P3190 Diagnosis

toyota-prius-p0a0f-and-p3190-no-throttle-response-solved

 

P3190 means the engine was caught at some point producing less than a fifth of the power it was expected to.

That counts as running pretty badly, though maybe just in the specific conditions when the code was set.

The freeze frame for the P3190 code will help identify what the conditions were when it was set.

Sometimes the diagnosis is helped by trying to focus on those conditions when test-driving.

 

Well, focusing on any one thing like the MAF sensor right now is probably premature. P3190 is a very broad "engine ain't producing enough power" code, with a very broad range of possible causes that have to be sort of methodically checked out.

The freeze frame is useful for seeing what kind of conditions seem to be the ones where there's a problem. The ECM had only clocked around 2½ minutes on this drive and the engine wasn't fully warmed up yet (124℉, request warm-up active). You were driving around 26 or 27 MPH at a low calculated load under 14%. The engine was turning 1100 to 1200 RPM.

The car was hoping for about 24 kW out of the engine, or so your attachment says. Here's where Toyota labeled something wrong in their data list PID descriptions:

Either the label on that PID, "requested engine torque", or the unit, kW, is wrong. You don't measure torque in kilowatts. Either they really meant torque and the unit should be Nm, or they really meant kW and the label should say "requested engine power".

The car was also requesting a target engine speed of 1300 RPM.

So if they really meant 24 Nm of torque, then at 1300 RPM that'd be about 3¼ kW, a nice light load near what it uses to charge the battery. Somewhere around 6 percent of the engine's max rated power.

If they really meant 24 kW of power, then at 1300 RPM that'd be around 176 Nm of torque, which is unrealistic for this engine. So I'm going with "they meant torque, and it's newton meters."

But either way, there's a definite problem, because the actual engine torque is negative. MG1 is still having to spin it to keep it going. The car has given the engine 6.1 seconds to do better than that, but no luck.

A non-zero misfire count is being shown for cylinder 2, but I'm not sure how much the misfire monitor data can be trusted at this stage of engine operation. The zero ignition triggers, zero misfire load and -100 misfire margins look to me like defaults; they probably don't start getting computed until the ECM believes the engine's actually running.

So the above is probably most useful, at this stage, for understanding the conditions where the engine is having trouble. Seems to be at light load early in the warmup.

 

If you haven't done so yet, I would suggest cleaning the throttle body bore where the blade is in closed position.

It can be difficult to figure a direction of cause vs effect from a freeze frame capture.

The data suggests that the ecu wanted the engine to run, but it wasn't really running. Both the A/F and O2 were reporting completely lean (like during decel fuel cut). Not sure why it didn't throw a P3193.

I would look closely at the MAF sensor- while the engine is running, try wiggling the connector and wires. See if the engine stumbles or cuts off. Recording data while driving might be helpful. If the MAF glitches while the ICE is running, then fuel trims will go high positive.

Another possibility is the fuel pump is "falling down". Might need to plumb a fuel pressure gauge in then go for a drive (PIA to do but sometimes that's what it takes).

Posted via the PriusChat mobile app.

 

Right. What it does in Park is nothing but a special case in the firmware for "I've got no reason for running the engine, but the human wants me to for some reason, so I'll rev a little bit." I think it's got at least two RPMs programmed in, one for if you press the pedal some moderate way down, and one slightly higher for if you press it near full.

 

So, if you want to check the ECT, just leave the car overnight. In the morning, turn the ignition ON - not ready (power x2, foot not on brake). Connect your scan tool and go into ECM data. ECT reading should be close to the IAT, and both should be about ambient.

Go to ready, the coolant storage tank and pump should do their thing (ECT should go up 80-100°). Then as you drive the car ECT should increase up to 180°F then plateau as the thermostat opens. Eventually it can go up to around 205° when the radiator fan should turn on.

You can "check" many things at once by monitoring fuel trims whole driving the car. The ECM receives data from sensors such as MAF, TPS, crank and cam, ECT, etc. It calculates the "correct" outputs for the injectors and ignition coils.

Then the ECM uses the Air - Fuel sensor to determine if the fuel mixture in the exhaust gases is where it should be. If it's "lean", the ECM uses Short Term Fuel Trim to add fuel (+ correction) to injector on time. If it's "rich", STFT removes fuel (- correction).

When everything works "just like it left the factory", fuel trim is near 0%. Should STFT average more than a few points away from 0, Long Term Fuel Trim makes a correction then keeps that stored in ECM memory.

STFT + LTFT = total Fuel Trim. Ideally total FT should be 0 +/- 5%. Up to 10% is "OK". 15% means something's not quite right but it can compensate. 20% or more is where you set codes.

There are FT values for different combinations of RPM and load. Idle is not the same as accel or cruise. Seeing how FT changes can help pinpoint problem areas. (example, high + FT only at idle but near 0 under load is often caused by a vacuum leak).

Point of all this, fuel trim data can be useful - most problems will affect FT. However, if everything "tests ok", you might not be able to find any cause for an intermittent fault.

Posted via the PriusChat mobile app.

 

I had P0302, P3190, U0073, and even C1310, in addition to the red triangle. C1310 went away somehow, but my mechanic actually tested my 5 year old 12 V battery, and found that, despite having what I thought was a healthy 14.2 open circuit voltage, discovered low voltage under load. I still can't believe that could have been the problem, but the replacing a 5 year old OEM battery seems to be working. I usually get 7 years out of batteries before they die. I did notice the smart key was less responsive, but the vehicle seemed to be working fine otherwise until the red triangle of death and all sorts of P codes appeared. I am guessing that the 12 Volt battery drew so much power trying to keep it charged that it compromised power to the coils causing misfire and poor power.

 

I would have my money more on an as-yet-undiagnosed problem that just isn't happening at the moment.

An interesting guess, but you would see that immediately with a voltmeter on the 12 volt system. The only way for the power the 12 volt battery "drew" to affect anything else is if it was enough to pull the voltage of the whole system down, so that the ECM or the igniters were seeing well under 12 volts in operation. If you see the usual 14-ish volts when the car is in READY, that just isn't a thing that's happening.

C1310 is a code the brake system sets just to tell you that there is a code in the hybrid system. You haven't listed any hybrid system codes, which suggests you were using a scan tool that just isn't able to show those. My guess would be the hybrid system code was P0A0F, which just means "not sure why, but I can't get the engine to run". So the chain would have been the engine codes detected by the ECM, summarized with P3190, leading to the hybrid vehicle control ECU showing code P0A0F, leading to the brake ECU saying C1310.

 

I got the P3190 code and the m going to try and sort it out, but it just seems strange to me that even with the techstream software you can’t get any more accurate than it might be a long list of things. I mean when the MAF sensor goes bad on my Honda you get a code for that sensor. Or if any sensor goes bad you get the code for that sensor.

 

You're expecting more of a computer in a car than a computer in a car can give you. It is located inside a box with some wires coming in that connect it to various sensors and actuators. What it sees on those inputs are all the information it has about what's going on in the car. It can detect some problems very specifically—for instance, your Prius can also give you quite specific MAF sensor codes sometimes, say when the voltage coming back on the MAF sensor wire is a voltage no working MAF sensor could ever produce.

Other problems, not so much. What if the voltage coming back from that sensor is in range, but doesn't match the amount of airflow the computer expects for right now? Is the sensor bad? Or is the airflow wrong because of some leak or obstruction and the sensor is showing that correctly? How will the computer sort that out?

Even when you get that specific code about an impossible voltage coming from the sensor, how should the computer know whether that's because the sensor is bad or the wiring to the sensor is bad? It doesn't. Lots and lots of people assume if they get any code that says MAF sensor in the name, the car is telling them to go replace the MAF sensor. Maybe they get lucky often enough to keep doing it, or maybe it's just there are always enough new posters coming along and making the same assumption even though others have learned their lesson.

And then there are codes like P3190, which means the computer is only seeing the engine produce less than a fifth of the torque expected for the conditions. So, the engine is running really badly. Back in the day, you would just have to use your gasoline-engine diagnostic skills and get cracking, and that's pretty much what you still have to do with that code. There just are a bunch of different reasons an engine can run poorly.

 

It helps to add more background. When did this start, what parts have been changed immediately before and after the problem coded?

Typical issues include maf contamination, throttle body buildup, pcv valve disconnected, clogged air filter and or aftermarket maf. Sometimes there are additional sub codes are Techstream can provide. A clogged cat could reduce power as well.