john banks

Recently bought my first diesel. My ECU would need bench work before it can have OBD access, and I would struggle to trust whether many or even most diesel tuning are really tuning or just file copying. Also have DPF. If I could source the connectors it would not be too hard to make my own two channel box, but equally it would not be a good use of my time just for my own car.

If a diesel tuning box alters the common rail pressure sensor signal, will the ECU attempt to lean the richer mixture by shortening the injector times via lambda feedback or would it increase the boost or just ignore? If you also alter the MAP sensor signal to increase boost it could mitigate this, but won't the MAF sensor show higher than target air mass?

Can VDCS show you lambda trims on a VAG diesel?

With a DPF, do you just watch EGT and lambda to tell if you are overfuelling? Unlike a petrol, EGT increases with richer fuelling? What sort of max torque or max power lambda are you looking for, or is it smoke and therefore limited by EGT and smoke/DPF loading?

RPM sensing by tuning boxes, do they look at pulsations on the rail pressure or is it a simple lookup based on pressure in which case why not just have a simple pressure in vs pressure out curve? Anyone have any sample values for rail pressure at various loads/engine speeds?

ALi-B

iTrader: (1)

The tuning boxes I've worked with work on two principals:

Single channel via altering the fuel rail pressure sensor values. By reading the rail pressure and sending an altered value to the engine ECU in real time. As far as I know the ECU has limited closed loop feed back on oxygen content in the exhaust as the oxygen sensor degrades through use, on modern BMWs (N47 etc) the oxygen sensor is mainly used for EGR monitoring (where the engine is restricted of oxygen) and DPF regeneration (where fuel is burnt in the exhaust after main combustion), however things may have changed on new cars - last time I refreshed on this was in 2011.

The better tuningboxes will retain factory idle/low speed fuel pressure, but ramp it up during acelaration and then peak out at the rail's peak pressure or slightly higher, if the rail pressure exceeds a set value it can "go back" to the actual measured pressure (fail safe). There are limits to this as the ECU generally knows what fuel pressure it should see at a given rpm, so if the values it is made to see by the tuning box are too low it will log low fuel pressure codes. This is why when the target rail pressure is reached you need to make sure the ECU thinks its seeing a normal fuel pressure. The same logic applies with idle fuel pressure, and even cranking fuel pressure; if the ECU thinks its too low the ECU won't even attempt to inject fuel.

Then we have two channel tuning boxes, which works the same on fuel pressure, but also alters the airflow sensor or MAP readings in the same manner (factory idle, ramp up rate, peak rate, go back level, etc). These sometimes can conflict with MAP sensors (if it alters MAF values) and interfere with EGR regulation as well as boost control during DPF regen (as the ECU uses the intake throttle to control boost, as the turbo vanes are held fully open during regen). But there is a window of error in which these boxes generally work in and take advantage of... however, step outside that window and the ECU starts logging fault codes.

I haven't seen a tuning box that takes a rpm signal. Just fuel pressure and airflow/airpressure. Although I have seen older units that alter fuel temperature (on PD engines IIRC)

The one main argument with raising fuel pressure is more fuel is injected at the start of the combustion phase. So even with shortened injection durations, there is a claimed advantage of rasing fuel pressures. Where that fits in with the "pilot, main and post" injection stages probably needs a bit more clarification...you don't want to inject too much fuel at the pre-injection stage as that would make the engine noisy. And too much post injection is just wasteful (unless you want to regen the DPF )

One hinderance of tuning with DPFs is making sure the engine doesn't smoke too much under acelaration, especially as the turbo spools (low boost therefore lack of air, but increasing fueling at this stage reduces lag) and at peak rpm. The problem is you can't see the smoke.

VCDS will show alot of things but critically it depends on if it has the correct "label file" for your car, otherwise many of the measuring blocks are just meaningless numbers.

john banks

Thanks. It seems that a 5V dual channel box would be easily made with some PICs which have built in DAC. Not found any open source ones yet.

ALi-B

iTrader: (1)

Its certainly possible.

The tuning boxes I've worked with (DPT) are Atmel based (AVR 32 I think), obviously they are pre-programed so the core program is already on there upon purchase, however the settings can be adjusted using telnet via a serial to usb adapater - saves the need for using specific software to adjust it.

steve rally

I have tried both tuning boxes and reflash and would recommend reflash as long as it is customised to the vehicle. I have an Audi A4 3.0 Tdi (2012) and have worked with Zi Motorsport to develop a "good" map using our DD Rolling road. We are at 297BHP and 495ft lb with just DPF deleted.Mpg is 47-52 on motorway runs with 42-45 on urban runs.First flash requires ECU removal but subsequent flashes are via OBD.Cost is £250.

john banks

Does the initial bench flash require chip desoldering or just plugging/probing?

It is an A8 3.0 TDI (2013) with 250PS and 550Nm factory.

steve rally

Just plugging/probing.
"Before" run was showing 245BHP and 410ft lb - basically as factory...
Concentrated on improving low rpm boost and fuelling mapping.

ALi-B

iTrader: (1)

Obviously you have a DPF delete, so that's not an issue, but i do wonder how mappers get around mapping cars with the DPF retained; As there is no real tell tale that the engine is producing excessive soot.

john banks

That is one of my concerns too, perhaps you could datalog the differential pressure sensor, but would one or two dyno pulls appreciably alter the value and then it wouldn't tell you which areas of the map are smoky? Presumably a safe AFR of say 18 or 20:1 might help, I know there is a good correlation between CO and smoke number on my oil boiler, but CO2 measurement which is more equivalent to lambda can gently get richer and then CO and smoke suddenly decompensate, also a bad air day can put a well tuned boiler into soot when air density is lower. On the GTR you can have a lovely lambda and still kick out smoke with poor injection timing. Don't really want to remove DPF, even stealthily.