Fuel vapors and charcoal canisters: need some tech on putting a homebrew system in

Adding charcoal canisters and fuel vapor purge circuits to modified cars appear to be an unknown according to my googles.

I'm converting my Fiat X1/9 to standalone EFI, and I need to be able to deal with fuel vapor from my sealed fuel system. The tank isn't vented, so I need to be able to mitigate expansion/contraction. The stock canister won't work - it's for a carb and has FIVE ports on it. Five. According to my research, the most simple form of a fuel vapor system goes like this:

fuel tank > check valve/fuel separator > charcoal canister > purge solenoid > engine vacuum

If the above is correct, what is the control scheme for when to open the purge solenoid? I can control it with my ECU really easily...but I don't know what operating conditions I should open it under.

Also looking for advice on easily retrofit components - specifically a good purge valve. Is there some nearly ubiquitous GM part that would be good to use here?

I have a similar issue with my TR6.  I probably have all of the parts, but need to know how to arrange them.  Mostly, I'm unsure of when to have the Megasquirt open/close the purge valve.  Listening in on this one...

The VW FI from the 1970's had a charcoal cannister encorporated into it.  Seems like you could just hook something like that up

For example, 914 AFC FI setup

Something like this...

You could vent the intake side before the throttle, but it may give some fumes if it sits for a while.

Another view:

I came in looking for moonshine tips. I'm very disappointed. 

If you can run a purge valve, it's cake.  No purge at idle, then trending up in duty cycle as engine speed and load increases.  Turn it off at wot because there is no manifold vacuum to run it. 

Hit the junkyard for a cannisters from a reasonably modern car, (American- Asians get a bit goofy) hook it to the tank vent, hopefully plumbed so raw fuel can't get to it.  Scrap the vent solenoid, keep the filter.  Enjoy.

Stampie said:

I came in looking for moonshine tips. I'm very disappointed. 

You must watch the documentary "This is the Last Dam Run of Likker I'll Ever Make"

https://youtu.be/glQjCKAI4gA

On the original topic, I'm also interested in integrating old and new emissions systems.  Will be interested to read what you find out.

That’s exactly how I setup my celica. Tank comes into canister. Canister has out to throttle body and solenoid. 

Port on bottom is for overflow. The 1uz takes care of the rest. Go pillage any old 80’s can setup. For reference, a ‘82-‘85 celica canister is about the size of a small Folgers coffee can. 

Streetwiseguy said:

If you can run a purge valve, it's cake.  No purge at idle, then trending up in duty cycle as engine speed and load increases.  Turn it off at wot because there is no manifold vacuum to run it. 

That right there is the data I needed. I can do easily do this with my ECU. Now to find a valve I like. Thanks!

Except you have ITBs.  Vacuum signal will be basically useless for doing much with.

If you have a single TB, then you have a larger and more stable vacuum signal.

Ditch the smart setup and make it fairly simple/stupid.

In reply to Stefan :

What's the simple/stupid option?

Would it work to put a shutoff valve between the canister and intake that only opens when the engine is running and not WOT?  I was thinking about plumbing it before the throttle bodies and relying on the vacuum generated by incoming air going by the port.  But that may not generate enough vacuum to clear the canister.  I also have ITBs and am out of vacuum ports on the intake.

So the EcoBoost engine I'm putting in my TVR has a dedicated vacuum pump.  This has piped connections to the turbo wastegate and to the brake booster.

I was pondering last night if I could pipe a gas tank vent to that.  That got me wondering where the exhaust on that vacuum pump ends up.  Probably straight into the intake at cylinder 4, since it's mounted right on the back/intake side of the head.

Too much vac on the gas tank??

I'm not even clear if the existing tank is vented.  Probably it is and I shouldn't even worry about it.

I'm pretty sure you can't put a vacuum port before the throttle blade. The throttle creates the restriction that in turn causes vacuum.

I would recommend connecting your intake runner vacuum ports to a separate vacuum reservoir, basically a container probably less than the size of a soup can down to a small pill bottle size. Then you can port your vacuum canister to that as well. Another benefit to this is mounting your map sensor to the vac reservoir will really smooth out the vacuum signal to the ecu.

the purge valve from a 92-95 e36/e34 is really easy to work with. one port in and one port out. the electrical connector is a really common connector found from many manufacturers. it is also rather small so it is easy to place where you need it.

One thing to think about is size- you want enough flow that you can detect it (know when you have flow, and know when you are done) but not so much that it's a huge leak that is hard to compensate for.  You can't have flow that makes the fuel too rich or air flow that makes the fuel too lean.

For older gas tanks, you also don't want to draw a big vacuum, as you will certainly collapse the tank.  

Drill a hole in your filler cap? 

Carbon said:

Drill a hole in your filler cap? 

Don't do that, that pretty much eliminates the point of the cannisters.  

Yeah I think there's more to it. The tank or cap might need a 2-way relief valve. 1 pressure to vent tank to atmosphere in case of overpressure (hot day). The other direction to isolate the purge/cannister circuit from the tank until the purge valve comes on. At least that's the way I understand the 96 Crammit system works.

So we're solidly back in the inconclusive category.

I'll have vacuum since I'm putting a small vacuum manifold in, but there still won't be much. I'm going to do some homework into the system the car came with - it should be a mechanical system that relies on a spring-loaded pressure relief valve. Maybe if I can determine which port on the canister does what I can eliminate a few, or better still go to another (read: smaller) style canister. At least one of those ports should be the carb bowl vent, and obviously no longer needed.

The search continues.

In reply to Mezzanine :

Have you ever seen the SPICA injection for Alfas?  There's a little vacuum distribution at the top of it- used mostly to allow a fixed idle air past the throttles.  But it also acts as a place to "dump" the vapor from the cannister.  So there's a way to get enough flow using individual throttles.

For a more modern look- check modern BMW's with VANOS- those should run almost no vacuum, and still manage to deal wtih modern control.

Really, the simple block that my Alfa has should be fine- most modern cars run near WOT kind of manifold pressure these days, and still flow more than enough to control vapors.

In reply to alfadriver :

Thank you, I'll do some research into how BMW solved the problem.

You could connect the canister to the engine side of the air filter the restriction of the AF will cause a slight vacuum to the canister and should move some air through the canister to evacuate it. Not ideal but functional.

Otherwise a purge valve that is closed at idle and open only at part throttle, with some kind of flow restriction.

Canister should not be completely closed, you do not want to apply a vacuum to the fuel tank, or have a direct connection between the fuel tank and the manifold.

The tank breathes through the canister (usually the bottom of the canister is open), the charcoal absorbs the lighter than air vapors, the suction from the motor draws the vapors out to be burned.

The small vacuum leak when the purge valve is open should not be an issue as the motor will be in closed loop at part throttle and will compensate for the air and fuel vapors drawn in.

With 5 ports your OEM canister must have a built in purge valve? These are usually operated by ported vacuum, which is only present at part throttle like spark advance or an EGR control port.

OK, here's the mechanical system the car came with:

Looks like the Fiat engineers used both heat and vacuum to make the charcoal work as well as it can. I might just move the canister closer to the exhaust header to warm it up in lieu of the "hot air purge inlet" since it looks to be atmospheric anyway. There's no port on the exhaust manifold - it just drew air from a heat shield over the manifold.

Having two liquid vapor separators (B) seems redundant, but no need to change that if they work. I'll poke around the remains of my factory system to see if I can determine if everything actually works.