Honda Odyssey Forum banner
1 - 13 of 13 Posts

bbarbulo

· Registered
Joined
·
401 Posts
Discussion starter · #1 ·
Background: 2000 Odyssey with 189K miles, american car. Transmission just rebuilt, driving for the first time - TCS light stayed on after starting.

Pulled the codes, codes 34 and 31 came up - TCS codes, not to be confused with ABS codes.

The codes are indicated as low reference voltage and permissive missing respectively.

I backprobed the relevant connectors and found 5V reference from the ECU to the TCS unit... at the time the car wasn't able to run due to a drained battery, so I didn't troubleshoot code 31 at that time. I reset the codes. Car is running now with a new battery so I went back to troubleshoot code 31, I started the car and the TCS light came on several seconds after starting the car. I backprobed the relevant permissive pin with an analog meter. Nothing there. It was wicked cold out so I didn't carry on with getting the ECU pulled and checking the permissive signal at the ECU itself. I wanted to do more research about the subject. The troubleshooting at this point says to replace the ECU. Seems a little excessive, and I wonder what the actual permissive signal from the ECU consists of? One has to assume that the ECU is looking for conditions to be met in order to give the signal to the ABS/TCS unit. So if I'm missing a condition, seems ridiculous to go and replace an ECU that is doing what it's supposed to. I should point out, this time only code 31 was set. I suspect code 34 was set when I was cranking the car with an almost dead battery, voltage prob dropped below the threshold necessary and the code was set. I'm no longer worried about that code.

In summary, the question is:

What makes up the permissive signal from the ECU to the TCS; ie. what conditions cause that signal to be withdrawn?

If there weren't any conditions, then one would assume that TCS would just be wired to run anytime the engine was running and the vehicle was not in Park or Neutral. However, this is a heartbeat signal, as in, a pretty robust 'permission' communication method as far as controls goes. The heartbeat is 50% duty cycle between 0 and 5V, meaning a square wave (at an unknown-to-me) frequency. All I know is an analog meter should show the signal as a 2.5V reading according to the service manual. Neither the service manual nor the electrical troubleshooting manual offer any further insight, presumably discussion of controls logic is beyond the scope of a shop manual.
 
Discussion starter · #3 ·
thanks for the link! I don't have any CEL, ABS, or TCS codes. Having any other codes would lead me to believe that the other code is the missing necessary condition for the permissive signal, so I would troubleshoot that first. That's why I took care of the battery and the VREF voltage first. I'll keep you guys updated, it could be a broken wire from the back side of the ECU but that's about it. If the ECU isn't outputting the heartbeat I'll leave it alone for the time being.
 
Sorry, I'm not familiar with that signal at all, so I'm no help. The 5v reference signal is a no brainer but I've never even heard of the permissive signal. Would be interesting to see the waveform of that signal on a working vehicle. It's possible it's generated based on the input from some other component but, again, I don't know enough about it to be of much help.

Maybe if you were to post a bit more info so that someone with a working vehicle could go check theirs and see if/when they get that 2.5v you refer to. Remember, if it's duty cycle square wave then to give you 2.5v on a meter it would have to be exactly 50% duty cycle of the 5v reference. If it's simply off longer than it's on then you would see very little voltage but would tell you the wire is OK. An analog meter is helpful in this case but this is where it's nice to have a scope.
 
Discussion starter · #5 ·
That's alright, I didn't expect a LOT of help from the forum, just a sounding board really. This is more of a record for others to follow and see what DIY troubleshooting looks like and potential pitfalls. For there to be any real help on the forums, there would need to be an engine controls engineer in here (not necessarily from Honda, most ECUs function roughly the same) and he or she would need to be familiar with powertrain controls like TCS, not just engine controls.

And turns out you were tremendously helpful anyways!

So went backward to the ECU and see if the ECU was outputting the signal. Initially I got too cold to dig back to the other side of the ECU (plug E on the pass side), so I was probing a 2 pin connector with the same wires on the left side of the ECU. The source of the signal is a 12 pin connector (with only 2 wires) on the right side. That left the possibility of a broken wire between the ECU 12 pin plug and the 2 pin plug. There was no signal there either.

Still not satisfied that I needed a new ECU, I figured I'd check the ECU for pending codes so I hooked up my scanner. Pending codes are failed modules that haven't yet met conditions to set a permanent check engine light. Lo and behold, my check engine light doesn't work and I DO in fact have codes stored. LOL what a n00b!!! I missed the simple fact that my CEL doesn't work!!!! So two codes came up P0135 and P1676. The P0135 is common to the other thread that John Clark posted, the O2 sensor heater circuit bank 1 sensor 1. The other one is FPTDR signal line failure. I'll troubleshoot these, weather permitting... and also the CEL light situation... someone prob just pulled out the bulb from the cluster.

But the indication is that the heater circuits for the O2 sensors (at least for sensor 1, presumably in each bank) would be a necessary condition for TCS to be permitted to run.
 
Discussion starter · #6 ·
The wires in location 3 and 4 of the engine side primary o2 sensor had become depinned. Black/White and Black/Yellow were out of their sockets. I tried to put them back in to no avail. I ended up depinning the connector and put them back in, one at a time. Since I was there, I verified the heater circuit was at between 10 to 40 ohms, and it checked out at 13.5 ohms. Buttoned everything up and started the car. Deleted the check engine codes stored earlier, and no codes set again. TCS light also went off.

Just for others doing this, I took a shot of the 'permissive' signal from the ECU, the needle bounces roughly around 2.5V as anticipated.
 

Attachments

Discussion starter · #7 ·
Oh, almost forgot, P1676 relates to the other wire in the two pin connector that goes from Plug E of the ECU to the TCS module. There is light green/red which is the permissive signal, and pink/black which is the FPTDR signal. Since that was unplugged during testing, that's what that code was. With everything plugged in and working, that code is irrelevant and did not appear again.
 
ECU's, on a Honda anyway, pretty rarely fail. I ran through a similar goose chase on my radiator fans a while back. I posted a thread about it.

This is a good lesson to always pay attention to all the red lights that come on when you turn the key on. Pending codes are usually codes that set when the PCM has seen something once but is waiting to see it again within a certain number of drive cycles. Not all scanners will show pending codes. Another feature of the higher end scan tools is showing "history" codes, meaning a pending code came up but then the computer didn't see the issue again and removed the code from the pending status, or the code even set a check engine light but then the code went away, either by repair (and without the codes being reset) or on its own.
 
The wires in location 3 and 4 of the engine side primary o2 sensor had become depinned. Black/White and Black/Yellow were out of their sockets. I tried to put them back in to no avail. I ended up depinning the connector and put them back in, one at a time. Since I was there, I verified the heater circuit was at between 10 to 40 ohms, and it checked out at 13.5 ohms. Buttoned everything up and started the car. Deleted the check engine codes stored earlier, and no codes set again. TCS light also went off.

Just for others doing this, I took a shot of the 'permissive' signal from the ECU, the needle bounces roughly around 2.5V as anticipated.
This thread is an excellent example of proper troubleshooting. So many people would have found that code and run down to the parts store for a new O2 sensor. If a code reads "blah blah blah O2 sensor blah blah blah" that immediately means it needs a new sensor. This is a good example of why you actually check wires. Not everyone would have noticed the unpinned connector but if an O2 heater circuit code comes up there's no excuse for not ohming out the heater. Even if you hadn't noticed the wires first you would have suspected something in the wires after finding a good heater in the sensor.

Nice thread!
 
Mind blown at JC and bbarbulo.. you guys are bad a$$. I am simple monkey wrench backyard wannabe mechanic... you guys are the real deal and glad to be on a forum with you two at the helm. Hope users on here realize how lucky the community is to have guys like you around. I have used forums for over a decade so trust me I know. Great work guys.
 
Mind blown at JC and bbarbulo.. you guys are bad a$$. I am simple monkey wrench backyard wannabe mechanic... you guys are the real deal and glad to be on a forum with you two at the helm. Hope users on here realize how lucky the community is to have guys like you around. I have used forums for over a decade so trust me I know. Great work guys.
Thanks for the kind words but I consider myself in training as I still have lots to learn. I've been to the GM Truck forums and, trust me, I know what you're referring to. Every time I go over there I see people posting things that make me face palm non-stop. They spend money on parts and never fix their vehicles. Then they give the same advice to others. It goes something like..."I went out to my truck this morning and it won't idle. Any ideas?" Then the masses just come in and say "Replace your Idle air control" or "Replace your elelctronic throttle body." And so many just do it. It's sad. They'd be better off just to take it to someone who knows what they're doing. The only problem is many techs in shops don't properly diagnose either and just fire the parts cannon at them. Dealers are notorious for that. The Honda factory service manual even calls for swapping parts "to see" if a given part is the problem.
 
Yeah I use the board to do the do it yourself stuff and if I cant do it myself, I take it in to get looked at and advise them on what I have learned. Some use the advise and it at least points them in the right direction, and if they really know what they are doing helps isolate the issues. I am on the Mitsubishi Evolution forums, and see some of the same there, but that community is pretty tight and even has sub sites by location , socalevo, base for evo owners in the socal area, and lots of times many of us have met or cross paths at some point. Glad to see this community is still active especially for an almost 17 year old car.
 
Discussion starter · #13 ·
Thanks mr0072003 at least we can hope to inspire or help someone else keep their cars on the road longer.

I agree, forums are a great resource that social media, twitter, instagram, or snapchat can't replace. For fear of losing documentation and pictures from forums (of stuff posted over 10, sometimes 15 years ago), I "print as PDF" and save it to my computer. Some of this stuff is like GOLD. Imagine finding turbo information and how-to's for a 1st gen Saturn!!! I have a 93 twin cam wagon w/ a manual tranny I wanna turbocharge, so anytime I see good intel on tuning and injector info I "print as PDF".

You guys are lucky in Cali, your cars don't rust. With a little aptitude, you could literally have the same car for life. Biggest challenge for me is finding a car worth investing my time in. I only recently started doing welding/rust repair - spent most of my summer restoring the rear wheelwells and rear trunk panel of an '88 Civic sedan.
 
1 - 13 of 13 Posts