Index for MembersPage/MarcellGal recent engine project
This car has been a reliable economical daily driver (running on VEMS for years now). It will stay normally aspirated (no turbo: the gains would be too little for the costs). MembersPage/MarcellGal/PowerAndTraction has the role of a turbo project car.
Water injection
We installed a damn-simple cruising water injection system called Aqua-mist that has been used in carburated cars since around 1989. It consists of
- small 2.5L plastic tank
- adjustable restriction (screw with conical end)
- tube to throttle
- this connection is placed so that vacuum appears above appr 8% throttle position (TPS).
- luckily we had a suitable nipple on the Daewoo throttle. We didn't have to drill. We just put a small tube into this nipple, and the engine can draw some water from the tank through a small restriction. Could not be simpler than that.
This system
- does not provide water during idle
- provides very little water during high power (TPS fully on, MAP=101 kPa).
- the water injection itself has no electronic control
- without a check-valve not suitable for a turbo engine
We'll publish results. After adjusting ignition (+3..5 degree advance) and lambdatarget (lambda around 1.08) at the cruising loadsites, we expect
- around 10% fuel consumption improvement. Minimum 5% is almost certain, and 15..20% can only be reached if the comparison point is somewhat suboptimal - far from the detonation point (as often in case of a factory engine).
It's rather just a training for the more complex MembersPage/MarcellGal/PowerAndTraction/WaterInjection
History
After fighting with weather, fuel plumbing issues, unknown ignition module, mysterious GND connections inside the unknown harness, damage due to static electricity, bad sparkplugs, bad injector connection, reverse information about VR sensor polarity, bad information about missing tooth position, missing injector connection again (after small rewiring - this was quite stupid), mechanical problem with hood, weather (continuously), dyno, the engine is tuned, not just VE (that was easy and fast with lambda-target and incredible ego correction) but ignadv as well.
After a fiasco with the fuel filter the official service did not replace (although they charged for it!), the engine was retuned
- [config and tables] (see file dates, mcd-saved.txt and mct-saved.txt are newest)
- With 85% (FWD car) drivetrain efficiency LogAnalysis/PowerCompare result was 167 Nm, almost dead on (factory data is 160 Nm)
- works well with 1.1.x overlapping dwell firmware just earlier 1.1.18, than 1.1.26. Some notes:
- h[1] TDC-reftooth array configured as 30,0,30,0 (60-2 wheel without camsync)
- h[2] set according to new standards: 0x10,0x00,0x10,0x00 ... that is i259 ch0 and ch1
- normal/missing tooth width 24 and 72 (quarterdegrees)
- works with multitooth advanced filter
- standard WBO2 config (normal, amplified nernst circuit, unlike my spare controller).
- MembersPage/MarcellGal/EngineSwap/TaskList - tasks checklist
- MembersPage/MarcellGal/EngineSwap/ToolList - tools checklist
- MembersPage/MarcellGal/EngineSwap/QuestForSpace - cannot fit the turbo
- MembersPage/MarcellGal/EngineSwap/DaewooHarness - old harness
- MembersPage/MarcellGal/EngineSwap/Wiring - wiring and pinout
- MembersPage/MarcellGal/EngineSwap/VrTrigger - 60-2 crank trigger
- MembersPage/MarcellGal/EngineSwap/Ignition
- MembersPage/MarcellGal/EngineSwap/AirFeed
- MembersPage/MarcellGal/EngineSwap/FuelFeed
- MembersPage/MarcellGal/EngineSwap/StarterMotor
- MembersPage/MarcellGal/EngineSwap/ElectronicProblemList
The preparations/brainstorming for this project is below - will be cleaned up some day.
cars
- 2001 year Opel (Vauxhall / GM) Astra 5door with 4cyl 1.6L 4cyl, 8valve X16SZR engine (Delco ECM, central injection, with 4-wire bipolar IdleControl/IdleValveStepper, charcoal canister, and EGR; no power-steering, no ABS), front wheel driven: available for project. I verified that the Delco ECM pinout I got from MembersPage/Fero matched the measurements I made myself before.
- 16 valve engine swapped in
- DROPPED: 1998 year Honda Civic 5door with 4cyl 1.4L engine, front wheel driven, power-steering, ABS (not available for project). The only thing I like in this car is the suspension. I decided I will stay away from modifing Japanese cars.
- MembersPage/MarcellGal/PowerAndTraction Convenient 4WD car. Just walk into a saloon and ask for their best car. They will offer sg for 50k+ EUR. It will turn out that the audi200 is more convenient, has more power, better traction, and better costs. The only thing they can offer is "looks better" (and newer) of course.
All with manual 5 speed transmission and air conditioner.
Constraints:
- I have little time and chance to spend on the mechanical stuff, so I don't intend to invest too much money
- I don't have my own mechanical workshop. See MembersPage/MarcellGal/WorkShop
- I don't have a well defined purpose (racing, or 600 Hp or something). I'm thinking on 1.6 .. 2L engine with turbo and GenBoard/VerThree for a reliable and clean street-car. I also have some AutoTrans or sequential gearbox tweaking in mind, but that could cost much time - postponed for now.
- after I've seen and helped many installs (and considering that I prefer smart solutions to brute force "power-by-displacement") my recent favorite are the 2.3L 5cyl 20VT audi engines in quattro car.
Because of several reasons, the overall project cost and performance seem exceptionally good for these cars (for me)
- MembersPage/NanassyPeter and his friend Karkus with exhaust-shop has much experience (and spare parts) with these engines (winning dragraces)
- no ECU other than VEMS supports the factory triggerwheel (so engines and car with these engines are cheap)
- the audi quattro solution is quite good (and surprisingly affordable)
local stuff
- car-ownership costs are changing, it seems power will be taxed (not displacement). This means NA registered car is cheaper. Hmm, it has a turbocharger ? Howcome did that happen ? ;-)
- Also, an engine above 4L consumes much gasoline in city. I mostly drive in city so 2.3L turbo delivers the same peak performance with lower consumption.
- If I needed >100kW power most of the time (SUV on autobahn), big NA would be better economy.
- But in city, where high power is rarely needed and most of the time the engine outputs 5..10kW, smaller turbo is better
Solved heating issue
After the engineswap a few years back, the AC got broken. Maybe something didn't get power
- relays
- solenoids
- clutch
I got a schematic for it, and some text to describe components (1 sentence each) and translate Italian color names. Most relevant parts:
- the multistage fan control (series resistors)
- the multistage snowflake light (yes, it is dimmed when headlights on - how featurecreep!)
- the CLT valve that allows hot coolant to enter or bypass the cabin heat exchanger
- normally this should not be needed, but it seems they expect some air sneaking through the heater even with the control knob in the coldest position.
- when the temp selector is in coldest state and AC switched on this (apparently the solenoid air-valve allows vacuum to the actuator of the coolant-valve =>), this bypasses coolant to prevent the hot water entering cabin to prevent chance for warmness sneaking in
- a known errata of this valve is that (according to the expert who installed the system) it needs vacuum to keep hot coolant out of cabin, so when the car is started hot, the first few revolutions of the crank (and the coolant pump) pushes hot air in the cabin exchanger, delaying the cool feeling 1..2 mins. Not sure if 1..10 seconds WOT (destroying vacuum effect) will allow hot coolant sneeking in or not.
- Marcell (being lazy to debug the wire-jungle with the Italian-named colors and wiring not matching schematic) allowed direct vacuum to this (directly connecting the 2 vacuum hoses), which fixed the bypass
- cold weather is coming now, and hot coolant (yes, I know from the LCD if and when it is hot) did not enter the cabin even with disconnected vacuum line. Turned out to the bowden actuating the heat/cold air flap came off, the flap in the cold position. Quite trivial, I knew it was one possibility. But earlier some warm sneaked in even with the control-knob in the coldest position => I found this very unlikely until pretty much only this remained. The funny part is that it's relatively easy to check.
- air internal circulation => this was the problem
- also controlled by a solenoid, but MAP vacuum signal from the engine was missing after the engine-swap ;-)
- IMPORTANT: MAP (vacuum) connection for both of the above easy to check - it draws in air I can feel with my finger, and MAP slightly drops when I put my finger there: or idle stepper steps up a few steps if TPS<iac_tps_threshold)
TODO:
- Latest in 2007 April I want to add raiesed idle to the firmware when the AC is on (simple).
- And maybe disconnect AC clutch at max TPS (?)