2013 Honda Fit PCM failure?

A friend contacted me yesterday asking if I knew anyone interested in buying his mom's '13 Fit, which currently runs but with no power. A local shop has determined that the PCM is bad (I'm not 100% convinced of this, unless Honda stuck it under the rain tray like VW does) and the dealer says they've been waiting a year for a PCM for another customer car. If this is indeed the problem, can I buy a PCM on ebay, swap it in and have the keys re-programmed so that she can get back on the road? She loves that car and in working order it's still pretty valuable, I'd like to see her get it fixed properly rather than get hosed on trade-in value and buy a car she doesn't like as much.

I have literally never heard of any problems with Fit electronics aside from sticky A/C clutch relays.  What was it doing or not doing?

In reply to Pete. (l33t FS) :

It's at a Monroe tire shop right now (I don't know why she took it there.) The guy who answered the phone told me it runs but with almost no power, she originally was having a misfire issue which is why she took it in for service. I suspect maybe a rodent got to a wire somewhere.

In reply to dannyp84 :

Or the misfire plugged the converter.

I have had two 2nd gen Fits back the spark plugs out enough to misfire.  One is weird, two is concerning.  I only saw the two, though.

I don't see how a PCM failure can cause a lack of power unless the throttle body driver died.

"Yes. Apparently one of the cylinders was misfiring and they changed everything but it was still showing a misfire.  They checked the computer and it showed the problem was in the computer. They called the dealer and they said they had been waiting for the part for a year for the part. I just wasn’t sure if it would be better to try and trade it or sell it for parts or scrap it. " - from the owner.

I wonder what the "everything" was that they changed, and how the plugs looked. Also the car has 140k on it for reference.

In reply to dannyp84 :

Must be awful rich to be willing to scrap a $10k car just because a muffler shop can't figure out that the valves are tight.

In reply to Pete. (l33t FS) :

Awfully trusting of a general repair shop I'd say, I'm gonna look at it myself though I'm no expert.

In reply to dannyp84 :

I am probably being uncharitable, it is easy to do from hundreds of miles away.  But I doubt it is a phantom misfire caused by a PCM issue, and "low power" without the secondary terms "shuddering" or "misfiring" usually means a failure to open the throttle (which will have codes attached) or a restricted exhaust (which won't, and can be a result of driving with an ignition miss).  Or a secondary issue like bound up brakes or a trans with a bad stator in the torque converter, or is stuck in 3rd gear, or something.

The only times I have seen Honda PCMs go south was on older Civics when a bad alternator would fry the circuitry that controlled the field.

140k is getting close to the point where tight valves cause a misfire.  It usually is only perceptible at mid warmup and only at light load/idle.

An aside, I used to work for a local chain of shops and it was company policy to never adjust valves on anything because of the liability of doing it wrong.  And, I am told, because it would have to be warrantied, and if the valves needed to be adjusted every 15k then we would own it forever.  Made no sense to me because valve adjustment is maintenance and we didn't warranty oil changes, but whatever.  

I'd put money that this is a lack of willingness and/or ability to properly diagnose the problem, not a PCM problem.

So it sounds there may have been a misfire and it probably has gone into limp mode. Among (many) other things, I wonder exactly what all the "everything' they changed included, and what exactly the computer showed them that indicated a PCM failure?

In reply to Driven5 :

Fits do not have limp mode.  They do not cut injectors when misfiring, they are not that sophisticated.  Really I don't recall any Honda ever cutting fuel when it detected a misfire.  They keep giving you all the throttle and VTEC as necessary, and happily dump fuel into the misfiring cylinders, plowing through the problems in a vain hope that the fault is only transitory.  No power to spare after all 

In reply to Pete. (l33t FS) :

Interesting... So you're saying that's what is actually happening when Fit owners experience a severe reduction in power in conjunction with the flashing CEL, and report their cars having gone into limp mode?

In reply to Driven5 :

They are confusing "engine running like crap" with "the computer is making the engine run like crap in order to preserve something" most likely.

A "limp mode" is something like Fords and BMWs cutting an injector or an entire bank of cylinders (in that order) when sufficient misfires are detected, for the duration of that key cycle.  Or GMs pumping fuel trim up to max (ALL THE FUEL!!!) and/or cycling banks of cylinders off (internal air cooling) when coolant temperature gets too hot.

When a Fit starts misfiring, it makes no power because it is a 1.5l engine running on only three cylinders and is dragging a dead one, just like if it had a carb and a points distributor.  The flashing MIL just means that something is happening bad enough to damage the catalyst so you should shut the engine off ASAP.  (ECMs with a proper misfire related limp mode do not flash the MIL when it is engaged, because it is not damaging the converter)

And yes, people saying "limp mode" when they mean "running badly" IS a peeve of mine . So nothing personal.

In reply to Pete. (l33t FS) :

No offense taken. I just thought (see what happens when I do that) limp had been a standard 'feature' on all ECU's for years now. Is it common for a misfire to also go away with the next key cycle, making it that much harder for plebes like me to distinguish whether or not the car actually has a limp mode? Because I'm pretty sure I've seen that reported on Fits too, but my memory could also be hazy on that since I've been Fitless for a few years now.

In reply to Driven5 :

It's... complicated.

In the 1980s, Ford had a somewhat revolutionary concept in the EEC computers, forgetting now if it was the III or IV, called FMEM.  Failure management effects mode, or failure mode effects management, and that is hazy too because it's been close to 20 years since I needed to remember things like our local barometric pressure is 155hz on a Ford baro sensor.  Basically if a sensor input was determined to be faulty, a substitute value would be used.  Because the computers of the time were fairly expensive for processing power, RAM, and ROM, the fault checking criteria and substitute tables were extremely basic, like having a very coarse TPS/RPM table to substitute in if the MAP failed, so the engine would not run "great" but it would still run, with things like fixed timing (easy to do, just stop sending a signal up the SPOUT line to the ignition module) and the like.

Other cars, if you unplugged the MAP sensor or airflow meter, it would just stop running, no value substitution.  Again, what Ford was doing seems pretty basic, but revolutionary in the 1980s.  Ford's methodology would actually insert a value into the computer in such a way that if you were looking at a scan tool, you would see the FMEM-calculated value instead of the actual sensor input (or lack thereof).  If you didn't know that you could chase your tail all day long.

Oddly enough, OBD-II has made this both universal and comprehensive.  Because they by law have to sanity-check the inputs, in most cars you can unplug major sensors and the thing will still run acceptably well.  Unplug the MAF in a GM and the engine might actually run better   GM in particular uses MAP and MAF and alpha-N tables and cross checks everything.  All this cross checking means that they will run quite well when there is an input failure and you wouldn't even know it if the MIL wasn't on.    It isn't a "limp mode" because the computer isn't limiting anything.  Heck I drove a 1.4t Cruze around a bit today with the MAF unplugged and it didn't even turn the light on.  There were a billion pending codes for open circuits, of course, but the computer was perfectly happy with just going off of alpha-N/speed density and it drove very well.  Must have all been two-trip codes.

There is no government requirement for the engines to not hurt themselves, only for the computer to tell the driver to shut the engine off if it IS doing damage: the MIL must flash if there is a "catalyst damaging event".  Mind you, they don't care if you hurt the engine, they just care if you damage the converter, because that will increase emissions in the long-term.  A broken engine decreases emissions because it becomes a statue

Hondas can be amazingly dumb.  Run a K24 low on oil so that VTEC can't engage, and it won't simply run off of the non-VTEC maps.  It will run like complete garbage past where the switchover point should have been.

It's also kind of a good thing, too.  Run the valve lash so tight that the valves hang open at light throttle at a range bracketing around 130-140ish F coolant temperature, it will misfire, and flash the light, but when it gets to full operating temp and the lash opens up again and it has reliable compression again, it will hit on all cylinders again.  It's transitory.

IIRC, Chrysler was the first to kill injectors with a detected misfire, on the LH cars.  Not everyone does it.

Really, when I think of "limp mode" I think of TRANSMISSION control.  A lack of control by the trans computer usually meant that the transmission would hydraulically be in one certain gear, and line pressure would be jacked up to maximum.   Find a solenoid control matrix for your given trans and look for what gear it would be in with all solenoids off - that is the "limp mode" gear.  Usually 2nd, sometimes 3rd.  Low enough that you can move from a stop, high enough that you aren't too much of a traffic hazard.

I can confirm Pete's comments on the flashing CEL vs limp mode thing.  Take 2 of my vehicles, for example.  The Jeep and the E38. 

In the Jeep, if you get a misfire (and the computer bothers to notice) you'll get a CEL with code for a mild miss, but a severe miss (especially under heavy throttle) will trigger a flashing CEL until it stops missing.  No limp mode is programmed in, so it'll keep feeding fuel to that cylinder.  Probably a good thing, as the misfire detection on JTEC PCMs is downright awful.  I've had it throw misfire codes for something so minor I couldn't feel it or detect it at idle with an IR gun.  But I've also seen a dead miss not throw codes unless it happened under heavy throttle, or the most common one, there's a misfire and it threw a code, but for the wrong cylinder. 

In the E38, I've never experienced a mild miss, but with a severe miss (such as when a coil pack decides it's done for the day), you'll get a few seconds of flashing CEL.  If the miss hasn't sorted itself out after something like 2 - 3 seconds, it cuts fuel to that cylinder, turns the CEL solid, and the car has been down-graded to a V7 until you shut it off and restart it.