Tightmopedman9

Tuners Rejoice! Free Tuning For M4.4!

Recommended Posts

Wow! Today i found out some very interesting stuff.

To those tuning on a 15G and still using the stock orange injectors BE VERY CAREFUL!

Today I hit 138% duty cycle at only 13 psi.

I believe this is because of all of my supporting mods. I was maxing the MAF at only 14 PSI on my 15G!!!!

I couldnt figure out why car my was knocking sooo badly at 16-18 psi around 4000-5000 RPM. No matter what i did to increase fuel my AFR would not get any richer.

I decided to calculate injector duty cycle and it was through the roof.

I talked a little about this yesterday, but today i realized how badly this issue really is. With your injectors maxed, when the ecu detects knock and tries to add fuel, it will essentially have no effect.

To those of you running orange injectors with a 15G above stock boost with supporting mods to aid in volumetric efficiency be careful and monitor injector duty cycle.

Edited by Simply Volvo

Share this post


Link to post
Share on other sites

Also, has anyone been able to go above 10.8ms of load? That seems to be the maximal amount of load I see, even though at that point boost and RPMS are still rising.

Great ideas on the MAF scaling, I was thinking along the same lines. However I was going to change the 'load threshold lambda control' to 0 ms for all values. I guess setting LTFT and STFT limits to zero would have the same effect? Also, since the VE map is a correction factor map and not a lambda target map should a correction factor of '1' correspond to 14.7?

Interesting. Did some digging around the DAMOS and noticed a parameter TLMAX.. TL being the usual DAMOS abbreviation for load. byte in 607/608 is 0xC937, comes from the factory as 0xE1 (225), 225*.048 = 10.8 -- Try setting it to FF, which would be 12.24.

Good point, the load threshold for lambda control should do it, although I've not tried this. It might still cross-count the O2 sensor at idle, just not under part-load. Zeroing it all out should be a safe bet.

Yes, in concept, 1.0 in the VE map *should* be 14.7, but not being an actual target lambda, it will only be 14.7:1 if the injectors and MAF are both correctly calibrated. Technically it's simply a correction factor applied to the injection time value.

And yikes, Simply. Time for an upgrade indeed.

I see most tendency towards retard on cylinder 3 and 5, although it's not always consistently that way, just seems more common. I'm fairly sure the cylinders are in the correct (physical) order, based on how the code is written and what the DAMOS says, I just don't really know how to confirm it. I need to do some more digging there - I was thinking if I could figure out how the ECU stores in memory data related to the crank and cam position sensors, I might be able to correlate from that... But my initial attempt was unfruitful and I haven't got back to that yet.

Thanks dream3r for the suggestion on ecuxplot. Will have to check that out. I use Excel for most things because I'm used to it, but it's graphs indeed suck!

hotpantslovebug, knock in higher RPM ranges is definitely not good, these cars seem sensitive to the low-mid 3k range, even the stock knock detection map has increased values around there to reduce sensitivity, and cam timing (retarding the intake cam a bit) helps that alot, but it seems knock in the higher RPM range under load is more concerning. How much load are you seeing when it knocks? Remember that knock retard numbers from the knock retard adaptive map (XFALTx) stick around for the RPM/load range, so look for a value >128 in "BITS" (bit 7 set) or >0 in "KRCOUNT" which indicates active knock detect between the previous and current log line. That'll show you the actual load/ignition angle/RPM combination where the knock occurred. Also, the retard numbers (XFALTx) stick around between drive cycles, and the logged values are for the RPM/load region (there are 15 of those) under which the motor is operating -- so it's possible to see numbers in an adaptive rpm/load region with values from knocks detected on previous drivecycles.

Edited by mercuric

Share this post


Link to post
Share on other sites

Hmmm... That's strange, I wonder why they would have a maximal load? I'll add it to my .xdf. I only scaled my load axis up to 10.4ms, so reading load higher than that won't make a difference, but it would be interesting to see what load I can hit ^_^

I tried the BMW map scaled by +20% and it was sitting at 17.8-18.2 at idle. I went back and scaled it by +35%, but my car wouldn't start! I think it's my CPS, but I don't really feel like investigating tonight.

Share this post


Link to post
Share on other sites

Good question, I was wondering the same! Hopefully it continues to raise injection time past 10.8 and the limit is just arbitrary.

I wonder if the difference in accuracy goes down as flow goes up, say idling or cruising at 2000rpm or so? My guess is the tube size change makes more of a difference at lower flows than higher flows, due to the low velocity of the incoming air. But I could be off base there...

Did some work on a routine to hold the TCV closed until a minimum load is reached, going to drop this in my binary and see what happens in the morning.

; ------------------------------ Compare TPS position to threshold, return if TPS position is less than threshold
90 F9 23  MOV  DPTR, #0xF923       ; Current TPS position XRAM address
E0        MOVX A, @DPTR            ; Get TPS position
F5 F0     MOV  B, A                ; Move it to B
E4        CLR  A
90 E0 F1  MOV  DPTR, #0xE0F1       ; 0x1E0F1 - TPS threshold for minload routine
93        MOVC A, @DPTR+A          ; Get TPS threshold to A
B5 F0 00  CJNE A, B                ; Compare A (threshold) to B (current TPS), set carry if A < B (B > A)
40 03     JC  +3                   ; Jump into routine (over RET) if B (current TPS) is greater than A (threshold)
74 00     MOV A, #0                ; Set A to 0 to indicate no duty cycle modification
22        RET                      ; And return to caller
; ------------------------------ Get minload for current RPM, bail if 0 (no minload for current RPM)
E5 3B     MOV  A, RAM_3B           ; Move RPM*30 (NMOT) into A
75 F0 0A  MOV  B, #0xA             ; For /10
84        DIV  AB                  ; Divide load by 10, RPM / 300 now in A
90 ED 00  MOV  DPTR, #0xED00       ; 0x1ED00 - base of RPM/load map
93        MOVC A, @DPTR+A          ; Get min-load byte from base + RPM / 300
70 01     JNZ +1                   ; Jump over return if value is not zero (zero value = no action at current RPM)
22        RET                      ; And return to caller
F5 F0     MOV  B, A                ; Minload in B
; ------------------------------ Compare current load to minload, set duty cycle = FF if load < minload
E5 40     MOV  A, RAM_40           ; Move load to A
B5 F0 00  CJNE A, B                ; Compare A (current load) to B (minload), setting carry if A (current load) is less than B (minload)
40 03     JC   +3                  ; Jump over return if A (current load) is less than B (minload)
74 00     MOV A, #0                ; Current load is greater than minload, set A to 0 to indicate no duty cycle modification
22        RET                      ; And return to caller
; ------------------------------ Current load is less than minload, set duty cycle to FF, set A to FF (non-zero value induces exit from rest of TCV logic) and return
74 FF     MOV  A, #0xFF            ; To indicate TCV duty cycle modified to FF, induces jump to 15180
75 35 FF  MOV  RAM_35, #0xFF       ; RAM_35: TCV duty cycle computed value, moved to XRAM address at 0x15180
78 BB     MOV  R0, 0xBB            ; TVMBEG 0xBB
76 FF     MOV  @R0, #0xFF          ; TVMBEG = 0xFF
; ------------------------------
22        RET                      ; Back to TCV routine

Edited by mercuric

Share this post


Link to post
Share on other sites

Boost 15-17psi on rev5 map, I believe that should only be 10psi? Not sure whats going on but need to do more logging and learn fast :) Looked into Injector duty cycle and 101.9% was my highest but dipped into 80's/90's so thats no good. I actually have a set of 550cc bosch injectors so maybe it's time to throw those in. Thought stock whites were fine at my level (stockish) but i guess not?

Heres my log, yes I know its long... will be data logging smaller portions from now on.

https://www.dropbox.com/s/m94kab68gokli8y/15psi.csv

Share this post


Link to post
Share on other sites

Can you post the XDL (raw unexported log) as well?

You are going way past target load, hitting 9.5ms! Top is about 8ms on the stock maps. XP_SOLL shows the current target load from the boost map -- notice it's 7.x when you're over 9...

Something doesn't seem to be right with your hardware or something is behaving really weird. Both the I-part and P-part corrections for TCV duty cycle are going _way_ negative, reducing duty cycle to pretty much zero. But your load still goes way up. Either your wastegate actuator is way too tight (or defective), your TCV is locked shut, or you have a vacuum leak on one or both of those hoses between the TCV and the wastegate actuator or intake manifold. If the system was functioning normally, load would be limited (assuming you're using the stock maps) to about 8ms, and would reduce as duty cycle decreased. Reduced TCV duty cycle allows more pressure to flow to the wastegate actuator, opening it. It appears you are essentially running the turbo uncontrolled, hitting >800kg/h peak airflow. You might try disconnecting the TCV electrical connector -- this will throw a code, which you can clear by removing/reinstalling the ECU, but will disconnect voltage from the TCV, leaving it open, causing you to run at wastegate actuator pressure only. You should see 2-4 psi maximum in this condition. More and you have hardware to adjust/fix. If you've adjusted the wastegate significantly tighter than the factory setting, put it back to stock, which is just a few PSI.

Or, you're running a MBC... :P

Edited by mercuric

Share this post


Link to post
Share on other sites

I was wondering why duty cycle was 0. ARD TCV is only 7 ish months old and all vacuum lines are correct. Disconnected tcv, boosting > 7 psi. that's enough for me to know somethings not right with the wastegate actuator. I set it to the correct psi with ipd's wastegate adjustment tool 6 months ago. So maybe its time for a new actuator? I believe ard's is out of stock. will give the car a once over again and post XDL when I get of work.

Share this post


Link to post
Share on other sites

Fun Fun! Sounds like getting your baseline boost level in shape is the first step. I suspect that will bring things into reason. You might be sure the line to the compressor housing is in good shape, that one likes to break at the nipple if it's old hose, there's a lot of heat there and you can't really see it from the top -- you can usually feel a problem with it, though.

I doubt the valve has died so soon, and given you see a lot less boost with it disconnected, it's probably working as designed. I'd bet the actuator is what needs attention -- personally I'd just lengthen the rod 2-3 turns and see if it plays nicer.

Share this post


Link to post
Share on other sites

I've been having problems with my TCV valve duty cycle as well. I've been datalogging TVMBEG, but it always sits at 0%. I assumed it was because I had to no TCV connected, but now that I've connected a MAC solenoid it still sits at 0%. I've set the TCV duty cycle pilot control map to 2% for the entire map, as well as the TCV duty cyle min limit to 2%, yet I still shoot past 20psi. Any idea what could be up?

Share this post


Link to post
Share on other sites

I'll have a look at the code for TVMBEG logging tonight, might have screwed something up, but... You want way more than 2%. When power is applied to the TCV, it shuts, blocking pressure to the wastegate. 2% is basically running it open, so you'll get wastegate pressure. Not sure what pressure your actuator actually opens the wastegate at, and the duty cycle doesn't direcly represent a percentage of the pressure applied, so some tweaking will be required, but start with 50% to be safe, and work up from there. If it behaves like mine does, you'll reach a percentage after which boost pressure increases quite quickly. With the stock hardware set two turns tighter vs factory (about 5psi on my car) I'm getting around a bar of boost with mid-70s duty cycle numbers. This varies depending on the actuator pressure, turbocharger response, wastegate condition, etc.

EDIT: derp, re-read your post. brain is half-scrambled with work BS, lol... are you sure you have all three hoses connected to the correct places? Do you get normal wastegate-pressure boost with the TCV electrical harness disconnected? Also, send along the XDL, curious what all the parameters are looking like.. :)

Edited by mercuric

Share this post


Link to post
Share on other sites

Yeah, I've tried every possible configuration. Interrupt and bleed modes, with varying duty cycle values for each one. It seems no matter how I hook it up or configure the values it just remains 100% closed. I really wish I had a portable scope so I could see what the actual output is. I really need to pick up the DSO Nano...

I'll post up an .xdl once I'm done with the MAF scaling, almost there!

Share this post


Link to post
Share on other sites

Man, I've been looking at a bunch of PC-based digital scopes, there are many coming around that are very tempting... Waiting on the market to mature a bit first. The Nano is tempting, can't beat it for the price, but holding out for something more powerful to come down in the $200 range.

Looked over the code, don't see any issues and I'm getting solid duty cycle off mine.

Disconnecting the harness will keep it open. Start by making sure you've got proper wastegate operation with the valve disconnected electrically. A voltmeter across the harness-side connector should show battery voltage if it's indeed being closed by the ECU. You might also try the non-P-part-modified binary if you're using the P-part hack. I'll have a look over that again -- F'ed up once on the 608 version already, maybe F-up #2 in there too :ph34r:

Also note the duty cycle will always show 0 unless the throttle angle is greater than the parameter "Throttle angle threshold for LDR active" (0xE8D4), since the valve is left open (no power) until the throttle is pushed enough to toggle the LDR routines. The factory value is 20 degrees.

Had some luck with the "minimum load" hack, which keeps the TCV closed until a specified load is reached. I set the table up to activate it at 25 degrees TPS and keep it closed until 7.5ms load. Spoolup off the line didn't seem to change much, maybe a bit, but it seems to make a more pronounced impact when rolling at speed and the throttle gets hit, seems to make the turbo come to full boost a little quicker. I think I was a little too aggressive with 7.5ms, 6.7 or 7.0 would probably be better, as it added a bit of a spike, which is likely partly related to increasing my duty cycle map in the lower 3k range in an attempt to aid spoolup in the last few flashes -- Backing down the minload to 7 or so and reverting the duty cycle map will probably do the trick. Target is 9ms around 3600.

Note the last column, duty cycle, staying maxed from hitting 25 degrees TPS until attaining 7.5ms load

9Brocf3.jpg

Share this post


Link to post
Share on other sites

Fun Fun! Sounds like getting your baseline boost level in shape is the first step. I suspect that will bring things into reason. You might be sure the line to the compressor housing is in good shape, that one likes to break at the nipple if it's old hose, there's a lot of heat there and you can't really see it from the top -- you can usually feel a problem with it, though.

I doubt the valve has died so soon, and given you see a lot less boost with it disconnected, it's probably working as designed. I'd bet the actuator is what needs attention -- personally I'd just lengthen the rod 2-3 turns and see if it plays nicer.

All vacumm line are new silicone, replaced when I got the car in july last year. 20 inHg vacuum at idle and 22inHg decel. I've kept stock boost till now and just focused on stage zero. TCV plugged in = 15-17psi TCV unplugged = 15-17psi. So I think the actuator has to be toast. Tested it and it still starts to open around 3-4psi like it always has. Since ARD is out of stock I was thinking this for a replacement. Although 17psi is fun the thought of no boost control makes it hard to sleep at night so I'd like to fix and start tuning :) .3bar spring should do the trick?

http://shopping.kinugawaturbo.com/turboadjustableactuatorvolvo850s70td04ltd04hl3spring101215bar.aspx?variation=2650

Almost forgot, https://www.dropbox.com/s/endwqaxnzc3gc42/10-21%2015psi.xdl Not sure if this is needed still...

Edited by hotpantslovebug

Share this post


Link to post
Share on other sites

Your hoses to the TCV are in the correct positions right? I know when I installed my ard tcv I had them backwards and it'd over boost like crazy lol....just something to check :)

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.