Purge control solenoid alternative
 

My 93 was converted somewhat shoddily to non sequential by the PO and I found the purge control solenoid to be no good. The precontrol solenoid shares the same connector and barb diameter although it is a physically smaller solenoid, would it be suitable for purge control?

I'm willing to bet in the past 30 years that no purge control solenoid in the history of the fd has ever done it's job. Not because of being broken but because the system has never had enough fuel in that little canister under the throttle body to need to purge it. My suggestion is to not worry about it and delete it if you are able based on your ecu. It is an emissions item but it's not a functional one.

The tank under the throttle body is a filter for the fuel vapors coming from the charcoal canister. You can try the valve but I couldn't find any way of verifying that it would work. I suppose the best way would be to check for vacuum at the solendoid and verify it opens when commanded by the pfc.

The purge control is about $65 bucks new if that doesn't work. Part number F285-18-741A

nothing comes FROM the charcoal canister. Theres 2 lines on it: the first is the 3rd "fuel" line coming from the engine bay and the second vents to atmosphere. The charcoal canister is a dead end. For the pfc to command the solenoid to do anything you would need to know what condition needs to be met for that to happen. Under what circumstance does the stock ecu command the purge control to do anything?

this particular solenoid isn't very useful. It's for a system that more than likely hasn't done anything since birth. There are ways to bench test the solenoid to verify it's operational but it's like troubleshooting the exhaust over temp light.......

From page F-129 of the service manual

"The evaporative emission control system temporarily stores in the canister the evaporative fumes generated in the fuel tank. The stored gas is then pased into the air intake system for combustion when the engine is running. This operation prevents evaporative fumes from flowing out to the atmosphere

Sending a large volume of evaporative fumes at one time into the air intake system deteriorates the airfuel ratio; thus, the ECU uses the solenoid valve (purge control) to regulate this volume"

"Operation
With engine stopped and no load applied:
The evaporative fumes from the fuel tank are absorbed by the charcoal canister

With engine running and load applied
The evaporative fumes absorbed by the charcoal canister are drawn into the engine via the solenoid valve (purge control). The volume of fumes drawn depends on engine conditions"

Fumes are collected in the canister from the fuel tank, and burned off by the engine when the ECU decides it makes sense, by triggering the solenoid. This prevents fumes from being vented to atmosphere.

It's a system you can technically live without, but it certainly isn't useless.

From the factory the charcoal canister is tuned to run on the EPA city cycle (FTP 75) which mimics stop and go traffic in LA on a warm day. In the EPA test the canister is intentionally loaded beforehand with vapors to confirm it is working during the test. This is basically set according to government regulation. The requirements weren't as stringent in 1992 when the car was certified but I can assure you they switched from the mechanical control of the 2nd gens to a duty solenoid for better emissions and driveability.

I'm not sure how the PFC controls the canister because the evap solenoid duty cycle is not a loggable parameter, probably because the PFC is really old and they had to make a decision on what could be tuned and logged within their RAM and sample rate constraints.

If you vent it to atmosphere you will smell fuel vapors on a hot day as the fuel boils. If you don't care about that, it's not a big deal. If you want a "normal" car, just buy a new solenoid from Mazda and be done with it.

See orange line on the right side of the diagram below. The canister goes to the solenoid valve, which is connected to manifold vacuum. The vacuum pulls the fuel vapors into the engine according to the duty cycle commanded to the ECU. I presume the stock ECU at least has some compensation in the fuel calculations to account for the additional fuel vapors, so that it doesn't run too rich when the solenoid is purging vapors. At least, that's how modern cars work. It's also possible that the purging system runs under such narrow conditions that they could get away without any compensation.

https://cimg7.ibsrv.net/gimg/www.rx7...61a0595a17.jpg

https://cimg9.ibsrv.net/gimg/www.rx7...34f0c180dc.png

Well now.......... =D

I have a haltech elite that doesn't control the purge.
should i just vent it or run the catch tank with a check valve to uim?

@arghx , are there considerations besides running too rich? I want to start using my purge control solenoid to avoid the fuel smell and pressurized fuel tank on warm days. I can adjust the duty cycle (AEM Infinity standalone) based on Engine Speed and Manifold Pressure to keep the fuel delivery and O2 feedback from getting out of hand, but is there any risk of damaged parts from pulling too much vacuum? I'm still trying to make sense of the check valves and diagrams, would the engine pull air through the bottom port of the charcoal canister after it has depressurized the fuel tank?

Here's what I would do. Go drive the car in 4th and 5th gear from say 40 to 80 mph. Look at the rpm and manifold pressure you run at holding steady rpm (flat or slight incline, no boost). Set those areas of your purge solenoid PWM map to 100% duty. After making those changes go drive the car around and see if you can tell any difference in driveability. If you can't tell any difference, and you don't see huge changes in AFR, consider it done.

You don't want it to purge off idle or during major acceleration. You just want low load light cruising, and you need to have some vacuum to draw the vapors out. Your goal is to not actually improve emissions, you just want to empty the canister before it fills and make it imperceptible to the driver. Go do that and report back. If I had to guess, maybe 40kpa to 80kpa ish absolute pressure and say 2500-3500rpm would do the trick. If the MAP is much higher than that, you don't have much vacuum, and if it's lower, you're in decel and pulling against a closed throttle. I'm pulling those numbers out of the air though.

There's no risk of damaging anything. I mean if it pegs 10.0:1 rich and gets rich misfire, that's not good, but I'd be surprised if that's an issue. When the car was new it was tuned so that the stock ECU can account for the fuel volume and not go too rich, which would increase CO and possibly HC emissions. The way it works is that in a Federal Test Procedure (the EPA city cycle), the canister is loaded with several grams of vapors by a technician before the test. The vehicle is run on a dyno in a controlled 75F environment, and in the middle of the test the engine is shut off and the hood closed. That generates fuel vapors which fills the canister further. It gets to the point that if the purge isn't tuned right, emissions will increase and the car will fail, and the car isn't allowed to be sold if it fails emissions. None of that is your concern. You just want to get rid of the smell and not have any driveability issues.

I dug through the service manual, and according to the ECU pinout on page F-158 the purge solenoid runs from 1500-3300 rpm, and does not run at idle. And according to F-132, one of the valves in the purge system takes 6 kPa to open. So yeah, try having it purge from 1500-3300 (or 1500-3500, whatever), 40 kPa to 80 KPa absolute. Maybe set it to 100% (ok not 100, but like 95%) or even try 50%. I'm not sure how much effort would need to be made for it to be a smooth transition of the duty cycle map. What you could do is ramp the duty cycle by rpm though. so 1500 rpm could be 50% duty, 2000rpm is 60%, 2500 is 70%, 3000 is 80%, 3500 is 95% , all within 40 to 80 kPa. If you're concerned about affecting off-idle area, narrow the window to say 2000rpm instead.

If this is too rich, scale that whole map back in 10% increments. so 50% becomes 40%, 70% becomes 60%, etc. Just keep dropping it in 10% increments until you don't see it go rich in those areas. However if your canister is full it's going to purge more and get more rich, so you don't want to chase your tail trying to get AFR right. You said you had closed loop fuel so that should be good enough. Modern cars can sort of calculate how full the purge canister is so that it doesn't overshoot and cause an emissions problem. You've got a simple speed/load look up table, which is plenty for your purposes.

Also, I cross referenced with some data I have from a V8 piston engine naturally aspirated engine which is much newer, and that is running purge from about 1000-2500rpm, 30-80 kPa, but that's on a stock ECU in a modern OBD 2 car, and it's a V8 that doesn't rev.

As for the plumbing, just hook it up like the factory diagrams is all I can say for now. The check valves and catch tank and charcoal canister basically all hold the pressure in. There is a natural build up of vapor pressure in the fuel tank, and then the solenoid applies manifold vacuum to draw the fuel vapors in. The actual regulation of purge flow is very simple; it's flow through an orifice, not that different from a carburetor. The complexity in evaporative emission systems comes from all the anti-leak technology. All the crap that makes your check engine light come on with a loose gas cap is what makes it complicated.

this is basically the closed loop area, so the O2 sensor would compensate for the extra fuel, yes?

you would think so. I imagine being such an old car the stock ECU has a very crude controls that are mostly reactive. Nowadays the software is pretty elaborate as you can't just sort of go rich for a second and wait for the o2 sensor to catch up as that would fail emissions.

Thanks for the advice, I set it up similar to you described. The solenoid is active between 1500-5000 RPM and 40-85 kPa, with max duty cycle about 70%. It still drives well, but the tank was still pressurized after a short 5-minute drive. The tank was probably pressurized before driving, but I didn't want to open the gas cap before driving in case that would mask drivability problems.