Ham radio operators have learned that modern cars are very prone to issues from aftermarket wiring. Accessory plug wires very marginal, electronics susceptible to upset with rapid changes in current flowing through the body, etc. Many mobile installs now run all their wiring, hot and ground on net new wiring from the battery, often adding a new fuse block from blue sea or similar.
1 Like
I replaced 4 heavy duty wires under my hood due to corrosion. The were 2 negative cables from batt to chassis and batt to transmission block, and 2 positive cables from batt to fusebox and batt to starter. I used heavier than needed cabling. Check it outā¦
But what is most interesting is that when I went for my test drive, of course the CEL was off due to battery power reset. It was very cold outside, and I switched Torque on within a few seconds of the startā¦ Immediately I checked for pending codes, and what do you know, P0141.
So the ECU is not supposed to even check for heater issue until a certain warmup has occurred. See the pages from Service Highlights a few posts back. Yes, I know its for 95-97, but I doubt its very far off from the behavior of my 01. So either the glitch from powering up the ELM327 bluetooth is to blame, or the ECU is checking that sensor way too early.
I think I will try the other scanner at DMS, reset battery and see what happens. I can also try to add 4 ohms of resistance to the heater coil. So that would be two things to try at DMS. Any other thoughts or suggestions?
So I wanted more information about the POP. Turn out it is VERY serious indeed, as the POP extends far above the ~12 volt power rail and dips far below the ground rail. I was able to estimate its frequency at 1/500ns or approximately 2 MHz. I saw lengths of this aberration lasting as long as one millisecond. It is indeed quite serious, and so it makes perfect sense that we should be seeing some quite unusual behavior from the ECU.
So what is the remedy? Well, I will, of course, be writing into this forum as I figure this one out. Hopefully, I can make it clear how I go about debugging this issue. Like @AndrewLeCody mentioned, I will be adding a capacitance to squelch this. The trick is picking the right value of capacitance, and also making sure I have solid cabling to deliver the max current required. More soonā¦
1 Like
You may have a bad alternator.
1 Like
New alternator. That is not to say its good, as we all know, however I am quite suspicious of wiring at this point.
Speaking of points, I have a new one that I have not documented. This āpointā has become plain to see as Tracy and I probed and poked around on the car. I have loosened the screws holding the ECU in place. So now it moves around while driving. And furthermore, ever since we did this, I have been noticing a lot of strange behaviors out of the car. For example, now the remote key fob entry no longer works. The A/C no longer works. I have had the car die twice when idling at stop light. Hmmmm. Lets review these facts in this paragraph. I loosened the screws to ECU, Tracy jiggled wires, and now we are noticing all sorts of strange behaviors from the car.
This looks like a job for Lisle, pictured belowā¦
Some ECUs ground through their chassis.
Not sure on this oneā¦
1 Like
Of the two ground connections that connect the battery to ground, one of them connected to a bolt on the firewall. The other connected to a bolt on the transmission (which is connected to the engine by way of being a big hunk of metal in an of itself). I know this because these two connections are two of the four I replaced this week due to corrosion. I used wiring of a heavier gauge than what was present from the factory, and therefore I expect a lower resistance of my new cabling alongside a higher current carrying capacity. And the big plus is the fact that there is no green corrosion adorning my new cabling.
When I replaced these wires, I saw identical performance of the engine and electrical subsystems after installation as to what was present prior to the effort. In other words, all the same benefits and issues carried through the operation. It was as if nothing had been done. This tells me that the corrosion was not an issue yet, but due to the fact that there was considerable corrosion at the tips of the cables where they attached to the battery terminals, I believe that I may have averted issues that were soon to come.
What has had a drastic effect on the behavior of the car has been fiddling with the ECU connectors. And while now I have many new issues that were not present prior to touching it, I am actually pleasantly surprised by this result. This tells me where I should be looking to resolve my issue with the O2 sensor. So while it is no fun stooping over and reaching my hands behind the ECU (located under radio), I have the advantage of having removed the radio this time, so I think that I may be able to get things nice and clean back there with good connection.
Iāll let you know soonā¦
So I am working on cleaning these connectors right now. But the other thing that comes to my mind as worth review would be the ground connections on the car. A bad or missing ground connection could cause this behavior. Fortunately, in my Mazda Millenia Wiring Diagram document, there is a rather lengthy section dedicated exclusively to ground connections. So after cleaning these connections to the ECU and making sure I have a good seat (and also checking my connections through to that troubled Bank 1 Sensor 2 oxygen sensor), I will walk through each and every ground connection that should exists. Seems I saw maybe 30-40 ground points. Seems like a lot, huh? Must be importantā¦lolā¦
You should also take the ECU apart and look for cracked solder joints.
2 Likes
You evidently read my mind, as I was up at DMS this AM and did exactly that. The ECU looks perfect to my eye, and I have a discriminating retina.
But the wiring was not so perfect. One of the connectors was in place, but yet unclicked. I hate to admit this, but I believe that this may be a large portion of my issues. I was sure to click it in well, and I also did not mount the ECU. I let it ādangleā in such a way as to prevent any tension on the cables. I left the radio out so I have easy access, and the ECU rests atop bracketing with minimal tension on the four connectors, now all firmly squeezed in and clicked.
At first, I still had many bugs, the same ones mentioned previously. But I went to auto store and purchased a fuse kit. I replaced a 15 amp labelled āRoomā, and it seems everything has returned to AOK status. The key fob works now, it thats any sign of improvement. The scariest of my previous bugs was the fact that neither Torque nor Actron scanners were working. I havenāt tested yet, but Iām pretty sure that they are good now.
But the most exciting thing happened during my ride home from getting the fuses. The CEL was illuminated during the ride to the store, but when driving back with the new fuss in place, I observed that the CEL had gone off. Since I have to locate my ELM327 to run Torque, and have not done this yet, I cannot speak about pending codes, but I doubt the computer would clear the CEL if it still observed failing status. I think that I may have actually remedied the O2 sensor issue, but time will tell for sure (and so will the scanners). It really may be as simple as loose ECU connector and subsequent fuse replacement saves the day. Fingers crossed.
Which would bring me to my last concern. I have disconnected a bunch of electronics considering the glitch and fuse situation. I will want to bring in a new tap, for sure off of swiched power. Any suggestions on how to do this? I want to power the following extras: Bluetooth stereo, audio amp, navigation, camera, phone and tablet charging. I donāt need any of these things unless ignition on. Any suggestions about kits available to simplify creating a fused and switched tap for my vehicle?
And I just checked. Sadly,the code is still present. I reset it and a pending code appeared immediately afterward while idling in the driveway. Exasperating, but unfortunately it is reality. Time to carefully check wiring, the sensor itself, and the circuitry on the ECāU that verifies O2 sensorā¦
Some thoughts that have come to mind:
I tried to measure resistance to ground through the pins of the O2 sensor, with sensor and ecu disconnected. No such luck. Fortunately, the wiring diagram I have has a whole section dedicated to ground connections. This is obviously a great starting point for this debug.
Sensors connected directly to this sensor include: maps (manifold absolute pressure sensor, egr position sensor, transmission fluid temp sensor). So any of these guys misbehaving could cause trouble as well.
1 Like
Iām not understanding this, as if both ends of the wires are disconnected, thereās no path to ground.
At any rate, if you have both ends (ECU and O2S) free, and can find a lead long enough, measure the resistance in each wire. It should be zero (but Iām sure you knew that, and that it wonāt be, and roughly how to calculate what it should be based on type of wire and length to see if itās āin rangeā).
1 Like
And at long last, and right after a few days break from the beast, I quickly discovered root cause, remedied, and am looking positive for inspection readiness in another drive cycle or two.
A young man named Tracy helped me for a few hours look at this car. Thanks, Tracy! He was very knowledgeable about cars, and immediately pointed at wiring or sensor problem. Now I got us off in the weeds looking at the glitch on the power supply, which is still of some concern. But this issue was much simpler than that.
The oil dripping on the wires was a problem as it was accumulating where I had spliced, soldered, and shrink wrapped my woirk. The shrink wrappoings were loose, and oil got all in my solder joints, and I believe shorted everything out. A new sensor arrived without any sort of splicing, and it fixed the problem, but I found the root of a whole bunch of problems/weird behaviors by pure chance.
So I noticed in the shop manual that there is a steering switch that tells the ECU when someone is applying turning force to the wheel. This tells the ECU to rev up a few notches so we donāt power down while turning slowly (which has been a problem with this car for years). I started to look at this single wire, and a previous tech had neatly fastened this wire against the engine block with an aftermarket fastener. Only one problem, though, and that was the insulation wore down and cracked and so the wire was rubbing on engine block. This connection is right next to that O2 sensor connector as well. So I guess this circuit being shorted to ground was pulling current from ECU, bringing down supply rails and injecting noise from rubbing (011001010100101010) 4 V to engine ground. This was making the ECU crazy. Soooo glad I stumbled upon this one. The car runs differently now. I am shocked that the ECU did not break. I come from PC world, and for sure shorting a signal to ground on your PC will bring down the power supply and often woirse.
And since it is running so very differently now, I have to cycle through the timing and idling stuff again and get it all set right (it was ātunedā for the wacky setup with noisy wire polluting ECU behavior). But after I do this, I will reset, and the tests will pass 1,2,3ā¦done Iām pretty sure.
2 Likes
Luckily the ECUs are built to withstand crazy environmental problems - poor power, EMI/RFI from stereo gear and transceivers, short circuits, wild temperature swings, incredible vibrations and physical shocks, and more. They are designed and built to standards that few PCs could ever match - even Panasonic Toughbooks donāt have that much overdesigning done. 
2 Likes
I am a believer!
The only thing I think that is tougher than that ECU design would be medical devices aka Medtronics type stuff, pacemakers, etc. When a personās life rides on the device and failures WILL result in court cases, the testing requirements are so stringent it is a wonder anyone can make money building them.
That ECU is spotless, perfect. It looks like what I would expect to see come off the line nowā¦ None of the electrolytics are popped, no burns, no stains, no dust. Ive shorted stuff working on this car maybe 5-6 times, like sparks flying sort of screwdriver from 12 to engine shorts! Iāve had to replace 7-10 fuses and a few relays (and the main 120 Amp fuse)ā¦ But the ECU is still kickināā¦ no sweatā¦ lolā¦
Well, I am driving down my readiness tests for state inspection, and while there are some minor glitches, the car runs like a dream. Thank u all who hung in and kept me encouraged in the face of unrelenting expense amd difficulties. The result is well worth the struggle. And I no longer fear internal engine workā¦
As for the lengthy threads, I would like to post this disclaimer that what is recorded here is my learning curve. I eventually got led to or stumbled upon the correct answers, and many theories I had along the way proved to be quite incorrectā¦ like the one about Iāll just clean the head mating surface a lil bitā¦lolā¦
Also I have accumulated a LOT of footage amd pics at various stages during rebuild. Iāll post when I figure best way to share allā¦
3 Likes