209 Stafford St, Worcester, MA 01603Get directionsMon–Fri 8AM–5PM · Sat 8AM–2PM
EV & High Voltage

Tesla Model 3, Dead High Voltage: Why We Inspected Before Replacing the Pyro Fuse

A Model 3 arrived on a flatbed with no high voltage and a blown pyro fuse. The obvious repair was the fuse. We inspected the PCS and the high-voltage harness first — and found the rear drive unit cable worn through against the body.

Dark blue Tesla Model 3 strapped down on a flatbed trailer at dusk, delivered to Iron Wolf Motors in Worcester, MA after a complete high-voltage shutdown

How it arrived

Delivered on a flatbed. No drive, no high voltage, and the car would not recognise its key.

Reported symptoms

  • Vehicle would not drive and could not be woken normally
  • Complete loss of high-voltage system — contactors would not close
  • Car did not detect the key
  • Pyro fuse had fired, with the pyro disconnect panel recommending replacement

Outcome

The chafed rear drive unit cable was replaced with a new original Tesla part and the pyro fuse renewed once the short was gone. The pyro fuse diagnostic passed, the contactors closed normally, and the car drives without fault and sees its key again — fixed once, not twice. The repair carries our 12-month / 12,000-mile warranty on parts and labour, whichever comes first.

Read from the factory tooling

The codes this car was reporting

HVP_w021_PyroMiaFault
The vehicle detects a condition affecting one or more of the pyrotechnic (pyro) fuses. Potential impact: the HV contactors are not allowed to close. Set condition: the diagnostic test on one or more pyro fuses is unsuccessful, which can indicate the fuse has blown.
PCS_a043_hvBusPrechargeFailure
PCS DCDC failed to precharge the HV bus to the target voltage. Logged repeatedly. Payload showed hvBusV 4.54 V against an hvPrechargeTargetV of 373.83 V, with dcdcLvBusV at 11.80 V — the HV bus was effectively dead while the 12V side was alive.
BMS_a081_SW_Ctr_Close_Blocked
Battery management blocked the contactors from closing.
BMS_a083_SW_Ctr_Close_Failure
Contactor close attempt failed.
BMS_a071_SW_SM_TransCon_Not_Met
State-machine transition conditions for going to contactors-closed were never met.
GTW_w117_unexpectedDestructiveResetStatus
The gateway lost its power supply abruptly rather than shutting down in an orderly way. A “destructive” reset is exactly that — power removed mid-operation, not a controlled restart.

This one is worth writing up because of the decision in the middle of it. The obvious repair was sitting right there on the screen, clearly labelled, and we deliberately did not start with it — and that is the only reason the car was fixed once instead of twice.

How it arrived

The car came in on a flatbed. No drive, no high voltage, and it would not recognise its key — which on a Tesla is often a symptom of the low-voltage side being confused rather than a key problem in its own right.

A car that arrives this way is a good reminder of something we cover in our independent Tesla service guide: the high-voltage pack can be sitting at a comfortable state of charge and the car can still be completely immobile. This one showed 66.3% on the pack and was not going anywhere.

What the data added up to

Connected on Tesla Toolbox 3, the picture was consistent and it pointed in one direction. The codes are listed in full above; the useful part is what they mean together.

Read as a set, they describe a car whose 12V side was perfectly alive and whose high-voltage side was simply not there — a bus sitting at a couple of volts against a target near 374, battery management refusing to close the contactors, and a gateway that had been reset abruptly. That is not a sensor disagreeing or a module misbehaving. That is an open circuit.

Which is precisely what a pyro fuse produces when it fires. So the tooling was not wrong, and it was not being vague: the fuse had gone, and the car said so clearly. The only question left was why.

The decision: inspect first, replace second

The tooling was pointing at the pyro fuse and recommending replacement. On a lot of cars, in a lot of shops, that is where the job would have started and ended: order the part, fit it, hand the keys back.

We did not do that, for one reason. A pyro fuse does not blow on its own. It is a one-shot safety device with no wear mechanism — it fires because the car detected something severe. We have seen enough of these to treat a fired pyro fuse as a question rather than an answer.

So before fitting a new one, the car went on the lift for a full inspection of the power conversion system and the entire high-voltage harness — every run, every connector, every point where a cable passes a bracket or an edge, front to back.

What the inspection found

The fault was a high-voltage cable from the rear drive unit that was contacting the body. Movement and vibration over time had worn through the insulation at the contact point until the conductor was finding the chassis. That is a dead short to ground on a several-hundred-volt circuit, and it is precisely the condition a pyro fuse exists to interrupt.

It also explains everything else on the list — including the symptom that looks unrelated. On a Tesla the immobiliser function is handled by the rear drive unit. With that unit off the CAN bus and not communicating, there was nothing for the car to authenticate the key against, so it simply did not see it. The owner experienced that as a key fault. It was never a key fault.

The gateway’s destructive reset fits the same picture: power being removed mid-operation rather than the module shutting down in an orderly way. Every entry in that code list traces back to one worn-through cable.

The repair

The damaged cable was replaced with a new original Tesla part — not a repaired section, not spliced. On a high-voltage run carrying that kind of current, a splice is a future failure and a safety question, and it is not something we are prepared to hand back to a customer.

With the short gone, the pyro fuse diagnostic passed, the contactors were allowed to close, the HV bus precharged to target, and the car came back properly — including seeing its key again. Full re-scan afterwards to confirm nothing was left stored.

What to take from this

If you own an EV and you are told a pyro fuse needs replacing, the question worth asking is not “how much for the fuse.” It is “what made it fire?” A shop that cannot answer that has not finished the diagnosis, and you may well be paying for the same part twice.

There is nothing clever about the fix itself — a cable was worn through and it was replaced. What decided this job was the order the work was done in, and that came from having seen enough of these to know that the part the computer names is not always the part that failed.

And if you are looking at a used EV, this is one of the reasons a proper inspection matters more than a clean dashboard — a chafe point that has not worn through yet looks like nothing at all until it does. Our EV battery health check guide covers the pack side of that inspection.

From the job

What we photographed

Dark blue Tesla Model 3 strapped to a flatbed trailer at dusk, arriving at Iron Wolf Motors in Worcester, MA with a total loss of high voltage
How it arrived. Plenty of charge in the pack, and no way to move under its own power.
Gloved hand holding the removed pyrotechnic fuse assembly from a Tesla Model 3, with the opened high-voltage compartment and orange HV cabling visible behind it
The spent pyro fuse. A one-shot device with no wear mechanism — which is exactly why it was treated as a question, not an answer.
Tesla Toolbox service mode HV battery screen showing HVP_w021_PyroMiaFault, a Pyro Disconnect replacement recommendation and a squib resistance reading of 10.671 ohms
The tooling naming the pyro fuse and recommending replacement — the obvious repair, and the one we did not start with.
Tesla service mode alert detail for PCS_a043_hvBusPrechargeFailure showing hvBusV of 4.54 volts against an hvPrechargeTargetV of 373.83 volts
The precharge alert, logged over and over. A high-voltage bus reading near zero against a target near 374 volts is an open circuit, not a glitch.
Tesla service mode active alerts list showing BMS_a081_SW_Ctr_Close_Blocked, BMS_a071_SW_SM_TransCon_Not_Met, BMS_a083_SW_Ctr_Close_Failure and GTW_w117_unexpectedDestructiveResetStatus
The rest of the car agreeing: contactors blocked, and a gateway reset consistent with power disappearing abruptly.
High-voltage penthouse area of a Tesla Model 3 with orange HV cabling and connectors exposed, Tesla diagnostic breakout leads attached and a multimeter probing the circuit during pyro fuse testing
Measuring at the high-voltage penthouse with the diagnostic breakout leads fitted. This is the stage where a code list stops helping and the meter starts.
FAQ

Questions this job raises

Why not just replace the pyro fuse the tooling asked for?

Because a pyro fuse is a one-shot safety device with no wear mechanism — it does not blow on its own, it fires because the car detected something severe. We inspected the PCS and the full high-voltage harness first and found a rear drive unit cable worn through against the body and still touching. A new fuse fitted into that car would have fired again straight away.

Can a blown pyro fuse just be a part failure?

Very rarely. It is not a wear item and it does not blow with age. It fires because the car detected a crash or a high-voltage fault. If a shop replaces one without explaining what made it fire, the diagnosis is not finished.

Why could the car not detect its key?

On a Tesla the immobiliser function is handled by the rear drive unit. The damaged cable meant that unit was not communicating on the CAN bus, so there was nothing available to authenticate the key against and the car did not see it. It presented as a key problem and was never a key problem — once the cable was replaced, normal key detection returned with no separate work.

Why replace the whole cable instead of repairing the damaged section?

On a high-voltage run carrying that current, a splice is both a future failure point and a safety question. We fitted a new original Tesla cable rather than repairing the worn section.

Would a generic OBD scanner have found this?

No. Every code on the car described a consequence — precharge failure, contactors blocked, pyro fuse test failed. None of them could describe a chafed cable, because no module can see one. Finding it required Tesla Toolbox 3 to read the real fault data and then a physical inspection of the entire high-voltage harness on a lift.