What is this part? (White cylinder)

H

hpmaxim

Member
:
2003 Mazda Protege5
35ja6b8.jpg


Located near the left rear tire, it's part of the fuel system. I have a P0455 (large evap leak). I checked my purge valve and it appeared to be fine. So I put a smoke machine into the hose running into the vapor canister (in front of the car) from the purge valve, and smoke started pouring out of the white part near the back of the car. It seemed to mostly be coming from the rectangular slots, although also from the seam which seemed to be separating the two halves.

I presume smoke isn't supposed to be coming out of there, so does this indicate the part needs to be replaced or does this indicate that something else which is supposed to cut off the flow to it (vent valve?) needs to be replaced?
 
Actually from looking at the manual it looks like its designed to vent positive pressure in the tank to the outside, and if there is negative pressure to draw it in from the EVAP system, which means behavior was probably correct. So, I'm back at square one. Do I try to seal it up with tape and then blow smoke again and hope the culprit is found?
 
Last edited:
Maybe try to feed the smoke in from some other point ?

It's hard to follow the EVAP schematics... They don't really say what the components are supposed to be doing and under what conditions.

A large leak should be easier to find than a small one though.
 
I can't imagine what other point to feed it in. The point I fed it in should pressurize the entire system, except for the purge valve which I don't think is leaking.

From looking at the directions, it's claiming the rectangular slots are PORT B. It says there should be flow from PORT A (the tank) to PORT B when PORT A is pressurized. I think I was pressurizing PORT A and seeing smoke from PORT B. Now, if in fact I was pressurized PORT C there might be a problem. But only maybe, because if PORT C is pressurized then it should flow to A which then might flow to B.

I presume the point of this is to allow you to fill the tank as if there is no where for air to flow when filling the tank it won't fill easily (I had this problem on another car when the vent hose got crimped).
 
I think you are pressurizing port C when you apply the smoke to your solenoid valve.
I don't think you should see smoke coming out port B.

I fit were me, I'd remove the check valve and blow through it... It should take a bit of pressure to blow through A or C...

You may be blowing right out port B with no back pressure.
 
hpmaxim said:
...

I presume the point of this is to allow you to fill the tank as if there is no where for air to flow when filling the tank it won't fill easily (I had this problem on another car when the vent hose got crimped).
Click to expand...

I'm pretty sure the escaping air is supposed to go out the vent tube in the filler neck to "blow" on the sensor in the fill nozzle from the pump to let it know when to shut off.

 
After looking at the diagrams again, it's kinda looking like you pressurized port A ???

Another thing to consider is to smoke test with the ignition on... I know the CDCV turns on and off with the ignition... Perhaps to direct the escaping air to the filler neck vent???
 
Oops, meant to say one of the two... the car was not running at the time. I've thought maybe the purge valve is stuck closed v and I should test it, but it sure appears closed. Its also worth noting I got the car from my father with a leaking gas tank. I replaced it and then let the car sit for a while, so I'd suspect the leak is more likely related to replacing the tank than some other random failure since he didn't mention anything about a check engine light when I got the car.
 
I'm not sure but I think the CDCV stops the vapor from flowing to the check valve when the ignition is on.

I don't think the car needs to be running. (that might not be safe)

That check valve only needs 0.14 psi to operate so you need to buildup a bit of pressure to find any leak.

You can check the solenoid valve by just blowing through it with and without power.

Check where the fuel sender unit connects to the gas tank under the rear seat. There's that big gasket that seals it.

 
Last edited:
Couple comments:

1) I bought the thing off eBay, its the typical paint can with a wire wound resistor partially submerged in mineral oil. There is a pressure regulator on the front of it. I was running it off a compressed air supply. I did not check the output pressure, but it didn't feel high to me. You barely feel anything from more than an inch or two away. I ran the pressure up to about 20psi (input) and the output pressure seemed to plateau by around 10 psi (input). But I don't really trust those numbers. I think the regulator is set somewhere between .5 and 1psi, so I don't think there's an issue.

2) That last diagram was very elucidating. Looks like CDCV definitely needs to be shut off and I need to re-try. I don't remember if the ignition was on, but I think the radio was on, so it might have been. I guess I can try again --- its also possible the CDCV is supposed to shut it off, but isn't either because of faulty wiring or because the CDCV just doesn't work. I'll have to look into that. Hence the leaking out of the evap vent might actually indicate a failure in some other component (CDCV).

3) I was actually thinking the tank seal was the most likely candidate. I could not see a leak from it, but the leak at the vent may have covered up the leak there.

You've really been a big help, that ignition on and the last schematic were great!
 
The CDCV does have a code associated with it which should help to rule it out as a culprit.

It does say that it checks the voltage as soon as the ignition is "on" ...

I actually threw that code and had my CDCV replaced.. It was sticking open or closed (I forget which) and the ECU "knew"...

 
By the way PCB, after changing the gas tank (with an aftermarket one) I've noticed that I seem to be undergoing fuel starvation while making prolonged hard right turns when the tank is nearly empty. I can't remember if I used a new fuel pump or not when I put the tank in, but I have to assume either they bungled the internal baffles (which seems unlikely to me), or is it possible that the fuel pump needs to go in to the tank in a certain orientation (I don't remember what the fuel pump looked like, but I know some are connected to the removable lid through a fuel hose which means the pump could get rotated. Any thoughts?
 
I pulled the electrical connector off the CDCV. I tried both with the ignition on/car off, and car on and there was no voltage across the terminals. It may have briefly pulsed on when I turned the key, unfortunately I didn't have a second person and I can't be in two places at once -- but it didn't seem to ever turn on. I believe that passage is normally open and closes when energized, which explains why I was seeing a leak through the vent. Question is, is this normal behavior or do I have another problem? Do I have to just put 12V across the CDCV and re-test?

Also, as for the fuel pump, yeah, I was replacing the tank so only that metal ring was ugly, and I think that I might even have gotten a new one with the tank. The question is, is it possible to assemble the thing wrong and create the fuel cut symptom? I just saw that new plastic piece at Rockauto for $28, might have been wise to put that in, I know I didn't do that.

Looking at the service manual it unhelpfully says:

Connect WDS to DLC-2
Turn ignition on
Close the CDV using ON-BOARD DEVICE CONTOROL [sic] function.
NOTE: The CDCV is closed for 10 minutes UNLESS the following any action is done [sic]:
-- The engine is started
-- The ignition key is turned off.
-- The fuel tank pressure sensor signal exceeds 6.43 kPa.

Since I don't have a WDS (whatever that is) (and I bet most independent mechanics don't either) I guess you either have to jury rig it with a power supply/battery or go to a dealer. Great!
 
Last edited:
hpmaxim said:
... Question is, is this normal behavior or do I have another problem? Do I have to just put 12V across the CDCV and re-test?

... I guess you either have to jury rig it with a power supply/battery or go to a dealer. Great!
Click to expand...

It's hard to know what normal behavior is... The FSM doesn't really say...

I think you should run 12 volts right to the CDCV.

You'll probably hear it clicking which will verify that it's probably working and then you can close it to try another smoke test.

You would probably throw that 1449 code if your CDCV wasn't working.
 
There's a couple simple tools you need to pick up...

I'm sure your local Autozone has a few in stock.

 
hpmaxim said:
The question is, is it possible to assemble the thing wrong and create the fuel cut symptom.
Click to expand...

I don't think so... It really only goes together one way.

The only thing I can think of is if a generic fuel pump was installed... It comes with hoses to connect it but you'd probably remember seeing that.




I think you should order a new high pressure filter (fuel pump housing) . Yours could be old and cracked causing your vapor leak.

You can check the seal when you install it and you can see if your gas tank has baffles in it... (I'll bet it doesn't or the ones it does have don't work.)

 
Last edited:
Please log in or register to reply here.

Similar threads

N
Seeking advice for my Protege5 error codes P0660, P0401 and P0456
Replies
3
Views
2K
nomadicsoul
N
leandoer
Are these Vapor canister purge valves the same..?
Replies
4
Views
797
leandoer
leandoer
T
O/D Light on at cold start. Can anyone explain this? 2002 Protege5
Replies
13
Views
2K
TX_SilverP5
T

Latest posts

Back