A standalone ECU spec sheet is written to sell the ECU, not to help you wire your engine. That’s not a criticism — it’s just what a marketing page is for. The headline numbers are accurate, but they’re chosen to look generous, and the details that decide whether the ECU can actually run your engine are usually a line or two further down, in smaller type.
Once you know what to look for, a spec sheet reads fast. This is a field guide to the lines that matter: the driver and output types, not just their counts; the difference between the three kinds of input; the reference and ground budget nobody advertises; and what the CAN spec really buys you. If you’ve already defined your build using the framework in How to Choose a Standalone ECU, this is how you turn a product page into a yes or a no.
Counts are the easy part. Types are the trap.
Every spec sheet leads with counts: so many injector drivers, so many ignition outputs, so many auxiliary outputs, so many inputs. Counts matter, but a count only helps if the channel is the right type for the job. An ignition output that can only send a logic-level trigger is useless for a coil that needs inductive drive, no matter how many of them there are. So read every count with a follow-up question: what kind, and can it do what I’m asking?
Injector drivers: peak-and-hold vs saturated
Injector drivers come in two flavours, and the distinction is about the injector’s impedance.
Saturated drivers switch high-impedance (high-Z) injectors — roughly 8 to 16 ohms — which most modern port injectors are. The driver just turns the injector on and off; the injector’s own resistance limits the current.
Peak-and-hold drivers are for low-impedance (low-Z) injectors — around 1 to 3 ohms — the big, fast injectors common on high-output and older builds. These drivers hit the injector with a high “peak” current to open it fast, then drop to a lower “hold” current to keep it open without cooking it. A saturated-only driver cannot safely run a low-Z injector; you’d need a ballast resistor as a workaround, which is a compromise nobody enjoys.
Many current ECUs use programmable drivers that do both, selectable in software — that’s ideal, because it future-proofs an injector change. What you want to confirm on the spec sheet: how many drivers, whether they’re peak-and-hold capable or saturated-only, and the per-channel current rating. A driver rated well below your injector’s peak current is a driver that’ll eventually let go. If the sheet says “peak and hold” or “programmable current,” you’re covered for most injector choices. If it only says “saturated,” check your injector impedance before you commit.
Ignition outputs: logic-level vs inductive
This is the single most misread line on any spec sheet. Ignition outputs come in two modes:
Logic-level (or “smart coil”) outputs send a low-current 5V trigger signal. They’re designed for coils with a built-in igniter — smart coils — where the coil handles its own high-current switching. LS coils are the textbook example.
Inductive outputs actually switch the coil’s primary current themselves, controlling dwell directly. They’re what you need for dumb coils that have no internal igniter — unless you fit an external igniter module between the ECU and the coils, which converts the job back to a logic-level one.
Plenty of capable ECUs are logic-level only. That’s completely fine if your coils are smart. It’s a problem if your coils are dumb, because you’ll need to budget for an external igniter (a Haltech HPI-8, an AEM or MSD unit, and so on) sized to your cylinder count. Factory Coyote, Barra and many older Toyota and Nissan coils are dumb; LS and most late-model Subaru coils are smart. The spec sheet tells you the ECU’s output mode; your engine’s coil type tells you what mode you need. Match the two before anything else, because getting it wrong changes both your parts list and your loom. The wiring side of all this — which coils drive from which output, and how an igniter sits between them — is covered in Injector and Ignition Wiring.
Inputs: analog, digital, and trigger are three different things
“20 inputs” on a spec sheet is three different resources bundled into one number.
Analog inputs read a variable voltage, usually 0–5V — pressure sensors, temperature sensors, throttle and pedal position, and so on. Some are dedicated temperature inputs with internal pull-up resistors, which is handy because most temp sensors are two-wire thermistors that need one.
Digital inputs read on/off or frequency signals — switches, a vehicle speed signal, a flex-fuel sensor’s frequency output, and — importantly — additional cam-position sensors on multi-cam VVT engines. Digital inputs are usually plentiful.
Trigger inputs are the specialised ones, for the crank sensor and the primary sync cam. Here the sub-type matters: a VR (variable reluctance) input needs a conditioning circuit to clean up the analog waveform, while a Hall input just needs a pull-up and reads a clean digital signal. Some inputs handle both; some are VR-only or Hall-only. Trigger inputs are typically the scarcest resource on the whole ECU — an ECU might have only two or three. The mistake is assuming you need a trigger input for every cam sensor. You don’t: crank and primary sync need dedicated triggers, but extra VVT cam sensors run on ordinary digital inputs. Miscount that and a capable ECU looks disqualified when it isn’t.
The line nobody advertises: 5V references and sensor grounds
Here’s the resource that quietly sinks builds. Every three-wire sensor needs three things: a signal wire into an input, a 5V supply tapped off a reference, and a return to a sensor ground. The spec sheet lists the inputs proudly. It rarely makes a fuss about how many 5V references there are, or how much current each one can source, or how many sensor grounds exist.
This matters because references and grounds are shared, finite resources. Each 5V reference has a current budget; hang too many sensors off one and you drop the reference voltage, which corrupts every reading on it. If a spec sheet mentions “three 5V reference outputs” and their current rating, that’s a good sign the designers thought about fan-out. If it’s silent, assume you’ll need to plan your reference distribution carefully — and be wary of a sensor-heavy build on an ECU with only one or two references. This is exactly the budgeting the I/O matching method walks through, and it’s where a build with “enough inputs” still comes up short.
Half-bridges and drive-by-wire
Generic auxiliary outputs are usually low-side drivers — they switch the ground side of a load, fine for solenoids, relays, VVT and boost control. But some jobs need a half-bridge or full H-bridge, which can drive a load in both directions. A drive-by-wire throttle motor needs a full H-bridge (effectively two half-bridges) and dedicated DBW capability in firmware — not just two spare half-bridge pins. A stepper idle motor and an electronic wastegate also draw on half-bridge outputs. On the spec sheet, look for an explicit “electronic throttle control” or “DBW” capability line, and a count of half-bridge outputs, rather than assuming generic outputs will cover it.
One more thing to recognise here: a few units list high-current power-distribution outputs because a PDM is built into the ECU (the Haltech Nexus R3 and R5, for instance) — outputs rated to drive pumps, fans and lights directly, replacing relays. That’s a different spec-sheet shape from a pure ECU, and it means the box also handles power distribution, not just engine control.
CAN: the expansion story
The CAN spec tells you how far the ECU can grow. Check the number of CAN buses and the supported protocols. CAN is how the ECU talks to a dash, a keypad, a power distribution module (PDM), and — crucially for I/O planning — external modules that add capability: a CAN wideband controller, a CAN flex-fuel module, EGT/thermocouple modules, and I/O expanders. These matter because they let an ECU that’s short on integrated wideband channels, flex inputs or spare pins still run your build — the module adds its own signal-typed I/O over the bus without consuming the ECU’s own analog or trigger pins. An ECU with strong CAN and a good module ecosystem is more flexible than its raw pin count suggests.
A spec-sheet checklist
Run every candidate through these questions:
- Injector drivers: enough of them, peak-and-hold capable (or programmable), and rated for your injector current?
- Ignition outputs: enough, and the right mode (logic-level vs inductive) for your coils — or budget an external igniter?
- Trigger inputs: enough dedicated ones for crank plus sync, and do they support your sensor type (VR or Hall)?
- Analog and digital inputs: enough for your sensor set, with digital inputs for extra cam sensors and frequency signals?
- 5V references and sensor grounds: enough, with current headroom, for a sensor-heavy build?
- Half-bridges / DBW: explicit drive-by-wire capability if you need it?
- CAN: enough buses and a module ecosystem to cover any gaps?
The bottom line
A spec sheet is trustworthy on counts and quiet on types and budgets — so read for the types. The count tells you how many; the type tells you whether it’ll work. Injector driver mode, ignition output mode, the three flavours of input, and the reference budget are where a “16-output” ECU either fits your engine or doesn’t.
Reading spec sheets one tab at a time gets old fast. The StandaloneHQ selector reads them for you — it knows each ECU’s driver types, input mix, reference budget and CAN modules, and checks them against your engine’s actual requirements, so you see fit and shortfall at a glance.
Always verify against the manufacturer’s manual and your tuner before wiring.