Jonas Olhssons BMW 1602
News
2007-08-20
We had our first tuning session with the car on the road tonight. As I expected the low load was pretty much impossible to get right with the basic Speed-Density approach and I can't use Alpha-N either because of the fact that this engine uses a turbo. It's the range 2000-4000 rpm that is especially hard. One can tune the VE map to work driving up hill, going down hill the same load site is used which makes it run pig rich. If it wasn't for the CDI ignition box we would've fouled a few plugs.
Any suggestions to a better approach to tune the low load with the feature set we got now ?
2007-08-03
Engine has been started, we had to add the R30 pull-up resistor because it was not mounted (I assembled and built this system myself) but after that the trigger worked. MAP is steady, about 85 kPa @ 1500 rpm idle. As expected this engine needs lambda 0.85 (or at least below 0.90) to idle properly. This engine could _really_ use mappable fuel injection angle to get a reasonable idle and low load quality, or at least development should be pushed towards Alpha-N with pressure compensation.
The engine was only warmed up today and an oil pressure problem appeared that must be resolved before any more is learned.
Engine specification
- BMW M10
- 4 cyl, OHC, 2 valves/cyl
- 90.5 mm Mahle flat-top forged pistons
- 2.0 litre crank (with over-bore 2.05 litres)
- Stock rods
- No squish area left, bad for fuel economy, good for boost
- 7.5 : 1 compression ratio
- Schrick cam shaft (328/316 intake/exhaust degrees duration)
- Individual Throttle Bodies (50 mm)
- Ford Motorsport 1600 cc/min fuel injectors, 5 Ohm impedance
- Run in saturated mode with no power resistors and a 30 V TS diode on the flyback wire
- Stock distributor with heavy duty coil and MSD 6A CDI box
- Custom aluminum trigger wheel on the crank, two metal points for simple trigger using Honeywell GT1 Hall-sensor
- Long tubular turbo exhaust manifold
- Holset HX50 turbo, 63 in/72 out with 22 cm² exhaust housing.
- Large front mount intercooler
- Two piston controlled wastegates (will be replaced with diaphragm based single wastegate for better control)
- I am interested to see how you do with the 1600cc injectors, when you have it running good I'd like to see the injector settings if you wish to share. I have a subpage for injector settings here: MembersPage/KevinBlack/InjectorSpecDatabase
- I will share the settings, but the parameters are _very_ hard to find since the VEMS firmware currently lacks the vital timed fuel injection function. Without that it is not possible to get good enough idle quality to get repeateable and discernable results. We will not focus much time to get good idle quality on this engine due to this.
MembersPage/PhatBob Just a few questions...
When are you injecting? It sounds like you're injecting when the inlet valves are open.
- We have no control over this with the current feature set of the VEMS Genboard. We only have a simple crank trigger, no cam sync. And even if we did have cam sync we would only have four fixed positions to try by rotating the fuel injector output table as there is no programmable fuel injection angle. Eventually we would like to add cam sync to get better fuel injection control, if for nothing else but to get those four fixed positions to try in hopes that we will get better repeatability in the tuning.
- With a multitooth missing trigger (36-1) you would be able to inject sequentially onto a closed valve, and have a modicum of injector start time control.
What PWs do you see at idle?
- I recall 1.5-2.0 ms at 1500-2000 rpm, will check/upload datalogs tonight.
Why are you not running power resistors?
- It's per recommendation from Emil, and it seems to work. At idle the lambda reading reflects changes in the VE table very well.
- Well Emil knows more than I, we'll stick to his recommendation.
What is your req_fuel and VE values at idle?
- I recall 3-4 ms req_fuel and 40-50% VE at idle.
- Try halving the req_fuel and doubling the VE, then try adjusting the injector_rampup value.
Why are you using such huge injectors on such a small NA engine?
- Ooops. Forgot to give enough facts about the engine, it's a low compression turbo engine, see text above :)
- In my experience, 2 bar boost on this type engine would easily require 50% duty cycle on these injectors with regular gasoline running lambda at about 0.80. At that point we are on the right side of 500 whp. We are considering ethanol (E85) and 2.5 bar boost which will require these 1600 cc/min injectors.
- I think in a case like this you're going to have to accept a comprimise.