Hi guys first time post here, been doing bits on a customers 240z with an RB26 fitted for a few years now and have found an issue which we think we have traced back to a back feed either in the ECU or the Adaptor block, (or both). The basic issue has been the car has an isolator on the battery (as its usually not driven in the winter and it allows the car to be isolated for that period). The weird thing is we only noticed recently the cars EMU Black ECU will turn on with the turning of the battery isolator, (not ignition key as it should do). This was first noticed when the fuel pumps would prime every time the isolator was turned to the “on” position. I started digging into this and my initial find was an unwanted 7.07V feed to PIN G18, (switched live to the ECU) on Isolator on but with the key out, when in fact that PIN should be at 0V with the key out. So I started looking for any possible back feeds. I found two potential back feeds where the Wideband Lambda and the MAC Valve were taking their switched 12V feeds from PIN G18 on the grey connector of the ECU instead of from independent switched 12v inputs. So I temporarily removed those inputs and the voltage dropped a wee bit but not enough. (I’m guessing the 7.07V at PIN G18 was enough feedback wise to convince the ECU it has ignition power and it was enough to initiate the ECU start-up?). So I carried on looking for other potential causes of feedback as the was still there and didn’t allow the ECU to operate correctly on ignition key up only). Eventually after much probing and testing I discovered a continuity connection between Pin 45 on the OEM Skyline connector, where it connects to the EMU adaptor block and PIN 4 of the OEM Skyline connector where it connects to the EMU adaptor block too. This PIN 4 (aux 5) traced back to the the AAC connector. (AAC is under the inlet plenum on the RB26 as most of you will know). If I unplugged the AAC connector from the AAC Valve the back feed disappeared and the EMU ECU worked correctly, (starting up on ignition key up rather than on 12V battery Isolator power up). We checked the pins and there was some resistance between them rather than being Open Loop so I cleaned them and the resistance improved but didn’t resolved the issue with the connector re connected. (plus it was still a constant 12V feed at the AAC connector rather than the correct 12V switched it should have at PIN 4 on the OEM Skyline connector). So I tried to look for any unwanted continuity in the engine wiring loom, the adaptor block (to the EMU ECU and the OEM Skyline connector), and in the end there was no continuity but there was resistance, (not Open Loop as expected) between the PIN B13 (12V live) and the PIN G12 (Aux 5, through PIN 4 on the Skyline OEM Connector) potentially connecting the Constant 12V live to the AAC Connector which is the cause of the back feed to the switched ignition 12v feed PIN G18 to the ECU even when the ignition key is out fo the barrel. So I’m wondering if there is any way there could be an internal ECU connection/leak between PIN B13 (constant 12V) and G12 (aux 5) or whether it could be corrosion causing the found 93.1kΩ ecu alone and 19.94kΩ (everything connected) leak between them? This also indicates a potential second, separate leak of roughly 20-25kΩ sitting inside the ECUMA02 adaptor housing itself? Any thoughts on this would be massively appreciated as short of sending the EMU BLack ECU off to get checked I’m now at a bit of a stand still until this has been diagnosed? Many thanks in advance for any help as big or small as you can offer. 
Cheers Bobby
Guys this might outline our problems a bit better than the post above, its a bit more detailed and to the point than my previous post, cheers Bobby:-
Subject: EMU Black + ECUMA02 adaptor - stray voltage between B13 and G12/G18 (Black/Grey connectors) - question re: idle valve PWM output behaviour at rest
21 July 2026
Hi,
I’m getting in touch about a fault we’ve spent several days diagnosing on a customer’s Nissan 240Z, running an RB26DETT engine, EMU Black ECU, and an ECUMA02 plug-and-play adaptor (adapting the ECU to the original Nissan Skyline engine loom/connector), purchased via Northampton Motorsport.
The symptom
The ECU behaves as if the ignition is switched on at all times, even with the ignition key fully removed and only the battery isolator switched on. This was first noticed as the fuel pump priming with the key out, and confirmed directly by seeing the ECU’s own software register itself as fully “live” with no key present.
What we found and ruled out first
We traced the symptom to pin G18 (Grey connector, an ignition-sense input) reading a phantom voltage of 6.8-12V depending on load and connector state. Working backwards from there, we found and fixed a genuine fault: corrosion inside a connector on the vehicle’s original wiring loom (the AAC valve connector, located under the intake plenum) was allowing two adjacent circuits to bridge together, feeding a stray voltage into G18 via the adaptor’s pin mapping. Pre-clean resistance at that connector measured 0.573M ohm / 268.7k ohm (asymmetric, direction-dependent - consistent with an oxide/corrosion-type junction). We cleaned this connector, which improved the reading to a symmetric approx 0.55M ohm / 0.57M ohm, but the original symptom didn’t change at all - G18 still read 6.84V - telling us there was a second, separate issue we hadn’t found yet.
Confirmed workaround
With the AAC valve connector unplugged from the AAC unit itself, the ECU behaves exactly as it should - it powers up on battery 12V but does not initiate start-up/ignition-on behaviour until the ignition key switch is actually turned on. This confirms the correct setup and behaviour is achievable, and that the remaining fault is isolated purely to the ECU/adaptor side rather than anything further downstream.
What we found next - and believe is the real root cause
We found a genuine resistive leak of 18.48-18.94k ohm (readings taken both at the AAC connector and again directly at the ECU’s own pins) between:
-
Pin B13 (Black connector) - the ECU’s permanent 12V battery feed, wired directly into the ECU’s Black connector, confirmed not routed through the adaptor or the vehicle’s engine loom at all
-
Pin G12 (Grey connector) - an auxiliary pin electrically tied to G18, on the same connector
We confirmed this is a genuine current path, not just a resistance reading: with the isolator on and a test light probing pin 4 (AAC side, downstream of G12), the test light glowed faintly, consistent with real current flowing through this leak into a high-impedance node.
We then isolated this leak in stages, disconnecting the Black and Grey connectors one at a time, and confirmed it exists right at the ECU/adaptor boundary - not anywhere in the vehicle’s wiring (a full unpowered pin sweep of both connectors against every other pin found no other continuity anywhere). Splitting it further:
-
Grey connector unplugged from the adaptor, Black connector (B13) still plugged into the ECU: resistance between B13 and G12 rose from 18.94k ohm to 93.1k ohm
-
Black connector also then unplugged (Grey still unplugged): resistance went fully open-loop
-
Black connector reconnected (Grey still unplugged): resistance returned to 93.1k ohm
This confirms the 93.1k ohm leak exists internally within the EMU Black ECU itself, between its own B13 and G12 pins - it’s present even with the Grey connector (and therefore the adaptor) completely out of the circuit. The remaining difference between 93.1k ohm (ECU alone) and 18.94k ohm (everything connected) indicates a second, separate leak of roughly 20-25k ohm sitting inside the ECUMA02 adaptor housing itself, between the corresponding pins/traces on its Black and Grey sides.
All of this was confirmed with the vehicle’s wiring completely ruled out at every stage - full unpowered pin sweeps at the ECU’s own Black and Grey connectors, staged connector-by-connector isolation, and direct voltage/resistance measurements taken right at the ECU and adaptor pins themselves.
Update since first isolating this - AUX 5 is our idle valve PWM output
Since first isolating this, we came across a thread on the Ecumaster Community forum (“Aux 5 giving 12 volts”) describing a case where another user’s AUX 5 read 12V even fully disconnected, with a reply from your team explaining that ECU AUX pins have internal pull-ups on unassigned outputs. We initially wondered if the same explanation applied to us.
However, checking our own configuration, we found AUX 5/G12 is not unassigned - it is actively configured as our idle valve’s PWM output (Idle > Parameters: Idle valve type = PWM, Idle PWM output = Aux 5). This means the AAC valve wired to pin 4 is our actual, functioning idle air control valve, not a spare/unused connector.
Given that, we wanted to ask directly: for an idle valve PWM output, is a resting voltage of roughly 11-12V on that pin (with the ignition off, isolator on, engine not running) expected behaviour for the PWM driver stage, or would you consider that abnormal? We noticed idle control was showing as active in the software (Idle Control Active: YES, Idle Valve DC: 5%) even with the ECU in an inactive state and 0 RPM, so we want to make sure we’re not misreading a normal PWM-at-rest characteristic as a fault.
We’ve since checked the EMU Black manual, which confirms AUX outputs are low-side type (switch to ground) and that PWM-driven solenoids like the idle valve require a flyback diode. Given that, we suspect the resting voltage we’re seeing on this pin may simply be normal open-drain/pull-up behaviour when the output isn’t actively grounding, rather than a fault - but we’d appreciate your confirmation, since we can’t be certain of the exact behaviour to expect under these specific conditions (idle control active, engine not running) from the public documentation alone.
The connector-side fault (AAC valve connector corrosion, described above) is confirmed and we are proceeding with a full repair/replacement there regardless. What we’re trying to establish is whether the internal leak we measured at the ECU’s own pins (approximately 93.1k ohm between B13 and G12, isolated with the adaptor fully disconnected) is consistent with normal idle valve PWM output behaviour, or whether it points to something we should be concerned about.
Separately, the approximately 20-25k ohm leak we isolated as specific to the ECUMA02 adaptor itself (only present once the adaptor is connected into the circuit) is still outstanding and we’d appreciate your thoughts on that directly, since we understand the adaptor is one of your own products.
One more wiring point we’d like to confirm
Earlier in this diagnostic (before the faults described above), we found the wideband lambda sensor heater wire and the Mac valve wire had both been incorrectly landed on pin G18 instead of their own supplies, and corrected this by giving each its own separate, independent switched 12V feed. We wanted to confirm with you directly that this is the correct wiring practice - i.e. that the lambda heater and Mac valve should each run from their own independent switched 12V supplies, entirely separate from pin G18, rather than sharing a feed with it or with each other. Just want to make sure we’ve got this right given everything else we’ve found.
What we think this means
We’re not yet certain whether the ECU itself requires attention - that depends on your answer regarding expected PWM-at-rest behaviour. The AAC connector corrosion and the adaptor-side leak are both confirmed as genuine issues on our end, which we are addressing/reporting separately.
Happy to share further readings, photos of our test setup, or the ECU’s serial number/firmware version if useful.
Thanks, Bobby (Zealou5)