2013 Mazda5 AT Power loss, No CEL and Throwing Codes

[whizbo]

It usually starts out just fine, and goes awhile before I sense it go sluggish. At that point, it does switch back and forth during the drive, but I don't notice the exact moments of the switch because i believe they happen when the engine isn't loaded. For example, I might accelerate off the line and everything feels fine. I come to a steady drive speed, then i decide to accelerate to pass or something, and that's when it might feel sluggish. If I ask for a little more, it will stay sluggish like I'm in too high a gear for the speed I'm traveling. Kinda like in a Manual when you're doing 20mph in 4th gear and punch it. It just kinda moans at you. If I ask for a lot more it will drop a gear increase the RPM, and it won't feel sluggish at that point, but it's hard to say if that's just thanks to the extra gear ratio. I have not ever noticed it become sluggish in the middle of an acceleration if the acceleration started off peppy.

Yes, I just have one of those bluetooth OBD readers that I link to with my phone. So, I can have my phone on the dash to watch variables. I can also graph up to two variables at a time. More if I pay for the pro version....but I'm not there yet

 

[edmaz]

Unfortunately, that performance pattern that you described would be really difficult to diagnose, mainly because it's impossible to identify the true start and end points. And another problem is that the issue could be with the transmission, just as easily as the engine. If you didn't have that wacko bunch of codes being set, then you would just do the standard diagnostic stuff like evaluating the typical performance hot spots such as plugs, O2 sensors, injectors, and coils. And taking a look at all of that would never be a bad idea, however none of it is going to resolve whatever is causing those codes.

By any chance, do you do have any experience programming microcomputers such as Arduino? A microcomputer programmed to check voltage would be the way around the limitations of the multimeter, because they operate at the microsecond level, and would be able to trap and record any abnormal voltage for the entire time the vehicle is running. Fairly easy to create the required code, for anyone familiar with that technology.

If you're not into that stuff like that, then I guess my only other suggestion at this point would be to leave the meter attached to the battery, with the hope that you will actually observe the voltage dropping, or see higher AC voltage, before too much time goes by. (and not rear ending someone because you're watching the meter )

 

[whizbo]

Hotter afternoon today. I’ve seen the AC voltage as high as .270v and the spikes correspond with the radiator fans spinning up. I also see about a 0.75volt dc drop when that happens. So it happens a second or two after the AC clutches engages.

An arduino setup might be a stretch. We’ll see how desperate I get. Ha!

Plugs are new, I ran a tank of gum out last month suspecting the injectors. Coils ohm out per the FSM. How can I test the O2 sensors? Monitoring the #2 O2 sensor shows 0-.9V while driving. The #1 O2 sensor varies from -1mA to +4mA do those values seem normal?

 

[edmaz]

That .27V AC voltage is interesting - likely not quite high enough to cause a problem, but certainly moving in that direction. How long do you estimate you saw it stay that high, or was it just a blip that immediately went down again?

O2 sensors typically become progressively slower to respond as they get older, and can have some effect on performance because of that, in particular with the upstream O2. However, most of the experts also agree that this is usually not a big impact, and the great majority of them never get swapped out, even after 15-20 years. Some of the sensor manufacturers claim they should be changed every 100K miles, but I wouldn't give that much credibility, given their vested interest in selling more sensors. No way to test this outside of a lab that has specialized testing equipment. But OTOH, YMMV and it's certainly possible that a new upstream O2 might help your vehicle's performance.

 

[whizbo]

Just a blip for the higher AC voltage while the fan accelerates. Once it gets to speed it goes away. This is a second or two after the A/C Clutch engages.

I'll consider getting a new O2 sensor on the way...But....

Here's a new development. While doing testing with the hood up. I've noticed the throttle body is noisy after I turn the car off. So, I opened it up and got some video. It hums from the time I turn the car off, until it goes through what appears to be a homing cycle. After all this, a couple seconds later I hear a "puff" in the area that I'm guesing is fuel pressure or vaccum relief through a solenoid in the area.

I've only recently notice this noise.....but I've only recently been paying this much attention to the car.
Does this seem normal? Certainly if I'm having throttle position issues it would affect performance.

 

[edmaz]

I have basically the same noise and click on my CX-5 after shutting the engine off. I haven't removed the air tube from the TB, but would expect to see something like what's in your vid, if I did remove it. I've never seen a writeup of exactly what that's doing, but suspect it might be some type of vapor flushing cycle.

And AFA some type of throttle position error, I believe that the PCM would quickly set a permanent code if that was happening. The PCM commands the throttle actuator based on the accelerator pedal sensor signal, and then checks the TPS signal to confirm that the throttle plate actually moved according to what was commanded. Not much wiggle room for undetected error with that design.

Here's one other possible diagnostic idea that occurred to me a short time ago. Early on I suggested reading voltage at the battery, instead of using the OBD live data value, mainly because I prefer looking directly at the source instead of having it travel through computer systems first. All well and good about that, but the continuous live data battery voltage PID is something that can be recorded and browsed later on. So even if the ODB values lose something in the translation along the way, they're almost certainly close enough for what we're looking to find out here.

Live data readers vary in their capabilities, and one big factor is how many values your reader can store, before it either stops recording or just wraps around in overlay mode. So if you wanted to try this, one key setting is how often battery voltage is sampled. And, assuming it's a modifiable parm, you can play around with the sampling frequency and see how that affects the maximum number of stored values.

Doing this testing of course won't provide AC voltage directly, but if bigger DC voltage drops show up, then it's reasonable to assume that some of that is a result of an increase in AC.

 

[whizbo]

Yeah, I'll start monitoring more with the Live data reader to see if I can capture larger spikes outside of what I've seen with the Multi-meter.

Today after I re-assembled the airbox after getting that video, I went for a quick errand. Then I got an actual CEL and a P0113 Intake Air Temperature Sensor - High Input. I believe air temp is handled by the MAF, so I cleaned the MAF and re-seated the electrical connector.

I was actually excited to see a CEL with the hope that something had finally completely died, but alas, another quick errand resulted in no codes.

 

[edmaz]

That code and CEL is actually helpful because it's evidence that the OBD system on your vehicle is not having any issues. And if OBD is functioning properly, I can't think of anything other than AC that can cause both a high and low voltage code to be set in quick succession on the same circuit. And the fact that nothing in the big pile of codes ever becomes permanent, suggests to me that the alt is sending just a very occasional, short dose of AC, which means the PCM doesn't get the number of fault reps required to set permanent codes.

If I had a different attitude about this stuff, I'd just suggest rebuilding or replacing the alt, However, I always prefer to have good solid diagnostic evidence before going to the parts department, so hopefully an OBD recording will show somewhat more abnormal voltage.

 

[whizbo]

I've been driving around monitoring voltages with the scanner. I'm seeing a drop when I accelerate off the line. In the attached image you can see the dip at the left of the graphs. The dips at the right were another smaller acceleration around a corner. Is a dip under acceleration expected? I also noticed last night while parked and idling, my headlights dim slightly on occasion when the rpm dips ever so slightly during the "normal" idle cycle. Feels like more signs pointing to the alternator.

 

[edmaz]

You definitely captured a very interesting and helpful test result! If you haven't already done so, here are some other things to check. The first would be the tension on the serpentine belt, and also the condition of it as well. Look for anything like glassy patches (indicating slipping), or damaged areas of the belt. Also, take a look the belt while the engine is idling, to check for wobble, particularly at the alternator pulley. And if it's not difficult on your vehicle, you might also consider releasing the tension on the serp belt, by removing it from the tensioner. Once the tension is off the belt, you can work the alternator pulley with you hand, checking for any sign of wobble.

Assuming you find no indication that the belt is slipping, IMO you've found enough evidence that the alternator is on the way out. A significant grounding issue is the only other thing I'm aware of that's in this ballpark, and there is some additional voltage drop testing that we haven't discussed yet, which you certainly can do if you want to run every possible DIY test.

If swapping the alt on your 5 is something you're up for doing, and you decide to go ahead with it, I recommend an OEM unit (reman or new). So find the alternator maker and part number on the I.D. plate. Then you can read through the specs for the various ones for sale on Rockauto, or other parts site you usually buy from, and make sure they're using that same part number.

One other interesting adventure is to buy a rebuild kit, including rectifier bridge, and rebuild it yourself. Not something that many folks want to get into, but it might be a good choice for someone who doesn't mind doing quite a bit of extra work, in order to save a small amount of $$

 

[whizbo]

The Serpentine belt is a little shinny. I'll get parts coming to replace that to see if it helps with the voltage drops during accel. I have some questions about that, but I'll start a new thread. There aren't a lot of threads about serpentine replacement.

 

[edmaz]

Ok, I hope you get a reply - not very many 5 owners on this forum. Perhaps you can find some info using Google search.

 

[whizbo]

Not very many 5 owners anywhere! It wasn't a hugely successful vehicle.
I added some links in that post if anyone is hunting serpentine info.

 

[whizbo]

Sepentine belt, tensioner, and idlers are replaced. The smooth Idler bearings weren't great, but they weren't wobbly yet either. Tensioner pulley bearings were bone dry, but still smooth. Grooved pulley bearings were ok, but I replaced them anyway. Alternator bearings feel good, water pump feels smooth. Problems still persist. I swapped in my old TCM to see if the issues followed the TCM, but they didn't. I'm now suspicious of the time the engine runs in open loop before it warms up, and switches over to closed loop operation. If it's acting normal before the switch, it would point at components in the closed loop. So, PCM, Upstream O2 Sensor, or the already identified grounding or alternator issues. Maybe MAF? I'll try to watch for when the switch happens from open to closed loop, and see if it coresponds with the sluggishness.

 

[hleboklop]

Sepentine belt, tensioner, and idlers are replaced. The smooth Idler bearings weren't great, but they weren't wobbly yet either. Tensioner pulley bearings were bone dry, but still smooth. Grooved pulley bearings were ok, but I replaced them anyway. Alternator bearings feel good, water pump feels smooth. Problems still persist. I swapped in my old TCM to see if the issues followed the TCM, but they didn't. I'm now suspicious of the time the engine runs in open loop before it warms up, and switches over to closed loop operation. If it's acting normal before the switch, it would point at components in the closed loop. So, PCM, Upstream O2 Sensor, or the already identified grounding or alternator issues. Maybe MAF? I'll try to watch for when the switch happens from open to closed loop, and see if it coresponds with the sluggishness.

Did you resolve this issue?

 

[whizbo]

Did you resolve this issue?

I don't believe so. It is definitely tied to outdoor temperature, and happens when it's warmer, so I haven't seen the issue through the winter. I did take it to the dealership last fall and they took it for two different test drives and their driver noticed nothing and said it's operating as it should. It was a cooler day, so maybe it wasn't happening. I'm tempted to trade it in on another Mazda5 at said dealership! Ha! Last fall I also opened up the top of the EGR to verify that it was operating smoothly. I thought maybe it wasn't opening/closing properly and that was giving a sluggish feel. It seems to actuate just fine. I left the EGR unplugged, and it threw some codes, so the downstream sensors are working well enough to sense the EGR completely not working. I'd like to do some more alternator testing on some really hot days this summer. With no obvious culprits, I still suspect some odd voltage/current jumps in the system might be at fault.