Adamw Posted March 5 Report Share Posted March 5 6 hours ago, Laminar said: In the log below, there's a failed crank event followed by a successful crank event. Do I need to adjust my sync tooth and/or window? It's still 50/50 on whether it starts or just chuffs and/or blows fire. It sounds like a mis-sync issue. Although in that log the RPM drops to zero shortly after the first revolution which would suggest it wasn't happy with the crank signal as well. Do us a few more crank scope captures. 6 hours ago, Laminar said: I calibrated my MAP sensor (factory IAT/MAP combo sensor Ford GV2A-9F479-BA "Germany") based on a known good MAP sensor and got 0V=10kPa and 5V=304kPa. The factory tune (via HPTuners) says the MAP sensor has a slope of 65.12kPa and offset of 0.47kPa. Using those numbers, I would calculate 0V=0.47kPa and 5V=326kPa. I can hook up a vacuum gauge to the manifold and compare it to the MAP reading while running but I haven't yet. All of that to say I'm not sure why my MAP signal is so high compared to what I would expect at idle. The HP tuners cal sounds more normal for a 3Bar MAP. Do either of the built-in bosch 3bar cals give you a MAP similar to BAP with engine off? The 3.0bar @ 4.65V one covers about 95% of bosch 3bar sensors. Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 6 Author Report Share Posted March 6 Switching to the [email protected] shows me 91.8kPa MAP vs 98.3 BAP, the 4.5V version was further off. It was enough that it wouldn't let me do a MAP sensor calibration. I hooked my vacuum gauge up to the MAP sensor and when I pump it down it all matches, so I believe the MAP reading is accurate. I reinstalled my piston stop and double confirmed my TDC marking is correct. I've attached 5 more trigger logs. I switched trigger2 between Rising and Falling to see if that would make any difference. Trigger Scope - 2024-03-5 9;16;31 pm trigger5 rising.llgx Trigger Scope - 2024-03-5 9;15;49 pm trigger4 falling.llgx Trigger Scope - 2024-03-5 9;15;15 pm trigger3 falling.llgx Trigger Scope - 2024-03-5 9;13;57 pm trigger2 falling.llgx Trigger Scope - 2024-03-5 9;12;59 pm trigger1.llgx Latest tune: https://drive.google.com/file/d/1qqKgCamkiNo7lV1rKo_M58vOJPo7blW8/view?usp=drive_link Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 8 Author Report Share Posted March 8 Still interested in a diagnosis on the triggers. I think my MAP issue was just the thing running so rich that I was down a cylinder or two and the extra throttle required to keep it running was dropping the MAP. I got that straightened out, you can hear where the 6th cylinder finally joins in and it revs clean for the first time: Next up is VVT. I was able to calibrate and have tooth counts and offsets for all of my cams. After some digging around, I assumed that the target for the inlet position should be positive to advance it and the target for the exhaust should be negative to retard in. I set my inlet advance tooth tolerance to 70 deg, retard tolerance to 0, exhaust advance to 0, exhaust retard to -70deg. In looking at my tune again, for some reason I still have negative values in the inlet target table, so that may be why I never saw any angle changes on those channels. I set my base duty cycle at 40%, which may or may not be correct. From the factory tune I was able to pull a feed forward duty cycle table. For a desired phasing rate of 0 deg/s and at operating temp I guesstimated about 40%. Maybe 50% would be a safer bet? From my log, I see the exhaust cam duty cycles jumping back and forth from 0 to 100% and the cam angle steps from 0 to -70 degrees with a Tooth Tolerance/CAM Signal Error. I picked my PID parameters based on a couple base tunes for other engines, maybe I need to pull those waaaay back so I can see what it's actually trying to do. Latest log: https://drive.google.com/file/d/1VuFxpntUiS2G_-Cd9eVvNilgDD5V1YDA/view?usp=sharing Tune: https://drive.google.com/file/d/1uBFR3lPPqtcR7k4hVpH-D17ytNKuTErY/view?usp=sharing Very interested in everyone's thoughts on the trigger issue and VVT as well. Thanks! Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 9 Author Report Share Posted March 9 I happen to have access to another 2.7 vehicle, so I went out and trended some parameters: Here's what I'm noticing. Base DC (lines 3 and 4) is 0%. When the ECU calls for VVT angle (lines 1 and 2) duty cycle jumps up and then holds. Exhaust angles are described as positive and intake angles are described as negative. What are the odds that the nomenclature is different for this engine? Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 10 Author Report Share Posted March 10 Okay, big step forward. I plumbed up the coolant system so I could run it for more than a minute a day. The key was changing the aux active state to Low for all four output channels. After that I redid the cam calibration and the engine started perfectly idling at 600rpm, it was nearly silent. I think with active High it was defaulting to all of the way advanced/retarded which is why it sounded like a ghost cam tune. I didn't have time to dial in the PID but I can see it actually trying to work now. Latest tune: https://drive.google.com/file/d/1qqKgCamkiNo7lV1rKo_M58vOJPo7blW8/view?usp=sharing Log: https://drive.google.com/file/d/1WL95rkQ-tZhLatSavl8NHwllVnvRO37K/view?usp=sharing Still curious on those trigger logs. I'm wondering if the active state would be trying to throw off cam position while cranking, or if it would have no effect due to startup lockout. Quote Link to comment Share on other sites More sharing options...
Adamw Posted March 10 Report Share Posted March 10 Yeah with the aux active state wrong the cam had possibly moved during cranking and given us an incorrect trigger scope that I based my cam window on. So can you do us a new cranking trigger scope with the cam actuator unplugged for at least the trigger 2 cam and I will take another look. Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 18 Author Report Share Posted March 18 Sorry, was out on spring break last week so I didn't redo the trigger log yet. I did spend some time on startup and idle parameters, and I worked on the VVT PID which looks pretty good now, I replicated the factory behavior of 0% base duty cycle. I did some rough tuning to the fuel and ignition tables and it revs fairly well now (I fixed that off-idle stumble after recording the video). When it starts it does it fairly easily like in this video. If it's not going to start, it just cranks and cranks. The oil pressure control solenoid works, I'm able to adjust the duty to that solenoid and manipulate oil pressure. I tried playing with the boost control solenoid but I can't get anything to happen. I swear I was able to buzz the solenoid when I first wired everything up, but I'm getting no response now, even though I'm measuring 12V at the solenoid plug and I can see the state change on the output channel when I go into test mode. I've ordered another solenoid in case I smoked this one somehow. I have turbo drain lines pulled off to repair some pinholes, so I'll have to get those resealed before I can crank it over for new trigger logs. Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 19 Author Report Share Posted March 19 Messing with temp sensor calibrations. Maybe I'm doing this wrong... Dipped the sensor in a pot of near-boiling water and compared that to my meat thermometer. Recorded temperatures vs. input voltage as the water cooled from boiling. Gradually added ice cubes until I got down to freezing. There's definitely error possible due to the update rate of the meat thermometer, but I was hoping the general recorded curve would reveal a common calibration curve. Once I had the temp vs. input voltage, I back-calculated the sensor resistance by R_sensor = (V_in * R_bias) / (V_ref - V_in). For some reason I thought the 10kohm bias was better on the ECT and 1kohm worked on the IAT. I graphed them with a logarithmic y-axis and added a trendline with the lowest polynomial that gave me the best R^2 value. ECT looks maybe similar to the Std Bosch NTC. IAT is from the combo MAP/IAT sensor and I'm not sure if it's close to a known standard. Quote Link to comment Share on other sites More sharing options...
Adamw Posted March 19 Report Share Posted March 19 The ECT looks similar to the Falcon CHT. About 37K at 20°C. The IAT looks like Std bosch, 2.5K @ 20°C. Quote Link to comment Share on other sites More sharing options...
Laminar Posted March 19 Author Report Share Posted March 19 Does a specific sensor or calibration require a specific pullup value? I set ECT to Falcon CHT and IAT to Std Bosch. My garage thermostat says 60degF and my meat thermometer says 58.7degF. ECT is reading 58 w/ 10kohm and IAT is 63 w/ 1kohm. Quote Link to comment Share on other sites More sharing options...
Adamw Posted March 19 Report Share Posted March 19 The pull-up will affect the resolution. For example if you have a sensor with a resistance of 100000ohm and a pull-up of 1000ohm then you are going to see a very small voltage change for a change in resistance. So the general idea is to use a pull-up value that gives the largest voltage change in the range of temperatures that are most important for your application. 10K for the ECT and 1K for the IAT would be the best choices in this case. Quote Link to comment Share on other sites More sharing options...
Laminar Posted April 1 Author Report Share Posted April 1 Finally got my new oil drain tubes from Aeroflow, will take some more fab work to get those installed. In the meantime it's spring and everything else in the garage wants attention, so I haven't gotten to them yet. Any opinions on Deatschwerks injectors? I think these should be good for over 500hp on E85 at 85% DC. https://deatschwerks.com/collections/injectors/products/16u-00-0090-6 Quote Link to comment Share on other sites More sharing options...
Laminar Posted April 8 Author Report Share Posted April 8 Got my injectors last week. Still need to get those oil drains fixed before I can crank it. Given this data from DW: Does this look right: Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 2 Author Report Share Posted May 2 Oil drains are repaired, we'll see if I got the leaks this time. Went ahead and threw the new injectors in last night. The worst deviation I see is injectors 115/116 at 2ms showing about 1.3% off of average and most are under 0.5%. Is that worth trimming out with an individual cylinder trim table? Quote Link to comment Share on other sites More sharing options...
Adamw Posted May 3 Report Share Posted May 3 No I wouldnt trim for that amount or error. Assuming their flow measurement is even accurate to that level, I would say in general +/-3% wouldnt make any detectable difference in the way it idles or drives. Laminar 1 Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 4 Author Report Share Posted May 4 If I'm changing injector flowrates, what adjustments do I need to make because the injector flow? Accel enrich calculates additional pulsewidth to add, does it take into account injector size? Pre-crank prime? Anything else that uses raw pulsewidth ms? Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 5 Author Report Share Posted May 5 Okay, got the oil system sealed back up and took two trigger logs with the VVT actuators unplugged. https://drive.google.com/file/d/1Xgm4ljHgrRzZuPWdxiVXaQPx-qBhak2k/view?usp=drive_link https://drive.google.com/file/d/1XijZ6nJWgktQ-NA-AbI4L26sUac5YpTD/view?usp=drive_link Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 6 Author Report Share Posted May 6 @Adamw Any useful info in those trigger logs? I have it running great on the new 90lb injectors, only issue is that at idle I'm sitting on the min pulsewidth and running about 0.76 lambda . I can only pwm the fuel pressure down to about 155kPa, so I may just have to live with a rich idle. Sits at 600rpm nicely though. Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 6 Author Report Share Posted May 6 Okay, tell me if I'm looking at this correctly. This is my unplugged trigger scope: I measure two crank gaps at 720 degrees. From there, if I measure back from the tooth before the gap, I get 160 degrees to the single tooth and about 41 degrees to the rightmost of the group of three teeth. I assume I want to look for that rightmost tooth so that I'm looking for a tooth that only exists during one unique cycle of the engine. The rising edge of that tooth is about 300 degrees after the gap, which puts it around tooth 50. If I made, say tooth 52 the sync tooth and used a window of 80 degrees, the system would check for the 80 degrees before tooth 52 which should give me a solid capture on that rightmost tooth. Does that seem like it will work or am I missing something? Quote Link to comment Share on other sites More sharing options...
Adamw Posted May 7 Report Share Posted May 7 Yeah you pretty much have it right, but a couple of small corrections: "Sync tooth 1" is on the right hand side of the highest point on the trigger 1 state, where it changes from "tooth after gap" to "counting timeout". So you are working from the wrong side of the gap. There is a further complication that I think may apply - Im not 100% sure with cam window sync mode but for all the other sync modes you cant have any active VVT cam edges move across the sync tooth. So I think the below will work, which is basically the same as you have arrived at. It is a bit of a busy picture so not sure how easy it is to understand. I have marked the window in pink starting at 300° & 660° (tooth 51) and 70° length (marked 230 and 590°). It doesnt look like any of the other cam edges will cross the 300/660 lines. Im using an oscilloscope overlay software called gabaritop which is quite useful for this sort of thing, there is a free version that is slightly more restricted but still very good. 21 hours ago, Laminar said: I'm sitting on the min pulsewidth and running about 0.76 lambda . I would ignore their minimum PW, that seems unusually large so not sure how they arrived at that. Try it at about 0.3ms, if you start to see erratic behaviour under over-run conditions or idle lambda isn't repeatable under similar conditions then bump it up. Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 7 Author Report Share Posted May 7 Awesome, thanks for the help! Quote Link to comment Share on other sites More sharing options...
Laminar Posted May 9 Author Report Share Posted May 9 Gave this a shot last night. Lowered my min effective pulsewidth to 0.3ms. It wasn't happy idling at 1.0 lambda, had some occasional pops. Richened it to 0.9 and it was happy. Extra bonus was my eyes weren't watering from all of the fumes of idling at 0.7. Used sync tooth = 51 and window = 70 degrees. Got a start on 4 of 9 attempts, still seems like it's 50/50. Tune: https://drive.google.com/file/d/19qBIifczMji_UOj1DnK4y8IkUmxW5Hl4/view?usp=sharing Log: https://drive.google.com/file/d/1Y90i_sJp0JXn6pNj9dRMyCgdvzyAFg_T/view?usp=sharing Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.