mattdavies Posted February 10 Share Posted February 10 Really enjoying the updates, keep them coming. Nice to see the developement of the platform and see products I didnt know existed, the PHR fuel tank Quote Link to comment Share on other sites More sharing options...
jagman Posted February 11 Share Posted February 11 Some complex stuff going on here , as you are pushing stress levels on the engine …I think CHT is an important parameter ….I am from an aircraft background and CHT is the main monitoring parameter EGT is fine but looking at heat energy leaving via exhaust valves at a relatively low load situation rather than the higher in cylinder loads at peak . Cheap and easy to install and will give in cylinder data ,that also links to oil temp control and water temp control (new rad fans)and handy for any nitrous use . I guess that the gearbox clutches are the main torque limiters ,any stronger ones available? Quote Link to comment Share on other sites More sharing options...
Mike2JZ Posted February 11 Author Share Posted February 11 29 minutes ago, jagman said: Some complex stuff going on here , as you are pushing stress levels on the engine …I think CHT is an important parameter ….I am from an aircraft background and CHT is the main monitoring parameter EGT is fine but looking at heat energy leaving via exhaust valves at a relatively low load situation rather than the higher in cylinder loads at peak . Cheap and easy to install and will give in cylinder data ,that also links to oil temp control and water temp control (new rad fans)and handy for any nitrous use . I guess that the gearbox clutches are the main torque limiters ,any stronger ones available? Hey, the coolant temp sensor on GE VVTi head is on the side of cylinder head at the front. Also have a k type thermocouple mounted to rear of cylinder head, so hoping that will give me enough scope to see how the CHT is doing. Yeah theres plenty of clutch pack kits and billet replacement parts for this box, so covered for a lot of future power endevours. 2 Quote Link to comment Share on other sites More sharing options...
Mike2JZ Posted February 12 Author Share Posted February 12 (edited) Back into the engine bay: - Mounted racegrade EGT on chassis rail - Installed all 7 lambda sensors - Ran vacuum lines - Made some hardlines for oil - Coolant hardlines for electronic gate + turbo - Changed the JUN cam pulley bolts with some ARP variants so they didnt rust - Started filling and bleeding any fluids where required etc. Starting to get closer to the goal of having mechanical systems ready to go, but I couldn't find a nice way to run 6 lambda sensors wires in a neat way, which was a bit annoying as everything else in the bay looks fairly neat and wireless. New plan with this is to build a breakout box that will sit on top of the exhaust cam cover using 8STA connectors for each sensor. Again a bit overkill, but I think it be icing on the cake in terms of looks and function. Currently have some 3D printed models that we've tested, but still need to finalize design and get it created out of billet before its ready for its unveiling. So more on that once its ready. Christmas 24' holidays were coming up, had a few weeks where I could focus on my car and get the wiring started. In the run up to that, spent a few more days staring at excel, going over pinouts and finalizing the plan. The holiday started, moved my car into the workshop. Closed the doors, told my family I was in spain with a girl so they would suspect I was doing something normal with my life. Instead I measured up all parts of the loom, and shut myself in the wiring room, huddled around the heater for two weeks. The long process of concentrically twisting the engine bay, engine, interior, rear and roof harnesses had begun. It was very time consuming but I'm happy with the results. All wiring done with M22759 spec wires ranging from 26AWG to 12AWG for various signals/powers/etc. All wrapped in DR25 sleeving with a sprinkle of ATUM/SCL in places. Opted not to label anything as I prefer the stealth look, and I know where everything goes. Some of the twisted sections going for a few meters really pushed my small table to the limits, and tested my patience many times. Very irritating trying to twist looms with 4 meter long wires when you can't spread everything out in a large area. But managed to get the job done, so can forget it happened now. Looms are currently installed on the car for testing and setup, havent finished booting the bulkhead connectors till I'm 100% happy that its all working as intended and I dont need to go in to modify anything. I'll grab some photos of the finished pieces when i remove them to do that. Next I made some mounting brackets for some of the bulkheads and did a final reshuffle of the electronics on their mounting plate. Also removed the X20 E888 expander, as managed to condense all my I/O onto everything else as I found a CAN based weather station module that cut down on the need for an extra 5 independnent analog sensors in the engine bay. Final thing on the wiring list was the ECU's to breakout bulkhead connectors. Went back and forth in my head a few times about whether to twist all these + wrap them same as the rest of the looms. In the end I decided not to as this section of loom is pretty modular and open to me making lots of changes in future compared to everything else that is pretty set in stone now. I still have some doubts about whether I'll keep the M150 ECU or possibly even change the gearbox to a different type of 8HP, so for now all this wiring is remaining like this until I have a better idea on what route to go with next after some testing. I'll probably neaten this up some more though once I've got everything tested. Edited February 12 by Mike2JZ (see edit history) 6 Quote Link to comment Share on other sites More sharing options...
Mike2JZ Posted February 15 Author Share Posted February 15 After installing all the wiring, I went through the process of powering on each loom, checking I didn't have any major issues. Then started a preliminary setup of the power distribution modules to get power to all the places it needed to go based on various keypads/buttons in the car. Started to get excited as starting up the engine seemed closer than ever. All the looms seemed to be behaving and nothing caught fire, so off to a good start. And then disaster! The PDM15 that I had bought second hand a while back was always a bit glitchy, even on my test bench. Technically it worked, but the communications over CANBUS were a bit hit and miss. Once I started setting up the rear of the car and testing the lights/fuel pumps and putting around 100amps through it with everything turned on, it made a strange noise and CAN communications failed entirely. I triple checked my rear loom, made sure I wasnt shorting out on anything, couldnt find anything wrong. Bench tested the PDM and still couldn't get any signs of life from the comms system, so I went home pretty grumpy. Decided I wasn't going to let this get in the way of me starting the engine, so got in contact with Dave @ EPS. He was happy to take the PDM and send it off for inspection at Motec, and in the meantime I bought another PDM15 so I could keep the party going. Depending on what they find is wrong with it and hopefully fix, then I'll either keep as a spare or sell on to recoup some moolah, as a last minute purchase on a new PDM was a bit of kick in the teeeth, but teaches me for buying second hand stuff The next morning the PDM arrived and I was back in business. Plugged it in and immediately worked like it should, phew. So that night it was crunch time. Continued setting up all the electronics and CANBUS communications. You know its getting serious when both laptops are on duty. The list of jobs was long, but after a few hours I had done enough so that all the electronics would do a thing, checked all the inputs/outputs and made sure all sensors had the correct linearlizations to read correctly. Also removed the factory immobilizer from the 8HP gearbox and got Kenneth from CANTCU to flash on a custom tune for a 8HP75 just to get me going and letting the box rev to 8K without trying to upshift. The following morning it was a scramble to get the rest of the mechanical systems all checked/cleaned and primed ready for first start. Ended up taking the entire day as I had a few leaks from coolant system and some other setbacks. But eventually got everything dialed in and with a small audience tried to get the engine to start. Unfortunately it cranked but didnt want to fire the coils or injectors. My brain and body was fried after a whole day or work so I called it there, took some startup logs and triggerscopes and went home. Sat in bed looking over my initial tune on the Motec, I noticed I hadn't set my threshold voltages low enough on the cam sensor for it to get detected when cranking. Cool easy fix, fell asleep like a log and got up super early the next morning. Uploaded the new tune into the ECU, fired the engine and it started like nothing ever happened. But it ran like shit, and wanted to cut out. Scratched my head a bit as to what was causing it as all vitals/sensors looked perfect. The issue was actually a very welcome suprise. The 90mm GM throttle body I chose to use went above and beyond what I was expecting. Normally on a larger throttle body you need less throttle angle at low engine speeds as it will still bypass a decent amount of air in order to maintain idle. However, due to the design of the throttle housing around the blade on the GM throttle you get a lovely deadzone at lower throttle angles that flows less air when throttle is closed. This picture sums it up between a regular throttle and whats going on with this one. So essentially I can get any potential flow benefits from a larger throttle whilst keeping amazing idle quality, just means I have to use larger throttle opening in the tune relative to what I'm used to seeing. So adjusted my idle and throttle flow maps and she purrs like a kitten. Idling at 900rpm at Lambda 1 thanks to the great throttle airflow and ID1050x primaries. 8 Quote Link to comment Share on other sites More sharing options...
Mike2JZ Posted Thursday at 11:29 PM Author Share Posted Thursday at 11:29 PM (edited) So with the car running, it was onto the dyno for some engine run in and some testing and setup of all the new pieces to see how it all reacted. Engine run in mainly consisted of loading up the engine at 150ft.lbs, then doing some low power sweeps to 5000. Varying that for a while, and do-ing some cooldowns and starting the process again. Noticed my oil pressure was a bit lower than i remember and noticed I left my thermostatic sandwich plate blanked off as I didnt want run in oil circulating through the oil cooler which was a bit of an oversight in a rush to get the engine started. The bypass orifice in the sandwich plate to the filer isn't massive, so was restricting flow a bit. Luckily wasnt running the engine super hard, so got away with that. Inspected the oil filter/cam caps/turbo filter and vvti filter and all clean with no debris. Took the thermostat completely off now till it the oil cooler is needed for the real world, changed the oil to 10w60, slapped a new filter on and oil pressure back to normal, healthy 8 bar at 8000rpm. Before ripping it to 8000+, I had to address the ECU firmware as I was still running on the Motec GPA package, just to get the engine started. With 12 injectors, lots of CANBUS data & 8HP integration to worry about I went to a JRR Motorsport Firmware, as it gave me some needed features. Overall the package seems like a step in the right direction, but not sure its going to fufill everything I'm requiring. Also had to track down a weird issue with the cam sync. For whatever reason, compared to the GPA package the cam sync hysterises seemed a lot more sensitive to noise. At the time I wasn't sure if this was down to a mechanical change or a difference between the coding of each package. As a sanity check, I went over the VVTi system and changed the cam sensor to another unit and swapped to a spare VVTi pulley in case the old one was jamming slightly and throwing off the cam position. Luckily with no front engine covers and a barebones approach to the engine, this was a really quick job I could do on the dyno in 15 minutes whilst keeping crank pulley on. If I lost cam sync randomly I could load back in the GPA package and it would fire up perfect and rev out no issue, then swap back to JRR and 50/50 it would drop out using the same settings. In the end had to adjust my hysterises values a bit higher then expected to get rid of any potential noise on the signal, even though the triggerscopes between both packages were the same. Weird, anyway been fine since. Next issue was the CANBUS and the M150 running out of available CPU processing power. Trying to bring in data from my EGT Module, LAM2CAN, CANTCU, PDM's, M1 Trans/Recieve Info, Turbosmart blackbox & Steering wheel controls seemed too much for the CPU to handle concurrently. So had to streamline some of the data incoming/outgoing to keep CPU load around 90%. The new firmware is also limited to how much custom CANBUS addressing can be used, so another limit I should of paid better attention to. At this stage of the development I don't need oversight over every little detail, so could pick and choose the most important information for now. In an ideal world I want all information being logged and referenced from a single point to make the data collection a simple process. My options going forward is to: a) Jiggle around some of my canbus setup, so a bit more things get recieved/transmitted and logged by the C1212 to spread the load, then I'll have to pull logs of ECU & Dash after each run and merge them in I2 for checking (not preffered). b) Get a developer license for the ECU and embark on the quest of streamlining a custom firmware package just for this car and I can get rid of any resource hogging things I'm not requiring and setup the canbus and various functions for how I'd like to see it. I'm actually up for this but comes at a decent price in terms of cost and time invested to get the result I need, so need to see what kind of mood I'm gonna be in as this progresses. c) Bin the M150 and go back to a Syvecs which can just squeeze on all the information via Canbus I require and does everything I need from an 8HP perspective from having done a lot of them for customer builds. Next issue with the Motec is the lack of granularity on some settings that I'm used to with the Syvecs. In fairness it runs the engine absolutely fine and is simple to use so I get the appeal, but feels a bit wattered down/restrictive now I have to worry about the 8HP and modifiers for things like what to do in each gear with idling/fueling/timing and reacting to torque reqests adequately. Some other functions like lambda control/boost control/throttle maps etc do the job, but again lacking some finer control I'm used to having. Annoyingly there is no package that I've found that ticks all the boxes for my requirements, even though the M150 hardware side of things is all there for it. Thats about the only issues I've had on the electronics/ecu side of things. Everything else has gone to plan so far. Got the steering wheel working, so I can choose P/N/D/R, drive modes/shifter modes and paddles all from the steering wheel. The C1212 dash displays all the info & warnings I need to know about whilst on the dyno and PDM's have been providing power just fine. Got a new revision for the keypad roll bar mount. Sits nicely above me near the roof of the car and houses some of the body/engine functions as well as the MSEL driver cut off switch. So with the oil changed, I could get into some power tuning and was a mixed bag of results. Generally happy with performance but need to make a few changes. The EM Raceport electronic blow off valve is sweet. Works just like a normal blow off valve under normal operation, then you can energise the solenoid and stop vacuum/pressure being fed into the top part of the BOV diaphram, essentially opening the gate. My main aim for this was to create a small air leak whilst staging the car for launch control to get the turbo speed a bit higher than normal to continue some momentum in the turbine after the launch. Might have to rethink this slightly as the Raceport does a fantastic job of depleting basically all the boost pressure in the pipe. Can do a run to redline with the BOV open and wastegate completely shut and I might see 0.3 bar boost by redline. I'm not super bothered about having the BOV functioning as a BOV should, so I'm either gonna have to move to a smaller orifice on the pipework to tune the amount of bleed it can do till I get the result I need. Or maybe if I can PWM the solenoid at a high enough frequency, maybe I can adjust the amount the valve lifts at various duty cycles. Need to check with turbosmart if they rate their solenoid for that kind of action. For now I'll leave it as a "normal" BOV and activate it in case of overboost as a safety backstop. Next is the turbosmart 50mm electronic straightgate and blackbox. What a cool piece of tech. So above is a log from a dynorun. Bringing in some of this info about the gate and the control system of the blackbox over CANBUS into the M1, so I can check what its up to and start to understand how its reacting to my commands. Currently I've started with the most basic way of controlling the gate by using a boost PWM signal from the M1 to the Blackbox. This is the same type of signal used to control normal MAC/Pierburg boost solenoids. The blackbox converts the PWM signal, into a target area signal which the blackbox converts into a valve position target as the target area and actual valve position are not on a 1:1 scale as the flow through the gate isn't linear. From there the electronic motor moves to valve into its required position and turbo makes some boost (Or not!). Overall this sytem works really well. I had a few struggles with the JRR M1 package, as ideally I need to send a high side PWM signal to the blackbox, so when 0% duty is requested the valve remains shut and 100% = open. However, I could only send a low side PWM signal so 0% duty = open and 100% = closed. This isnt a massive issue, other than if the blackbox actually recieves 0% duty, it will close the gate, so I have to cap my signal to 0.1%, when requested the gate to stay completely open or it shuts itself which was annoying cause a boost spike. In future I will move away from the PWM signal and send a target position area directly through canbus, so I can avoid the above situation. Although I've got this system working ok at the moment, I need to do a bit more work on it. The gate itself moves quickly and the current draw averages 2-5 amps, spiking to 25amp momentarily if you really snap is hard against the upper/lower limits of its travel. However as it flows differently than a typical poppet style valve took me a while to understand when and how much to open/close it. The PID system controlling the PWM output signal is also very slow/gentle at the moment whilst I was getting my head around it. I think if I make that a bit more aggressive and increase the frequency of changes, then I can get snappier changes on the gate itself, more like how an OEM electronic wastegate is controlled. So bit more work to do there, but the gate does work. Also super cool to open/close to the wastegate whenever you like, even at idle if needed. Look ma im a drag car. Only one small issue with the boost control, which I don't think is actually a gate issue but I'm getting some decent boost creep even with the gate wide open. If I leave the throttle wide open during a pull with the gate completely open, it slowly creeps from no boost to 1.7 bar after 7000rpm to redline. Guessing this is a manifold/wastegate priority issue, but I'll do a run with the gate completely removed to confirm. If it still does it, then will need to modify the design of the manifold routing slightly. Manifold needs to come off at some point to get the gate recirced & wrap the downpipe etc, so can tackle it all then. For now if I shut the throttle to 80% WOT, then I can control the creep, but trade off some engine efficiency/power at the top end which I'd rather avoid. So after playing around at a few different boost levels and dialing in the map, went to start making some decent power runs at 2.0 bar boost Not too shabby so far, but max RPM limit for gearbox was being reached before the power started rolling over completely. Quick conversation with kenneth at CANTCU and he sent me a revision that would let me rev to 8800rpm. ZF rate the max output shaft speed of the gearbox at 8000rpm, so in 6th gear at 1:1 i was exceeding that slightly, but figured a few hundred RPM more wouldnt hurt in a quick burst. However for anyone trying this, do not do this in 7th or 8th gear as you will easily overspeed the output shaft at much lower RPM's. So next run tried to bring it on slightly more progressively but overshot my mark a bit, closed up the VVTi angles at the top end after 7500rpm added a degree of timing and 0.1 bar boost more and she screamed nicely to 8600rpm. Before/After TCU changes Impressive powerband for stock GE VVTi head with some springs and cams on petrol. In reality I'm coming on boost a bit quicker than the dyno shows, but my inductive pickup keeps having a shitfit with the IGN1A coils and inteference, so had to calibrate RPM using wheel speed calibration which works, but I lose some accuracy down low as the wheel speed/gear ratio changes through the run due to tyre deformation. I think for petrol, I won't be running much more than 2 bar boost as its getting a bit close to detonation threshold on top end, so not much point chasing a number. If anything I'm gonna remove a few degrees up top to give me a bit more headroom, and offset that versus whatever gear / enviornment conditions the engine is going through in real world, even if it costs a few ponies. But it's looking good for making 1000whp at 2 bar of boost on E85, so wasnt a million miles off in my hopes. Some data from the sensors I've got hooked up so far. ETS 5'' doing a lovely job of keeping things cool even with 145C going into it. Got around 2-3psi pressure differential through the intercooler, which dosen't upset me that much for the cooling performance. No pressure drop going through the throttle body WOT so that good news too. I need to get my fabrication on 5'' intake pipe sorted, then I'll get it flowed and scale the MAF sensor correctly so I can really track what airflow through the turbo is, but as a rough guide matching up turbospeed and pressure ratio of the turbo gives the above (with a pinch of salt). Looking fairly well sized for the job though, and more room to grow with some more boost pressure in future. Can even see on my fast rate IAT sensor when the compressor wheel starts going closer to the choke point of the compressor map as temperatures rise. So overall decent progress with bringing this machine back to life. Gonna take a week spending some time doing normal people things and seeing the outside world then I'll get back to it after some more development and changes. Should have all my EGT's, lambda's, EMAP/DMAP sensors installed on the next go, so interested to see what thats going to show. Hopefully get some E85 action in as well. Edited Thursday at 11:35 PM by Mike2JZ (see edit history) 7 Quote Link to comment Share on other sites More sharing options...
mattdavies Posted Friday at 08:50 AM Share Posted Friday at 08:50 AM Lovely Update. enjoyed reading that with my morning coffee. Its nice to see a Motec vs Syvecs comparison for a UK Supra, not that I ever considered a Motec. 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.
Note: Your post will require moderator approval before it will be visible.