i believe this routine convert value to signed Could someone confirm im on the right way and wtf is that +300 offset ::) Or maybe some input? I was try to find can-bus messages as the input to my research but no luck here Title: Re: Need help with DQ250 reverse Post by: prj on September 21, 2024, 07:10:10 AM There is no such limits that you are talking about.
Diagnostic limit, you can rescale if you like. Clutch torque has a hard limit, but 410 or 450 is not it. Title: Re: Need help with DQ250 reverse Post by: fknbrkn on September 21, 2024, 07:20:07 AM There is no such limits that you are talking about. Diagnostic limit, you can rescale if you like. Clutch torque has a hard limit, but 410 or 450 is not it. iirc its 600nm What is the right way if i want to increase pressure without fckup ikp adaptation? Should i search for the 600nm hardcode cap and raise it a bit (with cal changes ofc) or its a more complex way? Or its better to go with some kind of torque-to-pressure tricks? Title: Re: Need help with DQ250 reverse Post by: prj on September 21, 2024, 10:35:12 AM It's not 600nm, it's 32767, and think about what the scalar is.
But 600nm is good to stop if you have stock input shaft because you will snap it in half. First make sure you are even hitting the max clutch torque, and make sure you are reporting correct torque to the TCU. And even if you mess with adaptations then it's not good (or you have to mess so much that it will drive like shit at part throttle). Title: Re: Need help with DQ250 reverse Post by: fknbrkn on September 25, 2024, 01:45:42 AM It's not 600nm, it's 32767, and think about what the scalar is. But 600nm is good to stop if you have stock input shaft because you will snap it in half. First make sure you are even hitting the max clutch torque, and make sure you are reporting correct torque to the TCU. And even if you mess with adaptations then it's not good (or you have to mess so much that it will drive like shit at part throttle). I mean the clutch torque Title: Re: Need help with DQ250 reverse Post by: prj on September 25, 2024, 06:23:43 AM I mean the clutch torque I told you what the max is ffs :D Title: Re: Need help with DQ250 reverse Post by: prj on November 12, 2024, 10:03:29 AM I mean the clutch torque Here's the patch for max clutch torque if you didn't figure it out yet: https://tools.vehical.net/DQ250Patch (https://tools.vehical.net/DQ250Patch) P.S. Clutch torque has nothing to do with the torque you transmit from the engine. It is internally back calculated torque. It's done by multiplying the clutch pressure with the clutch roughness constant. So no matter what you do with torque, you're still going to hit this limiter, and the gearbox always knows the real torque for the same reason. The only workaround for this is fudging the adaptation values, but that makes the transmission run too much pressure all the time and it's really not ideal, because you're fudging them super hard as the pressure hits a wall from 500+ clutch torque without the patch. Title: Re: Need help with DQ250 reverse Post by: fknbrkn on November 12, 2024, 11:08:15 AM Here's the patch for max clutch torque if you didn't figure it out yet: https://tools.vehical.net/DQ250Patch (https://tools.vehical.net/DQ250Patch) P.S. Clutch torque has nothing to do with the torque you transmit from the engine. It is internally back calculated torque. It's done by multiplying the clutch pressure with the clutch roughness constant. So no matter what you do with torque, you're still going to hit this limiter, and the gearbox always knows the real torque for the same reason. The only workaround for this is fudging the adaptation values, but that makes the transmission run too much pressure all the time and it's really not ideal, because you're fudging them super hard as the pressure hits a wall from 500+ clutch torque without the patch. we have now run adaption trick but thats a silly way and im still want to figure it up by myself without a2l and hardware )) anyway thanks for the link and btw do the vcds value able to run over 500 or its a diagnostic limits also? Title: Re: Need help with DQ250 reverse Post by: prj on November 12, 2024, 11:18:08 AM we have now run adaption trick but thats a silly way and im still want to figure it up by myself without a2l and hardware )) anyway thanks for the link and btw do the vcds value able to run over 500 or its a diagnostic limits also? IIRC on Fxx/Exx you have diag limit and on MQB you don't have. But I might remember that wrong. I think easiest way to buy the patch for 200 one time and then you know how to do it on every DQ250 gearbox after that. At least if somebody offered me the solution for that money when I was developing this patch 10 years ago I would have taken it in one nanosecond, because it took me some time to understand what is going on. The patch is quite simple, just tedious as there's over a dozen places that need to be modded. |