Symptom
Power comes back after switching off and restarting
The car loses power or drops into a restricted mode, and a stop-start cycle returns it to normal — sometimes for weeks. That single fact is one of the most useful things a driver can report, because it says the module withdrew power deliberately and then let it back: a mechanical failure does not repair itself in a car park. What it points at is a reading that went briefly out of range rather than a part that has died, which is why the fault is so often gone by the time anybody looks and why the stored code matters more than the road test. Note what the car was doing at the moment it happened — uphill, hot, wet, accelerating — because that is the half no code records.
Also described as: limp mode clears when i restart, power returns after restart, goes into limp mode then recovers, restart fixes it, fault clears when i turn it off and on, با خاموش و روشن کردن درست میشود, بعد از ریست عادی میشود, se arregla al apagar y arrancar, vuelve potencia al reiniciar, apagando y encendiendo se soluciona, ترجع طبيعية بعد الإطفاء والتشغيل, تتحسن بعد إعادة التشغيل, أطفيها وأشغلها فتضبط
What this usually means
This is one of the most useful things a driver can report, and it is worth saying so to them, because most people mention it apologetically as though it undermines the complaint. It does the opposite. A mechanical failure does not repair itself while the car sits in a car park. A limit that lifts on an ignition cycle was imposed by a module and released by a module, which rules out most of what an inspection would otherwise chase.
The mechanism is that protective limits latch. Once the management has decided to withdraw power, it generally holds that decision for the rest of the driving cycle rather than restoring power the instant the reading returns to normal — because a condition that goes in and out of range is exactly the kind it should not be optimistic about. So the driver experiences a cliff, a period of restricted running, and a complete recovery after a restart.
What this points at is a reading that went briefly out of range, not a part that has died — and that is why the fault is so often gone by the time anybody looks at it. The stored code and its freeze-frame data are worth more than any road test, because the frame records what the engine was actually doing at the moment the limit was applied: load, engine speed, temperatures, boost, the lot.
The half no code records is what the car was doing, and only the driver has it. Uphill, fully loaded, in the wet, at motorway speed, twenty minutes into a run on a hot day — an intermittent that only appears under sustained high load is a different fault from one that appears in traffic, and no amount of workshop testing recovers that information. Write it down at the time.
Working through it
| What you find | Likely cause | How to confirm it | What fixes it |
|---|---|---|---|
| Power drops suddenly, restart restores it fully, sometimes for weeks | A latched protective limit released by the ignition cycle — an intermittent signal or a brief excursion out of range | Read stored and pending codes with their freeze-frame data. Do not clear until the frame has been recorded — it is usually the only evidence that exists. | Diagnose the named parameter under the conditions the frame shows, rather than at idle in a workshop. |
| Happens on long climbs or when towing, never in ordinary driving | A boost or temperature excursion under sustained high load — the management protecting against overspeed or overheating | Log live boost, requested against actual, and the relevant temperatures on a road test that reproduces the load. Static testing finds nothing here. | Repair what the log names — commonly a charge leak, a sticking vane mechanism or a cooling shortfall. |
| Happens when hot, never from cold | A sensor or connection whose signal drops out at temperature — resistance rises, or a connector expands out of contact | Reproduce hot and watch the suspect signal live. A cold bench test proves nothing about this population. | Renew the sensor, or repair the connector. Warm the suspect area gently to reproduce it if a road test will not. |
| Happens in heavy rain or after driving through standing water | Moisture in a connector or module briefly corrupting a signal | Inspect connectors in the affected circuit for water and green corrosion, and check for grommets and drains that have failed. Reproduce with a controlled water spray if necessary. | Dry, clean and seal the connection, and repair whatever let the water in. A dielectric grease repair without fixing the ingress path is temporary. |
| Diesel, and it happens near the end of a long run | A particulate filter regeneration inhibited or interrupted, so the filter reaches a load at which power is restricted | Read soot and ash load and the regeneration history: repeated aborted regenerations are the pattern, and the ash figure is the one that does not fall. | Complete a forced regeneration and address why they were aborting. Replace where ash rather than soot is the load. |
| No codes at all, and it still happens | A limit applied without setting a fault, or a fault that has aged out of memory | Log live data on a drive that reproduces it, watching torque limitation parameters directly. Some limits are conditions rather than faults and store nothing. | Act on the logged parameter. An absence of codes is not an absence of a fault, and this is the symptom where that is most often true. |
-
Power drops suddenly, restart restores it fully, sometimes for weeks
- Likely cause
- A latched protective limit released by the ignition cycle — an intermittent signal or a brief excursion out of range
- How to confirm it
- Read stored and pending codes with their freeze-frame data. Do not clear until the frame has been recorded — it is usually the only evidence that exists.
- What fixes it
- Diagnose the named parameter under the conditions the frame shows, rather than at idle in a workshop.
-
Happens on long climbs or when towing, never in ordinary driving
- Likely cause
- A boost or temperature excursion under sustained high load — the management protecting against overspeed or overheating
- How to confirm it
- Log live boost, requested against actual, and the relevant temperatures on a road test that reproduces the load. Static testing finds nothing here.
- What fixes it
- Repair what the log names — commonly a charge leak, a sticking vane mechanism or a cooling shortfall.
-
Happens when hot, never from cold
- Likely cause
- A sensor or connection whose signal drops out at temperature — resistance rises, or a connector expands out of contact
- How to confirm it
- Reproduce hot and watch the suspect signal live. A cold bench test proves nothing about this population.
- What fixes it
- Renew the sensor, or repair the connector. Warm the suspect area gently to reproduce it if a road test will not.
-
Happens in heavy rain or after driving through standing water
- Likely cause
- Moisture in a connector or module briefly corrupting a signal
- How to confirm it
- Inspect connectors in the affected circuit for water and green corrosion, and check for grommets and drains that have failed. Reproduce with a controlled water spray if necessary.
- What fixes it
- Dry, clean and seal the connection, and repair whatever let the water in. A dielectric grease repair without fixing the ingress path is temporary.
-
Diesel, and it happens near the end of a long run
- Likely cause
- A particulate filter regeneration inhibited or interrupted, so the filter reaches a load at which power is restricted
- How to confirm it
- Read soot and ash load and the regeneration history: repeated aborted regenerations are the pattern, and the ash figure is the one that does not fall.
- What fixes it
- Complete a forced regeneration and address why they were aborting. Replace where ash rather than soot is the load.
-
No codes at all, and it still happens
- Likely cause
- A limit applied without setting a fault, or a fault that has aged out of memory
- How to confirm it
- Log live data on a drive that reproduces it, watching torque limitation parameters directly. Some limits are conditions rather than faults and store nothing.
- What fixes it
- Act on the logged parameter. An absence of codes is not an absence of a fault, and this is the symptom where that is most often true.
Codes that produce this symptom (16)
Listed with the codes that most typically present this way first — the likeliest place to start.
- P2544 Torque Management Request Input Signal A Typical of this code High
- P2545 Torque Management Request Input Signal A Range/Performance Typical of this code High
- P2546 Torque Management Request Input Signal A Circuit Low Typical of this code High
- P2547 Torque Management Request Input Signal A Circuit High Typical of this code Moderate
- P2548 Torque Management Request Input Signal B Typical of this code High
- P2549 Torque Management Request Input Signal B Range/Performance Typical of this code High
- P2550 Torque Management Request Input Signal B Circuit Low Typical of this code High
- P2551 Torque Management Request Input Signal B Circuit High Typical of this code Moderate
- P2637 Torque Management Feedback Signal A Typical of this code Moderate
- P2638 Torque Management Feedback Signal A Range/Performance Typical of this code Moderate
- P2639 Torque Management Feedback Signal A Circuit Low Typical of this code Moderate
- P2640 Torque Management Feedback Signal A Circuit High Typical of this code Moderate
- P2641 Torque Management Feedback Signal B Typical of this code Moderate
- P2642 Torque Management Feedback Signal B Range/Performance Typical of this code Moderate
- P2643 Torque Management Feedback Signal B Circuit Low Typical of this code Moderate
- P2644 Torque Management Feedback Signal B Circuit High Typical of this code Moderate
Information, not instruction
CarGeek is a reference, not a service manual. Its pages are written from public sources and from what other people have found, never from a manufacturer's own service information, so they can be wrong or incomplete for your particular vehicle. Some of what they describe carries real risk: high voltage, fuel under pressure, stored energy, hot and moving parts.
Have anything you intend to act on checked by a trained person against the manufacturer's own procedure for your vehicle. Nothing here replaces that procedure.
Others with this symptom
Be the first to write this up
Nothing posted about this yet. Whatever you found is worth writing down — even a dead end saves someone the same afternoon.
What makes one worth reading:
- The car — make, model, year and engine. The same fault behaves differently on each.
- What you measured, not only what you replaced.
- What actually fixed it — or that nothing did, which is worth knowing too.
Sign in or create an account to add what you found about Power comes back after switching off and restarting.