The epic saga of my 2002 Sonata GLS trying to pass emissions took another weird turn. Seems to be a stumper for the dealer. After it's all said and done, the dealer who's been working on my car for 3 months seemed to figure out my P0456 code but now the only code being thrown is P0442. The last thing they did to stop P0456 from happening was to drop the tank, check it, check the connections, hoses, etc and put it back. It also seems they replaced a hose to the canister but it didn't help. The last thing we did after my MIL went on was to replace the gas cap. The MIL is still currently on but it's only been since Friday that we changed the gas cap. I would imagine if it was the cap, MIL would go off by itself after so many miles if it was the cause of an evap leak, right?
P0442 was thrown after I spent a bunch of time trying to get the Catalytic Converter Monitor and Evap Purge Monitor to set as Ready by way of TSB 0236030 (which was *REALLY* hard to complete successfully):
formatting link
The P0442 code popped up the *second* I completed Step 6 on page 10 which is two successful drive cycles. The result from the emission test showed the Catalytic and Evap monitors as being Ready but of course I still failed the test because of P0442.
After 4 smoke tests, tons of time and plenty of hoses replaced, the dealership says there is no leak and it's just the computer being wonky (read: false positive). Do you think that makes sense? Anything else that might set this code? Anything else they might be missing?
The only difference between a P0442 and a P0456 is the size of the perceive= d vapor leak. P0442 represents a "small" leak and P0456 represents a "very= small" leak.
My take on this is that your car is just plain not fixed, and, in fact, the= leak is now a little worse.
The Evap monitor is the one that conducts the tests for codes P0442, P0455,= and P0456. That is why the code set immediately upon completion of this m= onitor. The dealer has the capability with their diagnostic equipment to m= ake this test (monitor) run. So, really, I'm at a loss as to why the car w= as given back with the problem occurring; the dealer can test it to make su= re it's fixed prior to releasing the vehicle. Well, I'm at a loss unless t= hey're basically giving up.
As for the idea that this is a "wonky" computer routine, I don't buy it. I= f that were the case, I'd have 1999-2005 Sonatas rolling through all the ti= me with this issue. (And I don't.) Nor do I buy a suggestion that the com= puter is defective and needs to be replaced. Sure, lots of things are poss= ible, but if I can say I've seen or heard of exactly zero cases (from credi= ble sources) that a small leak DTC was repaired by a computer replacement o= n a Hyundai with a 7-year production run that ended 7 years ago, that's a v= ery good indicator it won't be the issue.
I don't know whether you have the 4-cylinder or V6. The commanded evap tes= t results on the V6 are in graph form, and if you know what you're looking = for, they're easy to interpret. On the 4-cylinder, you just get a message = from the diagnostic tool regarding whether the test has passed or failed. = =20
Has the canister close (or canister vent) valve been replaced? A leaking c= anister close/vent valve is often missed during a smoke test because it's n= ormally open and must either be blocked or manually actuated shut to check = for the leak. And in fact, that's the purpose of the valve in the first pl= ace, to close the system so that the ECM can use the engine to vacuum the f= uel tank and monitor the pressure (vacuum) to check for vapor leaks.
I don't think I just jump to replacing it without some test results that ba= ck me up (such as evap test passes when vent hose is manually blocked while= the close valve is actuated but fails when the close valve is allowed to b= lock the system on its own). I do think, though, it's something worthy of = specific checking if it hasn't been ruled out.
On Monday, December 3, 2012 10:08:19 AM UTC-5, Thee Chicago Wolf [MVP] wrot= e:
Thanks so much for the reply and suggestions. I went back and looked at the receipt for the work I had done in September when my fuel lines and brakes lines rotted out. In addition to the lines, they changed out the Canister Assembly (31420-38100), Canister Close Valve (31430-2920), and Purge Control Vale (28910-2240) and I am not sure if they did the smoke tests using the method you suggested. You're thinking maybe they put one of the valves in wrong or the valves are just defective and not acting right? Seems in line with what you mentioned.
On another receipt when I went back in for MIL a month after the work was done, it states on the report that "All components working as designed per GDS scanner: activation tests of purge and canister close valves pass." So am I to interpret that as they simply activated the valves to test that they work and not smoke tested with the valve(s) actuated like you said? I'll bring all this info to them if my MIL comes back on again. The service manager has been really good about me coming back in a bunch of times for this and I am lucky that I got a 4 month extension and waiver for my emission test failing so if it turns out it is what you stated below, they can figure it out and it should be fixed for good. After all, warranty on the parts and labor is one year.
vapor leak. P0442 represents a "small" leak and P0456 represents a "very small" leak.
leak is now a little worse.
P0456. That is why the code set immediately upon completion of this monitor. The dealer has the capability with their diagnostic equipment to make this test (monitor) run. So, really, I'm at a loss as to why the car was given back with the problem occurring; the dealer can test it to make sure it's fixed prior to releasing the vehicle. Well, I'm at a loss unless they're basically giving up.
that were the case, I'd have 1999-2005 Sonatas rolling through all the time with this issue. (And I don't.) Nor do I buy a suggestion that the computer is defective and needs to be replaced. Sure, lots of things are possible, but if I can say I've seen or heard of exactly zero cases (from credible sources) that a small leak DTC was repaired by a computer replacement on a Hyundai with a 7-year production run that ended 7 years ago, that's a very good indicator it won't be the issue.
results on the V6 are in graph form, and if you know what you're looking for, they're easy to interpret. On the 4-cylinder, you just get a message from the diagnostic tool regarding whether the test has passed or failed.
canister close/vent valve is often missed during a smoke test because it's normally open and must either be blocked or manually actuated shut to check for the leak. And in fact, that's the purpose of the valve in the first place, to close the system so that the ECM can use the engine to vacuum the fuel tank and monitor the pressure (vacuum) to check for vapor leaks.
me up (such as evap test passes when vent hose is manually blocked while the close valve is actuated but fails when the close valve is allowed to block the system on its own). I do think, though, it's something worthy of specific checking if it hasn't been ruled out.
On Wednesday, December 5, 2012 9:42:45 AM UTC-5, Thee Chicago Wolf [MVP] wr= ote:
If the person performing the smoke test actuates the close valve to close i= t, then a faulty close valve would be apparent (presuming it's not an inter= mittent problem and just not acting up at the moment). On the other hand, = if that person simply blocks the close valve outlet, they'll never know if = the close valve is the culprit.
If we're in a scenario where the system sometimes tests okay (during the ve= hicle's own evap self test) and sometimes fails, then it's very likely that= a moving component is the source of the problem, and I'd go straight to th= e close valve (as this is the only moving component that can cause a vacuum= leak in the system).
The verbiage regarding the GDS test seems to suggest that each of the compo= nents was actuated individually (which this device also has the capability = of doing). What cannot be done is to do this with the engine running, so w= e're not really testing sealing or anything like that-- just that the plung= er or solenoid moves. The evap self-test gives a much better idea of the i= ntegrity of the system as a whole, but as I said earlier, if you have the f= our cylinder, the information provided is still much less than desired.
The smoke test is only particularly useful for finding a leak once you know= one exists. Basically, the system is filled with pressurized smoke and cl= osed to the atmosphere. The technician then looks for any smoke emanating = from the system. My efforts to find leaks with this method have been very = disappointing, to say the least. More often, I'm able to successfully loca= te leaks by blocking off various parts of the system and re-running the eva= porative self-test. I don't recall how feasible this is on your particular= vehicle.
then a faulty close valve would be apparent (presuming it's not an intermittent problem and just not acting up at the moment). On the other hand, if that person simply blocks the close valve outlet, they'll never know if the close valve is the culprit.
vehicle's own evap self test) and sometimes fails, then it's very likely that a moving component is the source of the problem, and I'd go straight to the close valve (as this is the only moving component that can cause a vacuum leak in the system).
components was actuated individually (which this device also has the capability of doing). What cannot be done is to do this with the engine running, so we're not really testing sealing or anything like that-- just that the plunger or solenoid moves. The evap self-test gives a much better idea of the integrity of the system as a whole, but as I said earlier, if you have the four cylinder, the information provided is still much less than desired.
exists. Basically, the system is filled with pressurized smoke and closed to the atmosphere. The technician then looks for any smoke emanating from the system. My efforts to find leaks with this method have been very disappointing, to say the least. More often, I'm able to successfully locate leaks by blocking off various parts of the system and re-running the evaporative self-test. I don't recall how feasible this is on your particular vehicle.
HT,
Sorry that I forgot to mention that I have a V6. Thanks again for all this detail. Sine I have no idea how the techs are performing the smoke tests and other methodology, I can pass this along to them and see what they say.
On Thursday, December 6, 2012 10:53:33 AM UTC-5, Thee Chicago Wolf [MVP] wr= ote:
If my memory is correct, the nice thing about the V6 is that the evaporativ= e self-test runs very quickly, and the diagnostic tool (GDS) will give a gr= aphic display of the results. So the technician should be able to do repea= ted tests and check for variation in the results beyond pass/fail.
self-test runs very quickly, and the diagnostic tool (GDS) will give a graphic display of the results. So the technician should be able to do repeated tests and check for variation in the results beyond pass/fail.
So ultimately the two parts that are likely the culprits are Canister Close Valve (31430-2920), and Purge Control Vale (28910-2240), right? Are either one of those two more likely than the other? It ought not be the Canister Assembly (31420-38100) because it has no moving parts? Thanks again for all the info.
On Tuesday, December 11, 2012 11:13:01 AM UTC-5, Thee Chicago Wolf [MVP] wr= ote:
Only if the problem is intermittent, and then only suspect the close valve.= The purge valve can stick open, but that'll result in more vacuum to the = tank, meaning a different problem than a leak will be perceived. The purge= valve can also stick closed, but the system will see this as a large leak = as it will be impossible for the engine to pull a vacuum on the tank.
If the problem is continually present, look for a leak at any source.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.