What is this part? (White cylinder)

It's hard for me to understand how the tank wouldn't have baffles in it, but I will probably double check at some point. That would be a major piss-off that I went through all that work to put in a new tank which wasn't designed correctly because of stupidity. In terms of installing it wrong, the thing I was thinking about was maybe the filter on the pump was oriented in the wrong direction, if that's possible and it kind of looks like it is.

I unplugged the CDCV and left it unplugged, and drove to work today (18 miles each way), and I got no additional codes besides the P0455 (either actual or pending). I wasn't expecting that based on what you said.
 
Aftermarket parts can have issues. Maybe they tried to save a few bucks by not installing baffles ???

I dug out my old fuel pump housing and looked at it. It doesn't look like you can assemble it backwards . There's a big white clip that holds the pump in and orients the filter and it won't clip on to the housing backwards... The clips won't line up.
The pump itself only plugs into the housing one way.

I don't know why you wouldn't throw another code with the CDCV unplugged ?? I just know that my CDCV was sticking and I threw that 1449 code.

I think it would be a smart idea to replace your high pressure filter... An old filter would have 15 years worth of crap in it which creates back pressure that can burn out your fuel pump.
You can kill a few birds with one stone when you change it and you won't have any rusted screws because everything is new.
 
I'm willing to give it a shot although I'm inclined to try the smoke test one more time (tomorrow evening first) with the CDCV powered directly. However, I neither see any liquid around the top nor smell any fuel. I guess I should pull it out to look into what's with the fuel cut.

Regarding the CDCV, I figure there are two ways it can detect a CDCV failure, one is check for an open or short (which if it did, it should have detected). Alternately it could detect it based on pressure level in the tank. If it's a big enough leak, perhaps the CDCV cannot affect the pressure in the tank and therefore it can't detect a failure. You'd think of that were the case the leak should be fairly obvious, but like I said, the only leak I noticed was the Port B of the check valve.
 
... However, I neither see any liquid around the top nor smell any fuel...

I'm pretty sure (but not positive) that you can have a vapor leak without having a liquid fuel leak at the fuel sender unit.

Another smoke test is certainly a good idea regardless.
 
I tried yesterday and was stymied by not having a good way to connect wires to it. I'll try again on Saturday, but when I did try to power it up, it was running about half an amp, and I got a nice arc when I first connected the wires. I couldn't clearly hear the valve operate though, which makes me wonder if the valve is stuck open.
 
Still no additional codes. I went to the junkyard and got a CDCV and wiring harness off a 2003 Protege, looks the same except for the mounting bracket. Will play around with them tomorrow (later today).
 
...looks the same except for the mounting bracket. Will play around with them tomorrow (later today).

Yea, I was going to tell you about that... The pictures in the FSM are of the sedan. Our bracket is more square.

There may be enough support from the vent hoses to hold it without using the bracket.... The hoses are pretty heavy.

Guess it's time to remove yours and do the bench test.

 
Last edited:
So, I did some more testing and here's what I found:

1) The CDCV was DEFINITELY not switching. I powered it directly with 12V and although it was drawing 600mA, it made no difference on the smoke test. I could not hear any noises coming from it when powered or unpowered. I also tried removing it, and did not visually see anything happen when it was powered up.
2) I tried the "new" CDCV. It's definitely switching. I powered it up with 12V and I'd guess a nice thunk sound, and see movement of the spring (inside one of the ports), and it blocked airflow. Only issue was that sometimes it'd take up to 2 seconds to switch once powered.
3) I installed the "new" unit in the system -- it definitely controlled the flow of gas through the check valve when powered externally.
4) I powered up the old unit (off the car), and started repeatedly wacking it with a ratchet. Eventually it started switching. It exhibited similar behavior where it sometimes, but not always would have a delayed switch of .5-2 seconds. Other times the switching was faster.
5) The "new" unit seemed to work okay even not mounted to anything. The hoses seemed more than adequate to hold it in place.

I've put about 35 miles on the car (1 trip) -- so far I have no pending codes, and the EVAP monitor is incomplete.
 
Apparently you can "lube" the plunger...

When I took my car in they said come back if it codes again...

Three months later it coded again and I got a new CDCV valve...

You might wanna get a new one...

(Canadian dollars... You guys pay 3/4...)

 
Last edited:
First step is to verify this fixes the P0455. If it does, then I will consider replacing it, but honestly... why bother? If it happens again, I'll know because I'll get the P0455 again and I can just wack it again. It doesn't take much effort. If it gets to the point that either wacking it doesn't fix it, or it fixes it but only for a short time, then I'll replace it.
 
There's a lot of conditions that are monitored...

It might take a while for the code to come back if it does.


 
So I just came back from a drive. The EVAP monitor is ready and no codes are logged or pending. When the CDCV was malfunctioning, the P0455 was thrown as soon as the monitor became ready. All things considered, I'm pretty sure this problem is solved. Now for the problem of the fuel tank... BTW, my low gas light was on (which I thought meant less than 2 gallons left) , and I filled up, and it appeared to take only about 9.3 gallons. But according to the all knowing Google, the tank is 14.5 gallon capacity. I recall it looking pretty similar in size to the original. I guess its possible it was just gas pump weirdness. Fuel gauge was showing a bit over 3/4 full, but I've also noticed its rarely consistent when its first turned on after changing the fuel level.
 
Back