ems3110 renault megane 3 rs 180 cv po p and bang

Hello,

A couple of things stand out from the file you attached before anyone can do anything useful with it:

• The name says OBD ... slave ... RS3-maps.bin, and the sampled header, middle and tail regions are all FF with 0.00 bits/byte entropy. That's consistent with a partial / calibration-only dump inside a 2 MB container, not a full flash. The only real content the string scan picked up is Renault part-number data (RD9970 / RD997010, 10169360AA, CARD9970RD99700010197627AA) plus map-looking byte runs.
• A slave OBD read is fine for a normal Stage 1 style edit if your tool writes it back and fixes the checksum, but it is a poor base for hunting overrun maps by hand, because you have no code area to reference.

What pop & bang actually is on EMS3110 (general, not read from your file — those maps are not identified in what was extracted):

• Overrun fuel cut (DFCO) — raise the cut-in RPM / delay or disable the injection cut on decel so fuel keeps going in.
• Ignition retard in the overrun/low-load region — this is what moves combustion into the manifold and makes the noise.
• Throttle / bypass air on overrun — a small opening is needed, otherwise there's no air to burn and you get nothing.
• Lambda control and adaptation disabled/frozen in that zone, otherwise you'll set fuel trim and cat efficiency codes.
• Optionally a hardcut limiter for bang on the limiter, separate job from decel burble.

Also note: the Hardcut Limiter tool in the Resources section here is VAG ME/MED only, it will not touch an EMS3110 file, so don't waste time on it.

Fair warning as well — on a 180 with the stock cat, aggressive overrun fuelling and heavy retard will cook the catalyst and can damage exhaust valves/turbine. Keep it mild if the car is still catted.

To move this forward, please post:

• Exact tool used for the read and the planned write (KESS, MPPS, Bitbox, MagicMotorsport, PCMflash...), and whether it's master or slave
• Whether you can do a full read (or bench/boot) instead of this partial one
• The ECU software/calibration ID from the diag tool, and the original untouched file (not a maps-only export)
• Gearbox (manual or EDC), and current exhaust / cat status
• Whether you already have a DAMOS or a WinOLS project for this SW, or you're starting from a blank auto-search

With a full original read and the SW ID, it's realistic to locate the overrun fuel cut and ignition tables by comparison. With this FF-padded partial, it isn't.

Gemini said:

Hello,

A couple of things stand out from the file you attached before anyone can do anything useful with it:

• The name says OBD ... slave ... RS3-maps.bin, and the sampled header, middle and tail regions are all FF with 0.00 bits/byte entropy. That's consistent with a partial / calibration-only dump inside a 2 MB container, not a full flash. The only real content the string scan picked up is Renault part-number data (RD9970 / RD997010, 10169360AA, CARD9970RD99700010197627AA) plus map-looking byte runs.
• A slave OBD read is fine for a normal Stage 1 style edit if your tool writes it back and fixes the checksum, but it is a poor base for hunting overrun maps by hand, because you have no code area to reference.

What pop & bang actually is on EMS3110 (general, not read from your file — those maps are not identified in what was extracted):

• Overrun fuel cut (DFCO) — raise the cut-in RPM / delay or disable the injection cut on decel so fuel keeps going in.
• Ignition retard in the overrun/low-load region — this is what moves combustion into the manifold and makes the noise.
• Throttle / bypass air on overrun — a small opening is needed, otherwise there's no air to burn and you get nothing.
• Lambda control and adaptation disabled/frozen in that zone, otherwise you'll set fuel trim and cat efficiency codes.
• Optionally a hardcut limiter for bang on the limiter, separate job from decel burble.

Also note: the Hardcut Limiter tool in the Resources section here is VAG ME/MED only, it will not touch an EMS3110 file, so don't waste time on it.

Fair warning as well — on a 180 with the stock cat, aggressive overrun fuelling and heavy retard will cook the catalyst and can damage exhaust valves/turbine. Keep it mild if the car is still catted.

To move this forward, please post:

• Exact tool used for the read and the planned write (KESS, MPPS, Bitbox, MagicMotorsport, PCMflash...), and whether it's master or slave
• Whether you can do a full read (or bench/boot) instead of this partial one
• The ECU software/calibration ID from the diag tool, and the original untouched file (not a maps-only export)
• Gearbox (manual or EDC), and current exhaust / cat status
• Whether you already have a DAMOS or a WinOLS project for this SW, or you're starting from a blank auto-search

With a full original read and the SW ID, it's realistic to locate the overrun fuel cut and ignition tables by comparison. With this FF-padded partial, it isn't.

Click to expand...

file readed with autotuner , file is decrypted so is master file NR (normal read) and the file have stage2 in , so its not the original file.

samperform said:

file readed with autotuner , file is decrypted so is master file NR (normal read) and the file have ...

Click to expand...

Fair enough — then the naming is just misleading (the file is literally called ...OBD_NR.slave RS3-maps.bin, which is why I flagged it). Two things to keep apart though:

What I can actually see vs. what you're telling me
The attachment sampler only reads the header, a middle window (~0xFFF40) and the tail — all three were FF. That doesn't prove the whole file is empty, but it does mean the populated area is a limited window inside the 2 MB container, not a wall-to-wall dump. Open it in a hex editor and tell me the first and last non-FF offsets; that instantly tells us whether you have code + cal, or only the calibration block padded out to 2 MB. On EMS3110 an OBD NR read usually gives you the cal region, while the code/boot area stays FF — which matters here, because overrun/DFCO logic is easier to nail when you can cross-check against code or a proper DAMOS.

For the burble itself, the real question is what map data you have
Being master and decrypted only means you can edit and Autotuner will re-checksum on write. It doesn't give you map positions. So:

• Do you have a DAMOS / A2L / WinOLS project for RD997010 (the CARD9970RD99700010197627AA ident in your file), or are you working blind with map finder?
• Since the file is already Stage 2, get the matching stock RD9970 calibration and run a WinOLS compare. That won't find burble maps for you, but it tells you exactly which regions the previous tuner touched, so you don't stack changes on top of someone else's edits.

The functional blocks you need to hit on this ECU family (general Continental EMS behaviour, not read out of your file):

• Overrun fuel cut re-entry/exit thresholds — you need injection to continue (or resume much earlier) on decel.
• Ignition retard in the low-load / decel zone — this is what actually makes the noise.
• Decel throttle/bypass air — without a small throttle opening there's nothing to burn, this is the part most people skip and then wonder why it's silent.
• Lambda closed loop + adaptation frozen in that zone, plus misfire/cat monitoring, otherwise you'll collect P0420/P030x and long-term trim drift.

Post the non-FF offset range and confirm whether you have a DAMOS or just map finder — and if Autotuner offers you a bench/BSL full read on this ECU, do that one instead. A full read makes this job realistic; a cal-only slice makes it guesswork.