Dear All –

First post here, so please allow for the usual awkward descriptions and improper lingo. A work friend has a nice 1985 GSL-SE that hasn’t run since 2022. From what he described, it sounded like the car suffered from the typical flooding issues RX-7s are known for, but at the time, he wasn't aware of the "unflooding procedure." As a result, the car sat indoors, and he siphoned out the remaining gas from the tank (non-ethanol 91).

Fast forward to today:

After adding five gallons of 91 and charging the battery, we attempted both unflooding procedures detailed here (regular & complex): Rotary Performance Technical Articles — but had no luck.

Of note, during Step Three of the complex procedure:

STEP THREE
Make sure all four spark plugs are pulled, blue connector disconnected, and ignition wires are away from accessories and belts. Get inside the vehicle, hold the accelerator to the floor, and crank the engine for approx. 15–20 seconds. The engine will not start during this procedure. Note: You will see lots of gasoline vapor misting in the engine compartment. This is normal and will diminish during the cranking process.

We did not see any gasoline vapor misting in the engine compartment.

With that, I turn to the group: Should we start looking at fuel filter or fuel pump issues? What’s the lowest-hanging fruit we can target next for diagnostics? Any input would be appreciated, as this is my first venture into rotary engines.

Best regards,
MJA

 

1. Make sure you have fuel to the injectors.
2. If no fuel, start at the fuel filter under the car.
3. If not fuel still, check the pump.
4. If no fuel still, it's either a bad pump or clogged/bad line in the tank.
5. If you have fuel at the fuel filter, check the filter in the rail.
6. Make sure you have spark.

This is not a complete list but it will get you started. There are also some good troubleshooting steps in the factory server manual here: Foxed.ca

 

Thank you! Much appreciated.

One thing I forgot to mention: we did get the car to fire up for a second by spraying starter fluid into the intake manifold through the air filter. With that, I’m pretty confident we have spark—it’s fuel that we’re missing.

We ordered a Haynes today, but if there is a quick how-to for checking the fuel at the injectors (i.e. where are they?) it would be appreciated. Again, sorry for the newbie questions/comments. We will be armed with a service manual soon.

Regards,
MJA

 

Page 4B-15 https://www.foxed.ca/rx7manual/manua...ssions_13B.pdf

The Factory Service Manual is free from this link: https://www.foxed.ca/index.php?page=rx7manual#firstgen

 

First off, welcome aboard. Secondly, your GSL-SE is a somewhat rare version in that most of the other information around the net and on this site is for carbeurated engines. That said, here are the common factors when you suspect EFI fuel delivery problems;

1) Trailing Ignitor is dead - easy test; remove blue spade connector from Trailing Coil and move it to the corresponding position on the Leading Coil. Start the car normally by pressing accelerator pedal to the floor ONCE, and then RELEASING IT. Crank as normal until it starts. What this does is moves the EFI Injector Trigger signal from the normal coil that it uses to your other one. The Injector firing time will be off, but the engine will still run acceptably well.

Alternately - Swap Leading and Trailing Ignitors at the Distributor Cap, leave all other wiring intact. If the car starts and runs = bad Ignitor (*the one that was in Trailing position, now in Leading position) OR it's a bad Trailing Coil. J-109 Ignitors are getting hard to find, but I always carry a spare in my travel kit. Ruling out a bad Coil just takes some testing and swapping of leads.

2) Fuel Filter Clogged - The SE uses an all-metal EFI Fuel Filter designed for the high pressure coming out of the pump (44psi-60psi), and These can get clogged badly enough to prevent fuel flow. Particularly sensitive is the micro-screen inlet at the Fuel Filter which collects every little contaminant.

There are several filters in line to the pump in this order: gas tank -> Fuel Feed line to Pump (*plastic screen filter, coarse) at the Fuel Pump inlet -> Fuel Pump (*unlikely to be a problem) -> Fuel Filter as stated above -> Fuel Line -> Fuel Rail -> Fuel Pressure Regulator (*not an issue here) -> Pulsation Dampner -> Fuel Return Line -> Tank. There's also a sock filter on the in-tank fuel pickup, but it's not likely to be a problem. Don't monkey with the 2× matching Pulsation Dampners - one on the pump and one on the rail - these rarely fail, but removing one can cause harmonic pulses that hurt performance.

3) Fuel Injectors Clogged - The SE has a Fuel Injector Test Plug at the Air Flow Meter at the front of the car behind the intake air box and filter. This connector is 2-pin plug, and if you run a shunt between the 2 plug ports, it turns on the Fuel Pump any time the ignition is on. This is risky, because leaving it shunted will lead to flooding, as the Injectors have pressure anytime the key is on. MAKE SURE YOU REMOVE THE SHUNT. By using it, you're able to hear the pump run, hear fuel flowing through the Rail, and can confirm that the Injector connections aren't leaking. Clogged Injectors can be sent to RC Engineering to be rebuilt. The SE uses low-impedance Injectors, 2× 680cc/min nominal flow - these are nearly impossible to find nowadays, as staged injectors are usually high-impedance and lower flow (*with more of them).

Fuel Injector cleaning will improve performance and ensure they're running properly. It's worth sending them off.

As more information, the above troubleshooting steps will work with or around what's called a Check And Cut Relay (*aka Fuel Pump Relay) that Mazda built into this system. It's located under the passenger football, but it's not likely to be the problem. This Relay serves to operate the Fuel Pump only when needed to allow a start signal from the starter to engage the relay OR the Air Flow Meter switch to engage the relay, depending on which is prevalent and correct for the engine running state.

Look into the above steps FIRST before doing anything electrically related to the Fuel Pump, as it's probably working fine, and when it should. Good luck, report back, and search on my Username and "SE"+ whatever your problem is, and you'll have 25yrs of board info to read...

Lastly, now that this is posted, scroll down a bit to RELATED TOPICS and I'll bet there's some gold in there to be mined, too!

 

Thank you so much for the responses — hugely appreciated.

We made good progress last night. Right out of the gate, we tried your suggested procedure:

1. Trailing Ignitor is dead – easy test: remove the blue spade connector from the trailing coil and move it to the corresponding position on the leading coil. Start the car normally by pressing the accelerator pedal to the floor once, and then releasing it. Crank as usual until it starts. What this does is move the EFI injector trigger signal from the coil it normally uses to the other one. The injector firing timing will be off, but the engine should still run acceptably well.
   

Unfortunately, no luck.

We were able to disconnect the fuel main and return lines from the hard pipes in the engine bay and confirmed our suspicion: no gas — they were dry.

Next, we jacked up the back of the car to focus on removing the fuel filter. One question came up: with the five gallons of fresh 91 octane we put into the tank, would we get a gusher when removing the fuel filter? As a precaution, we clamped what we believed to be the fuel tank output line.

We then soaked the fuel filter’s banjo bolts with PB Blaster. Once they loosened up, we got them out and a good amount of old gas drained. With the filter out, we did a blow test on both sides of the filter, and a ton of reddish goo came out of the inlet side — we’re hoping this was the culprit. From the engine bay, we also blew air back through the main and return lines to clear out the rest of the old gas from the now-open banjo bolts.

As a test, we removed the clamp on the fuel tank output line, but no gas came out of the open banjo bolts. Is that normal? In light of this, is there any tank cleanup we should do before installing a new fuel filter? We were hoping we could open that line and let the five gallons drain into a bucket to make sure the lines are free from crud.

Any thoughts are welcome, and thanks again to everyone who’s lent a hand!

Best,
MJA

 

The gas in the tank won't drain from the filter inlet side. It gets sucked from a pickup tube in the tank. There is a drain plug on the driver's side rear on the bottom of the tank. Sounds like a tank inspection and cleaning may be in order. You could first drain the gas and then use an air hose to push air into the tank and listen via the fill cap. Even if you do hear air, a tank cleaning might be in order. The level of tank cleaning really depends on a visual inspection via the tank pickup tube assembly being removed. Also, a good time to replace every fuel hose on the tank.

 

You seem to have accurately identified a Fuel flow issue. The reddish goo which was blown out of the Fuel Filter is likely rust, and will certainly gum up the flow enough to prevent injector pressure from building sufficiently. As stated above, drain the tank and inspection is your best bet - easier if you have a fiber optic camera you can poke down the fill hole, otherwise its removal and inspection through the in/out access cover. Note that the drain bolt (1×12mm) is directly under the fuel pickup tube and sock filter, so you may get lucky and your sediment is right there and comes out first. Check your drain pan for what it might be.

At least the Fuel Filter has caught this crap before it entered the lines and the Injectors. Get TWO new filters when you order them; replace the one, and keep the 2nd on standby because you'll probably have to change it twice. If you don't, you'll have a spare and you'll eventually need it down the road. Also, recall this part:

Me: The SE has a Fuel Injector Test Plug at the Air Flow Meter at the front of the car behind the intake air box and filter. This connector is 2-pin plug, and if you run a shunt between the 2 plug ports, it turns on the Fuel Pump any time the ignition is on. This is risky, because leaving it shunted will lead to flooding, as the Injectors have pressure anytime the key is on. MAKE SURE YOU REMOVE THE SHUNT WHEN YOU'RE DONE TESTING. By using it, you're able to hear the pump run, hear fuel flowing through the Rail, and can be sure you're getting good flow. It's hard to hear once the engine is cranking.

Also, now you know the proper engine startup sequence: push gas pedal to the floor ONCE and FULLY (*sets Fast Idle Cam), release the pedal completely, crank the starter, let it idle. There is no accelerator pump like on a carb. Pumping the pedal does nothing but drop manifold pressure as it's trying to auto-choke the intake.

Let us know how it goes,

 

That screen is cone-shaped, and can be easily overlooked. It's usually at the end of the hose at the fuel filter, so look up inside there. Needle-nosed pliers can make removal easy. Some gas will run out when you do this, so have a coffee can, or other small pan to catch the spilt fuel. Hose it out with some carburetor cleaner. I'd recommend putting it back in the way it came out, especially since you seem to already have crud in the tank from lack of use. Yes, it'll clog easily, but yes, it's also a pretty easy thing to clean out and not burn up so many fuel filters.

Our 84 GSL-SE had rust in the tank, which clogged that screen. I got to where I could R&R that screen in about 5 minutes, so it's not that big a deal. A floor jack, 10mm socket to pull the shield away from the pump/filter assembly, slip-joint pliers to do the clamp and some carburetor cleaner to hose out the screen is all it takes.

Good luck!

 

FYI - the screen filter you're talking about is clear plastic and located inside the feed hose to the Fuel Pump. The micro-screen Filter I'm talking about is inside the metal inlet tube of the Fuel Filter, is metal screen with very small pores, and catches rust and any other very small contaminants. You can see it looking into the Fuel Filter and is a pre-filter for the main cartridge.

That clear plastic cone filter won't catch rust. It's gaps are too big, and it's sole purpose is to prevent anything big enough to block or stall the pump from getting to it.

With Fuel flow issues, really need to check them all - and I think he's on the right track.

 

Thanks, all. Last night we had good success with removing the fuel tank drain bolt, dumping the fuel, straining that fuel, and then using it to flush the tank three more times. I think we will try to to this a few more times as we await the arrival of the new fuel filter. It's satisfying to get that crud out of the tank.

Regards,
MJA

 

That looks like sediment to me - what did it look like to you? The sediment that builds up in the tank is primarily from fuel contaminants during refueling, but could also be fragments of the rubber and composite gaskets from the fuel tank connections and such. These deteriorate over time and start to fall apart, dropping into the tank. This is why using that drain bolt usually pulls out the sediment, and also washes gas past the sock filter on the Fuel pickup tube which usually cleans it up a bit. Given that there are baffles inside the tank to help prevent slosh and starvation during cornering, your rinse using the old gas probably helped, too.

I'm betting you guys have gotten this solved, and I hope you took my prior advice about buying TWO fuel filters, because you may find that the replacement gets clogged up quickly from the last of it in the tank and lines once it's flowing smoothly again. I always keep a spare Fuel Filter around just in case because it's a cheap and easy fix when you start to notice symptoms.

Report back once you get it up and running. You've learned a ton already,

 

It's been a minute since I've been under there doing that job, so I'll trust someone ore up to date on the matter. That cone DID catch stuff from our gas tank, so it did its job. Symptoms, since it was a DD car, was that the speed we were able to maintain on the highway would drop. 65, then 60, then, 55, ..... Luckily, the last mile of the drive home was less than 45. And even more-lucky, it never happened on the way TO work, just homeward bound when it happened on the road.

But yes, he's on the right track. Stick a magnet into that sediment to see if it's just dirt, or actually some stuff flaking off the inside of the tank.

 

Thanks, all. I agree, sediment seems to be what is coming out now.

After a few more flushes with gas our next action-item will be to replace the fuel filter with new and then I was hoping we could turn the car on (key in the on position) to see if we are getting gas from the fuel main line (line removed from the hard-pipe in the engine bay). With the car in the on position will this actuate the fuel pump, or must we jumper the test plug, as described below, to run this diagnostic?

Me: The SE has a Fuel Injector Test Plug at the Air Flow Meter at the front of the car behind the intake air box and filter. This connector is 2-pin plug, and if you run a shunt between the 2 plug ports, it turns on the Fuel Pump any time the ignition is on. This is risky, because leaving it shunted will lead to flooding, as the Injectors have pressure anytime the key is on. MAKE SURE YOU REMOVE THE SHUNT WHEN YOU'RE DONE TESTING. By using it, you're able to hear the pump run, hear fuel flowing through the Rail, and can be sure you're getting good flow. It's hard to hear once the engine is cranking.
If gas is flowing, put the line back on the hard-pipe and move forward with the engine startup sequence: push gas pedal to the floor ONCE and FULLY (*sets Fast Idle Cam), release the pedal completely, crank the starter, let it idle......


Thank you again,
MJA

 

To be very clear; the Fuel Pump on an SE will only run without the engine turning if you short the Test Connector that I referred to above. This Test Connector was provided by Mazda for techs to test the integrity of the Fuel Rail before reinstalling the Upper Intake Manifold. BE SURE TO REMOVE THE TEST SHUNT AFTER YOU'RE DONE TESTING.

More; the Test Connector actually closes the microswitch inside the Air Flow Meter which would normally be closed by the Air Door opening as when air is being sucked into the intake. As soon as the Air Door opens, the switch allows continuity to the Check & Cut Relay (*aka Fuel Pump Relay), and turns on the Fuel Pump. If the engine stops, air is no longer being sucked into the intake, the Air Door closes, and the Fuel Pump shuts off. This prevents flooding. If you want to test this explanation, remove the Air Filter cover and Air Filter, and with the key in the ON position, stick a chopstick or pencil into the mouth of the Air Flow Meter and push the door open gently - the Fuel Pump will turn on. Pull the pencil out, and the Air Door closes - the Fuel Pump turns off.

I can't stress this enough that when you're done verifying fuel flow to the Rail and Injectors, REMOVE THE TEST SHUNT (*and throw it back in your toolbox). Every year, a few people show up asking why their SE is flooding regularly, and inevitably someone has read the FSM and decided that they want the Fuel Pump to run whenever the key is on, or forgot that their mechanic tested the fuel rail, and then it floods all the time. Removal of the shunt always fixes it...

 

Hi All -

Sorry for the delay, but we are back! Last night we flushed the fuel tank two more times with gas to get the remaining sediment out. We are in a pretty clean-state now. Calling it good.

The new fuel filter came in and there is a noticeable difference between the new and old when you do a blow test through them, so that's promising.

We got the new fuel filter mounted and buttoned up using the old crush washers. And the fuel tank drain plug has been put back in.

What we would like to do next is the aforementioned "shunt test" to make sure we can hear the fuel pump kick on. Once that has been confirmed we will add fuel to the tank to see if we get gas to the engine bay. If successful, we will try a start-up.

Re : Shunt test

We found these two electronic hook-ups just behind the air filter. The black unit, with rubberized coating, was a two-pin plug so we fashioned a jumper for that unit, installed here.



Before we do the test, I just wanted to double check if this is correct :

1. Place jumper in Fuel Injector Test Plug.
2. Turn the keyed ignition only to the on position (lights on the dash lit up).

With this, we should hear the fuel pump hum, is that correct? Also, I assume the fuel pump is the unit just beyond the fuel filter that is covered with a bash plate?

Let us know if we are on the right path.
Thank you in advance.

Getting closer.

 

Yep, that's right. With the shunt in place where you have it, turning the key to Run position (*NOT Start), the fuel pump will run continuously. I wouldn't let it sit and run for more than a minute or two or you risk flooding it if you have a leaky injector.

The Fuel Pump will take about 15-20 seconds to get Fuel Pressure up to the Rail and back to the tank. You'll be able to hear fuel rushing through the lines. Let us know how it goes,

 

Hi All -

Thanks for the encouragement.

Last night we had mixed results but picked up a few more clues. First, we put 2–3 gallons of fuel into the car, used the jumper in the fuel injector test plug, turned the key, and waited patiently with the fuel main and return lines disconnected from the hard lines in the engine bay — but no gas. Next, we tried to draw fuel from the lines (suck test). After some effort, we were able to get fuel from the main line. When we blew into the return line, we could hear bubbles forming in the tank. We then turned our attention to the fuel pump. We dropped the pump but left everything attached.




With a multimeter on the fuel pump leads and the shunt in place, we were seeing what we assumed was the correct voltage to actuate the pump.



However, the pump neither hums nor vibrates — nothing at all — which leads me to think it’s dead. The fuel pump on my Vanagon is actually pretty loud, and you can feel it working. Shouldn't this one behave similarly? As a future test, we’re planning to remove the pump, attach some pigtail leads, and try running it directly off the battery. In the past, I’ve used this as a definitive fuel pump test on my van, but I wanted to check with the group before proceeding.

We know we can get fuel to the bay by pulling it through the lines manually (suck test), so the lines are clear, but flow just isn’t happening on its own.

Any tips appreciated.

 

Well, that's not an OEM Fuel Pump for one, secondly its missing the Fuel Pulsation Dampner that comes on a factory Fuel Pump which would match the Dampner on the Fuel Rail. These work together to be sure there's no harmonic hammer effect at the Injectors, but that wouldn't keep the pump from running. If you're getting 12v (+/-) at the Fuel Pump connections and the pump isn't running - the Fuel Pump is bad. Simple as that.

Here's a Factory OEM Mazda Pump from 1984;

Note FPD on the left side banjo bolt, the harness to connect to the factory plug under the DS storage bin inside the car, and the rubber bung used to seal the wiring under the car.

Search the Related Threads below for a replacement pump which a member bought recently and he demonstrated to work. The OEM Mazda Fuel Pumps are NLA.

 

Thanks for all of this info. Super helpful. Tonight we will try to wrangle the old pump out and do a quick direct-battery test to make sure the pump is bad.

Re : pump replacement, I'm hoping to take this aftermarket unit to NAPA and just get something similar. Obviously a new unit will also be missing the FPD. Am I setting myself up for failure going this route?

Related, do we need to drain the 3 gallons of petrol from the tank before we pull the pump?

Again, thanks for all the help!

 

I seem to recall that Walbro has an EFI pump that a member here hooked up and was working correctly. The critical dimensions are the flow rate and pressure range, and then the inlet and outlet pipe sizes to fit the existing hose configuration. You'd have to search a bit, but that thread was less than 6mo ago.

When you go to remove the pump, do this OUTSIDE and nowhere near any source of sparks or flame (*obviously). Unless you have a lift, you'll be under the car and unable to do much if there's a problem. The fuel will stay in the tank, but whatever is in the lines will run out when you disconnect the pump. I like to leave the gas cap in place to create some tank vacuum, but it doesn't keep the gas in the lines from running out. You don't need to drain the tank, though.

Test the pump away from the car, just to be safe. Let us know what you find. My bet is the pump is bad and the motor is burned out.

 

Hi All –

Good progress last night. We sourced an identical aftermarket pump and were able to replace the old unit without issue. With the system shunted, we ran tests:

1. The pump ran with the key in the on position.
   
2. With 5 gallons of fuel, the pump delivered fuel to the engine bay.
   

Hooray!

We buttoned everything up and gave her a go, and with great smoke, she started — hooray again!

All in all, great progress. However, we’re still struggling to get a consistent idle — the car just won’t stay running. It turns over well, fires up, and with feathering of the throttle we can keep it running at high speed, but even after doing this for 8–10 seconds, it won’t settle into a good low idle; it just dies. I have an example video : Trying to start.

Any thoughts on moving forward would be appreciated.

Also, for a quick sanity check, these fuel lines, input & return, are on the correct hardlines, right?



Thanks, all. We greatly appreciate all that has been offered here.

Best regards,
MJA

 

Yes, your fuel lines are hooked up correctly. Search on my username and "SE Idle"; plenty of info out there when you're ready to tackle it. Good luck,