[george w]

For what it's worth, the replacement O2 sensors are Mopar. So.. when codes first appeared first thing done was check and clean the O2 sensor plug contacts with contact cleaner ( they appeared as new, but I used contact cleaner anyway.) Second, engine compartment grounds checked to include battery cables and body ground lug on inner fender behind battery These also appeared tight, clean and as new. Cleaned 4 connector plugs on the PCM and unplugged and replugged these plugs several times.
All of these things would seem to work of short periods but the two codes always returned. Next was a trip to an independent shop that replaced both O2 sensors with Chrysler replacements as a first step. Codes came back. The shop said that all their technical resources indicated that if the O2 sensors are changed then the only other choice was to replace the PCM. I did not have the shop replace the PCM as the quoted price was "about" $ 800. I then ordered a "rebuilt" PCM with a lifetime guarantee from an on line company. Installed the PCM and returned my OEM PCM to them as a core. They would not "rebuild" and return my original PCM but provided exchange units only.

All appeared well for about 2 months or so and the P0031 & P0037 have returned. Coincidentlally the car battery ( replaced about 30 months ago ) started to appear a bit weak when first turning the motor over when starting the car. The car has been used for mostly short trips since the first of the year. The starter just didn't seem to initiate that first rotation as quickly but would always start the car in a few seconds. On a hunch I put a full charge on the battery overnight. So far the CEL and the two codes have not returned. So for now I'm in a holding pattern. If the battery appears to weaken again and if the codes return I'll replace the battery.

It doesn't seem to make sense to try replacing the O2 sensors or swapping out the PCM again. All the while, the car has been running perfectly as it always has. maybe the battery is flakey in some way ??

 

[boxdin]

Are there any parasitic power losses when shut off, parked ?

 

[george w]

Just a follow up. There appear to be no parasitic power losses and all electrical equipment in the car works as it should. I gave the battery a good overnight charge about 10 days ago and the problem has not reappeared so far.

 

[astro_ext]

I've been having the 31/37 issue and recently found out yesterday what the real problem was. I will mention the solution a little later in my post. I'm a small car dealer and bought one of these PT's (03 GT 2.4 Turbo) with the 31 & 37 issue. The internet has been invaluable to me in fixing issues in any car that I buy, and believe me, they all have them. So I thought I'd give back a little and share with everyone how to fix this problem. The 31 & 37 issue is very common with these cars, but all too often, someone posts a problem, ends up taking it to the dealer, then never bothers to take the time to share the solution. The internet is filled with dead-end threads regarding the 31/37 issue. This one is by far the best, but still incomplete. I hope to put this issue to bed once and for all.

While following this thread, I did what George W did and ordered a new PCM from ACR. Installed it, cleared the codes, reset the monitors by driving, and was able to run it through emissions testing. I figured this was the end of it. Well, while driving home one day, the codes came back. They were stubborn as hell, my Innova scanner wasn't able to clear them. All that it could do was reset the monitors. The codes were also both stored & pending and I couldn't reset the monitors no matter how far I drove the car. I took it to a small indy shop and they were able to do a 'relearn' on the system, which did clear the light and I was able to reset the monitors, but the light came back after about 100 miles. In between repairs, I pulled my hair out checking grounds, heater circuit resistance, wiring harnesses. The 02 sensors were new, probably replaced by the previous owner in an attempt to fix the problem. There is a coolant line that runs behind the engine that had leaked some coolant on the wiring harness below which I thought could be a problem. I thought I was getting somewhere, but once I took the insulation off, it turned out the wires were fine. I finally bit the bullet and took it to a dealer for a solution.

Diagnosis---PCM drivers defective

The solution is to replace the PCM (unless you have other obvious problems like bad 02 sensors/grounds/etc). My original diagnosis (based on George W's experience) was correct. However, at this point I'm not overly confident that ACR can fix it. They're sending me a replacement PCM, but for all I know, it could be the old core that I sent them and I feel therein lies the problem. This issue seems to pop up more as these cars age. 11 years isn't too terribly old for today's modern cars, but it is for the electrical components. If you've noticed, most PCM's have heat sinks on them. Even though they're usually not directly placed by a heat source in the engine bay, they still generate a lot of heat through electrical resistance. I feel that this is the real issue. Whoever gets my old PCM will be getting an 11 year old component subjected to 11 years of heat and wear and tear. This combined with what I feel is sketchy electrical design/software by Chrysler and you have a recipe for disaster. This is why I'm not sure that the replacement unit will cure the problem any better than the first. The only other solution is to go with a new factory PCM from the dealer, but that could run $400-$600 depending on your location.

I hope this helps anyone out there having the same issue.

 

[george w]

Logic seems to indicate that setting these codes falsely is pointed to the PCM going out of spec. I agree with astro-ext that someting's deteriorating within the PCM. maybe this is more common with the turbos as they generate a lot of underhood heat that's trapped. I've baked several windsheild washer nozzles so far.
My PT GT started these 0031 & 0037 issues at just about 100K miles. Currently I still have the rebuilt PCM in the car and the problem came back once since it was installed. Thinking this could be somehow related to the battery or charging voltage, I disconnected the battery ( a two year old replacement of the original ) gave it a full charge as the car has been sitting a lot lately, and reconnected it. So far the codes have not returned. If they do I'll consider going to the dealer and ordering a new one ( if they're available ) but only if it's plug and play. I may even try buying a new battery first, as the replacement seems a tad weak when first hitting the starter.

I really don't want to get hit up blind with a full price PCM, dealer diagnostic fees and any "programming" fees that could be added to the tab. If the issue returns I'll go to the local dealer, explain the issue with the service writer and see if I can get a quote for the PCM replacement up front. It's certainly worth a few hundred extra bucks to get this fixed permanently. Other than the front control arm bushings going out ( twice, but the roads have been beat up in the winters ) and one set of motor mounts, the car has been almost trouble free and is still a blast to drive.

 

[ImperialCrown]

With reman anything, it is possible to go through a couple before you get a good one. Some reman modules are probably stuck in a revolving door at the parts counter.
My dad's PT had a P0601 code that we replaced the original PCM for (it still ran OK). A few months later a charging system fault showed up that was in the PCM. It was covered under the reman warranty.
Two of the same fault in this situation may be unusual, but could certainly still be in the reman PCMs that both had the same issue. I suspect that they get bench tested for a few minutes, then spray painted and boxed up and shipped out with a Hail Mary.

 

[george w]

Ain't that the truth ! Considering how electonically complex the PCM is with all the IC's, small circuit components and very likely, even smaller SMD's, I wonder what "remanufacturing" really means. It's pretty self evident on a mechanical part but on a complex electronic device I wonder how far they really go other than maybe checking for leaking caps and dirty connectors. I can't imagine they only reuse the case and metalwork and replace everything else.
Some slightly out of spec small component could be causing these spurious check engine codes when nothing's really wrong at all. Short off all new parts how would a rebuilder find such things unless it became a well known problem.

The rebuilder would have to have an inside track to Chrysler engineering or at least to the OEM testing criteria and equipment.
Anyone know what's involved with installing a new out of the box, sold over the counter PCM ? Does the dealer need to be in the loop to set or initialize certain parameters in the new unit in order to be compatible with the other electronic parts of the car like the keyless entry, ignition key chip or the odometer ?

 

[ImperialCrown]

Mopar offers a brand new 'generic' PCM that comes as a 'blank page'. The dealer then downloads the correct software application for what the PCM is going into. This saves money and having to stock pre-programmed units for every conceivable PCM part # out there. One part # on the shelf will cover many different part #'s out in the field. These are brand-new PCMs and although they cost less than a pre-programmed factory unit, the dealer has to tack on a programming charge. To be fair, this should only take 0.5 hr, but many dealers will tack on an hour or their 'minimum charge'. It varies between dealers.
http://www.moparpartsoverstock.com/p/__/MODULE-Powertrain-Control-Generic/6956938/RL044743AD.html
Mopar also offers remans (used PCMs) from companies that meet the Mopar quality standards. They are given factory support to meet factory quality standards. They can say that they are an 'Authorized Mopar' reman supplier or repair depot. Again, I don't know how intense the bench testing is and there are bound to be returns for very intermittent failures that aren't spotted the first time. I used to write the fault code (if known) on the PCM housing before sending it back in case the paperwork for the return unit got separated from it and to give them something to look for.
I don't think that they actually unpotted the circuit board to do a repair down to the SMT component level. That probably would be outside the scope of many reman facility technicians. The box can say '100% computer tested', but that doesn't tell me that they would have found anything in the short time on the bench. They are shiny again and have fresh paint and stickers.
http://www.jeepsunlimited.com/forums/showthread.php?550330-Need-a-new-PCM-anyone-dealt-with-allcomputerresources-com

 

[george w]

Because of the highly intermittent nature of the codes and the fact that the engine runs perfectly and gas mileage is unaffected, I'd have to assume that whatever algorithms are used to monitor the O2 voltages or the components associated with the same are what's a touch out of tolerance. Certainly typical for sensitive electronic components that can deteriorate either over time or from heating and cooling cycles. This tells me intuitively that unless that every active part that's involved in the 02 control or monitoring circuitry can be tested under operating conditions that it's unlikley that such a intermiitent issue could be solved in the reman process. If they replace the entire board and only re-use the PCM housing would I feel confident that a reman would solve the problem,at leat for another 100K miles.
What I've learned so far is that the PCM is clearly the fault and that my reman unit is probably also just on the verge of being out of tolerance. Since it has a "lifetime guarantee" should it crap out again then I can allegedly swap it for another and probably either have the issue either A-fixed the next time, B- Get the same problem back at some point or C: trade for a different problem to pop up. I'd do a swap under the lifetime warranty once, but that's it. Any recuurances of false codes then I'd go for the new PCM no matter how much it may cost. The question would be do I get the new generic PCM from Mopar parts or the GT Turbo specific unit ? As long as the pricing between the two choices isn'r in the order of hundreds or dollars i think I'd want the model specific version. On the other hand by the time that I go through the "swap" cycle again it's possible that neither the generic or the specific PCM may not be available new !

Do you know if it's possible to get the new, specific to the GT, PCM new over the counter already programmed so it can be plug and play later ? I'd seriously consider buying the new one now when I can get it and having it ready should it ever be needed. I don't intend on trading in the car as I'm perfectly happy with it and there's nothing out there that I would like to replace it with so I don't mind investing in mission critical spare parts. Only a total loss, really serious collision, would take the car out for me. I have other Chrysler/Mopars and I just sold off to a co worker my 94 LeBaron convertible with 354K miles on it. It's had all the required maintainence over the years including a new top, most areas repainted, head/valves serviced, a Chrysler reman trans. at 250K plus all the usual other maintainance ithems that you would expect. The car still looks great and drives just fine. The only reason I sold it was to buy an identical 95 with 45K on it so I can drive the same car for another 300K miles ! I clearly don't have a problems investing in the PT as I like it as much as the LeBaron. I've sold many nice cars over the years because i simply tired of driving them or I found that they didn't "fit" me all that well after driving them for awhile. I've learned to keep what I like and maintain them mechanically and physically so they run and look like new.

 

[ImperialCrown]

The generic one is just a 'blank' module. It would be programmed specifically for your GT turbo with only that software. It would be illegal to install anything else other than what yours came EPA-certified with. Flash updates give the PCM a later part # (eg: ending with AB ☞ AC, etc). You generally want the latest and greatest part #.
In this case, generic is a good thing. Only the dealership is authorized to do this under current and foreseeable emission laws. Once a PCM is 'written' it can't be 'unwritten'. Programming must be correct the first (and only) time.
Google your PCM part# for results. Add an 'R' or 'RL' in front of the # for a pre-programmed reman version. There are Fed or CA emissions differences, Auto/Man transaxle differences and possibly other criteria.
The P0031/P0037 fault is pretty simple to figure out. It is the O2 element heater circuit and could only be the O2 sensors themselves, the heater wiring/connectors to and from the sensors or the PCM itself. The problem is thinking when you replace a part and the fault comes back, that the part you replaced can no longer be what the problem is. Aftermarket or Universal sensor may have different heater curves that the PCM may see as a fault. The intermittent nature makes it even more frustrating if you can't catch it in the act to 'prove' the problem:
814f7cd5.gif

 

[astro_ext]

Replacement PCM is due to arrive Friday. I'll install and reply back after I've driven it some to see if the problem is fixed. If I end up having to go the factory PCM route, I think I'll just keep the car for myself for awhile since my profit will be basically 0 at that point. It's a good little car to tool around town/go to auctions/run errands in.

 

[astro_ext]

Just installed the replacement PCM and didn't even get a chance to drive it before the light came back on. I forgot that when I installed the first PCM that I had trouble turning the light off. I tried 2 different scan tools and was able to do it with a simple Actron pocket scan. I figured it was the tool, but it turns out that you have to grab the back of the port and hold it firmly together with the tool plugin while clearing the codes. It seems as if the port doesn't fit that snugly in its holder. I must have been lucky the first time I did it but now have figured it out. Unfortunately, while idling for about 10 minutes, the codes returned, both stored and pending.

I will take a little time to figure out what to do with this car, but the auction is the most likely place it will end up. I hate to do that to what seems like a very solid car aside from this issue, but am at the end of my rope with it right now.

 

[george w]

The codes are back ! The O2 sensors were replaced first and they're Mopar. I guess it's time to try exchanging the reman. ECM as it's under the rebuilders warranty. Here we go again !
If this doesn;t pan out long term then it's off to the dealer for a new replacement. I guess I should get a price quote first.

 

[george w]

Finally got around to calling the PCM remanufacturer ( All Computer Resources ) and ordering an advance PCM replacement. We'll see if this solves the problem. If not, then it will go back to a Chrysler dealer. I'll call ahead and make an appointment to make sure that they're aware of the generic PCM that they program in the shop. I'll get a cost estimate though I have a feeling that it's going to be pricey !

 

[Husky_Jeff]

And here I am back, my first posting was on March 4th, and after procrastinating for 5 months about what to do with my 2005 PT Cruiser, GT Turbo High Output, with the Po031 and 37 codes -- I ordered and received a replacement ECM from autocomputerexchange.com in the Miami, FL area. I installed the ECM and immediately... immediately had the 31 and 37 codes. I was so disheartened, as I'm sure george w has been.

I hadn't followed this info posting in the ensuing period since April and now see so much more valuable info that I'm both amazed and grateful for all of you continuing to post. I read astro_ext's info from May 28 with great anticipation as to what the real source of the problem would turn out to be, and discovered to my dismay it's the ECM (if I read it correctly) and that a factory replacement is probably the only way it will be right. After seeing his subsequent posts from June 8th and 10th that he had received a replacement ECM with the 31 & 37 codes coming right back, I feel it's most likely an impossible situation to correct short of buying the factory new unit and having all the programming done at the dealer, a course of action that is out of the question for me.

At this point I'm thinking I will call autocomputerexchange and talk with my guy and see what he recommends as to a replacement ECM from them. He stated when I ordered my unit that the turbo model ECM's are in very short supply, and considering that at most only 3% of PT's were turbo models this comes as no surprise. Their listing in ebay states they test the ECM's both hot and cold, so I felt pretty confident they would "weed out" all the intermittent problems...wrong.

I have not driven the car more than about 5 miles after the replacement ECM installed, and noticed astro_ext said he drove a distance to clear out the pending codes, so I'll do this today to see if it helps, maybe about 25 miles. And I'll be checking back daily to see any progress any of you are making in this. Again, thanks for being the only place I've found on the web with a real "in-depth" continuing effort to provide answers to the 31 / 37 problem.

 

[Husky_Jeff]

Followup to yesterday - disconnected battery overnight, charged it fully, and checked for codes at Ignition On, and no codes present. Good so far.... Started engine and Po031 and 37 codes came up within 1 minute with car idling. I agree with astro-ext when he stated that the PCM's are in a loop with problems that aren't addressed at the "rebuilders" shops - the supplier, in my case, agrees that this Po031 / 37 failure mode is typically caused by the PCM microprocessor, which they can't replace.

I called the supplier and they were are sending another PCM today. They are aware that they cannot monitor every failure mode since they don't have a test vehicle and cannot fix what they can't see. As a retired associate test engineer in a major electronics plant, this was one of our top tasks - address every possible failure mode with comprehensive testing under all conditions... very hot, very cold, vibrations, etc. I don't mean to be hyper-critical of the suppliers, but I sure feel like they should address the failure mode / modes the PCM was returned with, as possible, before sending it back out and "hoping" it will work ok for the next guy putting it in his car.
Enough said, so I will report on the next round of the battle when the PCM arrives late this week/early next week.

 

[george w]

I too contacted the supplier of the "rebuilt" PCM, All computer Resources of Miami, FL. which sounds like the same place Husky Jeff has used even though the business name is slightly different. After a wait of about a week and a half an advance replacement was sent to replace the "rebuilt" PCM that was generating the same intermittent P0031 & 37 codes as the original PCM was doing. BTW, the "rebuilt" PCM worked OK for about 6 months before the problem re-occurred. The replacement was installed and immediately upon start up the CEL light lit and the code P1604 was set. I tried disconnecting the battery several times, but in every case the CEL light would come back on immediately. The car would run, but ran improperly. There was little to no turbo boost and it seemed to run somewhat rough.

I reinstalled the original "rebuilt" PCM and the car returned to proper running condition and after a few miles the CEL light came back on with the same P0031 and 37 codes. As before, this PCM occasionally would reset itself and the CEL light would go out. Both with the original PCM and the "rebuilt" PCM the engine ran perfect even with the CEL light lit.

I returned the replacement ECU to ACR and got a phone call about 2 weeks later. The caller said that they "tested" the returned ECU and no problem was found ! We had a brief discussion, as I found that the ECU was OK hard to believe. I don't know how they test such things but I suspect that the ECU they sent was for a non turbo model or some other Chrysler product. One aspect about the "rebuilt" ECUs from ACR that I don't like is that they paint the entire ECU with a gray paint that obscures all the original labels, part numbers, bar codes, etc. They add a label that they generate, so you have no idea of what the original part information may be. The paint job is also somewhat sloppy as the four electrical connector sockets color coding gets somewhat obscured by the paint overspray.

As of now I'm awaiting yet another replacement, but like Husky Jeff said above, I suspect that they don't really "rebuild" these units at all. A "rebuild" to me would mean that the case and connectors would be reused but all the electronics inside to include the PCB and such would be new. I could also see if they reused the electronic components but replaced all the parts that would have affect on the trouble codes that were reported.

I'm fairly sure at this point that if the next replacement doesn't work correctly that it will be a trip to a dealer for a new ECU. This car runs hot under the hood and has crossed the 100K mile mark. These used ECU's are probably all deteriorating from the under hood heat. The car had already baked two sets of windshield washer nozzles from the under hood heat. I made two small heat shields out of several layers of folded aluminum foil and tucked it behind the hood insulation where the nozzles are located. So far, these newest nozzles are still OK.

 

[ImperialCrown]

There is no way that they can test for every possible contingency on the bench. There are too many variables, even if they had a PT there to plug it into before they shipped it. They do not unpot the board and 'rebuild' it. That is beyond their realm.
Many PCMs come from a junkyard and are painted 'reman' silver. Some aren't even cleaned off first.
Part quality is variable at best. They are shipped off to the customer with a 'hail Mary' and hopefully won't come back.
Many reman parts are stuck in the 'revolving door' as you suspect. Many are repackaged and returned to the stock shelf with a 'better luck next time' attitude.
Mopar does have higher standards and are only 'remanned' by a factory authorized remanufacturer (with factory support). They cost a bit more than the aftermarket, but any concerns are promptly handled. I hope that you get a good one.

 

[george w]

We'll see how the pending replacement works out. If it doesn't then I'll write this off as an educational experiment and get a new one installed at the nearest dealer. I haven't got a quote on that but I just feel vulnerable as to what they may want for a new ECU, programmed and installed. If Chrysler offers a reman I would assume it could have the same issues. Probably the generic, programmed on site, ECU is the way to go if the dealer even knows about it. I'd feel much more confident if I could speak to a shop mechanic who may have experience in changing these out rather than most of the service writers who typically seem to know little about the cars other than the basics. I had a great local Chrysler mechanic who did all the service work that I couldn't do myself. He's a top flight guy but he recently retired. He's done major work for me at his home garage, like doing a head gasket replacement on my LeBaron convertible 3.0, but he doesn't have the necessary tools at home to do a ECU replacement.
Is it possible to order the factory replacement over the parts counter already programed so I could install it myself ?
The replacement I received from the Florida rebuilder was preprogramed. I just gave them the VIN and the mileage.

 

[ImperialCrown]

You will want to review TSB (technical service bulletin) #18-006-04B:
http://www.ertyu.org/steven_nikkel/neontsbs/2004/18-006-04.pdf
This TSB describes the generic PCM. Mopar is phasing out pre-programmed PCMs. The replacement is a brand new PCM that is basically a 'blank piece of paper'. It can be told by a software download what it now is and what it can do.
Having one part # cover a whole lot of cars saves money for Chrysler.
The service desk writers job won't be to determine the best method of repair at write-up. You can request a generic PCM replacement/programming and try to by-pass any unnecessary diagnostic charges.
The PCM will probably have to be special-ordered, so you may want to make sure that one is in the parts dept. waiting for you before you drop the car off (or wait for it) while they install and program it to your car.