P3191

First thing of anything. How is the 12v battery condition? especially check the start position load 12v battery test. Put the ignition to on but don't start the car. Test the battery voltage at this position. If it reads 12.1v or less, the battery's sufficiency for booting up the ECUs properly is questionable. Like computers, the ECUs are voltage sensitive and too low of voltage could cause the ECUs to send faulty signals to components. Do this regardless of how "New" you think the battery is.

Otherwise, there are 2 sensors, when it comes to running, starting and stalling, the tend to get overlooked, and I have been fooled by overlooking them in the past. The Camshaft Position Sensor on the RH side (looking from the front of the car) of the cylinder head, and the Camshaft Advance Sensor on the LH side of the cylinder head. If they are dirty with old oil scum, or if they are damaged or terminals damaged, they will cause loss of power and startup and stall conditions. Not very hard to get too and to remove, clean, or replace. Not very expensive either.

If you check your OBD2 live date mode readings, you can observe the cam timing advance, and see if the advance at different RPMs is with in Specification. If out of spec, it indicates an issue with one of these or a cam timing jump.

 

P3191 air induction faults can include (other than throttle body and MAF) vacuum leaks from anywhere in the car from under hook to around fuel tank and emissions. Fuel Pump & Filter issues. Cam timing issues, Manifold air leak issues. EVAP valve and switch issues, such as with the Purge system will cause this too.

 

P3191 isn't specifically an air induction fault. It means the same thing as P3190—a large output power deficit compared to expectation—with the added info that the engine was being started at the time. (If such a power deficit is observed while the engine is already running, you get P3190, and if it is observed while the fuel gauge is also reading a low level, you get P3193.)

What any of those codes tell you is that you pretty much have to put on your old gasoline-engine mechanic hat and test and eliminate your way through all of the ways an engine might have its power significantly reduced. The grand categories of air, fuel, compression, and spark are the big headings at the top.

 

No. The ECM doesn't have that information, so it's not a question a scan tool could ask and have the ECM answer.

The real published DTC Detection Condition for the code in Gen 1 is just that the engine has cranked for too long without starting (at an adequate cranking RPM, and with normal communication with the HV ECU). How long is too long depends on the coolant temperature reading at the time. (So even the coolant temperature sensor could be one suspect, if it reads too high so the ECM thinks "too long" too early and gives up.)

Beyond that, it's really up to the human to put on the old-school gasoline engine mechanic hat and work out why.

Of course, you can make use of a scan tool's live-data displays to answer various questions you might have while doing that.

 

What scan tool provides live data, or am I not using TorquePro correctly. I only get a few datapoints - system voltage, coolant temp, rpm. IIRC, and nothing else. I downloaded the Toyota bundle and it displays all the rest of the pids/gauges, but none of them display live data/anything.

 

Are you saying that the Torque Pro Toyota bundle puts a bunch of additional gauges on the screen but they never show any data?

Sounds like something wrong with the PIDs in that Toyota bundle. If the usual PIDs common to all models are showing real data (coolant temp, rpm, etc.) then the tool must be communicating with the ECM so it should be able to fetch the other PIDs too.

 

Yep, that's what happens. The bundle has probably a couple dozen or more gauges, but only a few monitor. You can see them all in the pid list, and all that work are actively monitoring in that view, but you can select from those to display as gauges of different types in your custom view as well. I have created a bunch of gauges trying to check the battery condition, but they don't work either.

I could believe it wouldn't read any except universal pids - maybe Toyota was that unique or Gen1 that old, but it seems to do a decent job pulling engine codes and some others, so yeah, it's communicating with the ECM. Have you used TP? I think I've asked the forum a couple of times if anything other than techstream monitors live data and got no response.

 

There are a bunch of different ECUs in the car, and the network connecting them in Gen 1 was pre-CAN and a real hodge-podge, serial data links and BEAN and AVC-LAN. Some ECUs shared the same serial link but communicated at different bit rates. From the diagnostic connector, you talked to some ECUs by getting some other ECU to pass messages back and forth. I would be very surprised if there's any tool out there besides Techstream that knows all the tricks to talk to all the ECUs in a Gen 1. There weren't even enough Gen 1s ever made for that effort to seem worthwhile to most scan tool makers.

Still, if you're trying to diagnose a P3191 from the ECM, I would think you should be able to see the data you need for that from the ECM, given the right PID declarations, even if the tool isn't good for connecting to other ECUs.

I'm pretty sure I was able to program PIDs into a ScanGauge II for the HV ECU and battery ECU also.

 

<bump>
P3191 - my first guess is water in the gas, somehow.
Usually it's a cracked/corroded fuel fill neck pipe (Toyota part 77201-47051) but could be a bad batch of gas.
The solution is a bottle of DryGas into the tank, wait 24 hours, then drive (as best you can)
to get the fuel level low, then put in clean premium gas.
I don't know any easy way to drain the gas tank of a Gen1 except to siphon it out through the fuel fill pipe.

 

If whatever scanner you've used showed you the P3191, it's good enough. It gave you the code.

The thing is, P3191 is not some kind of "here's your exact problem, do this to fix it" code. (Very few trouble codes are. Generally, the car's ECUs are in no position to be able to tell you why they're seeing what they're seeing, or what you should do to fix it. They can only tell you what they see.)

P3191 is set when the car has tried to start the engine but never got it to start within the time allowed for cranking. So, the code means you have to put on the hat of an old-skool car mechanic and dig in and see why a gasoline engine isn't starting.

Water in the fuel is just one of the possible causes ... but a likely one if you notice holes in the fuel filler neck and if you noticed the problem after driving through a lot of rain. (The gen 1 filler neck situation is that rust holes develop there, and then the rear wheel flings rainwater upward and it gets into the tank through the holes.)

For somebody else with a P3191, if that isn't the pattern, then they've just got to methodically look elsewhere at all the usual suspects.

 

I'd add that I expect the filler neck issues are more prevalent in snowy climes. I think the wheel throws salty snow up that that wedges in place on the top side of the filler neck near the cap. Just sits there for days if it stays cold.

Easily replacement, and easy diagnosis - just reach up there and feel across the top. Be careful. You can't see up there and, if it is damaged, there could be rusty, sharp edges of corroded metal, and you'll be getting a tetanus shot later.

Still, there should be codes for the issues caused by a bad filler neck. You can also easily disconnect the fuel line under the hood to see it's pumping fuel, and connect a gauge there to confirm the pressure.

Honestly, I wouldn't think a rusted filler tube is the most likely cause, but takes 30 seconds to check for rust, so it can be eliminated easily, or at least moved to the bottom of the list of likely causes. I'd be looking at injectors, MAF, throttle body, plugs.

But, I got a car doing the same thing, so I'm no expert.