broken93

Some may have seen my thread a while back:

https://www.rx7club.com/3rd-generation-specific-1993-2002-16/running-one-rotor-possibly-372884/

in which I was having problems getting it started.

I still haven't been successful, and I'm out of ideas. At the conclusion of that thread I was fairly sure I was having a flooding problem, and I have since pulled my primary rail and replaced all the o-rings, etc.

I am still having the same problems - I put everything back together tonight; I cranked it and it fired up and ran for a bit - very poorly, about 10mmHg vacuum, stumbling, and when I let the clutch out it died again. I tried for another 15 minutes to get it started again, with no luck.

I pulled the plugs and cranked it with the EGI fuse out and as usual I was rewarded with a liberal mist of (probably) fuel vapor. A good bit of vapor comes out of the exhaust while cranking.

I'm out of ideas, and pretty smug about this whole thing now. I've pretty much lost hope. Anyone have any ideas?

The only thing I can think of so far is that perhaps one (or both) of the secondary injectors are stuck open? Perhaps I have lost a coolant seal and the vapor is vaporized coolant (?) - there's no coolant smell, however.

All the obvious things have been gone over many times:
- no vac leaks
- no timing problems (RPM reads correctly, 250 cranking)
- PFC's set correctly (850/850)
- MAP sensor OK
- plugs OK
- coils OK
- plug wires OK
- etc, etc, etc

thanks, anyone...

J.S.J

I don't remember if you had the intake off but if you might have switched connectors around involving the coolant sensor and the fuel temp sensor, injector connectors maybe? You need to look at the fuel pressure in the system. If its a leaking injector the pressure will bleed off and you can isolate the leak by pinching off the lines leading to the tank as to isolate the engine side. I spent 2 weeks on a problem every night 4 or five hours and I ran out of things as well. Turns out I left a rag in the lower intake runner and it could not be seen. The engine was a fresh rebuild and hadn't ever run right. I found my problem with just keeping my cool, most of the time. I have 20 years of drive ability experience and tested everything 100 times. Good luck.

broken93

What's strange is when I had the primary rail + injectors out I tested the fuel system by pressurizing it - held 50 psi no problem, no leaks, and the pressure held steady after the fuel pump shut down for at least 2-3 minutes.

Now, after reassembling everything, I pressurize the system and it goes to 40 psi and rapidly falls of (10-15 seconds) to almost zero - that's fuel leaking into my intake. So, I suppose it's got to be the secondary injectors - I'll take that rail apart tomorrow and see what I can see.

Thanks for your help. I'll update if I come up with anything.

J.S.J

I took a look at the history of your problem and one thing stood out to me. Being a drive-ability specialist I often see direct failure's meaning a shorted or open sensor and they are no trouble because the computer sees the problem and directs you to the fault. The hardest are intermittent in the harness or in the sensor itself. I would guess its likely when the fans would come on for you unexpectedly followed by the engine springing back to normal is likely the system seeing a hard fault and substituting a back up parameter. My ideas are this would likely be a core input signal. Like the coolant, map, and throttle sensor / circuit issue. The air and fuel temps are more trim in my view. If it were me I would try just turning on the ignition but don't crank the engine. From there assuming the fans are not coming on now for no reason I would carefully manipulate the engine harness and input sensors with the idea in mind that if you find a problem even with just the ignition on the fans would trigger on when or if you duplicate the problem. In my experience a skewed coolant sensor wont turn on the check engine light unless its open or shorted. If its real skewed but not enough to trip a code it would be hard to pick out. The harness connectors must be looked at. If the female terminals are spread it could cause this type of fault. Just a thought. John St. John

mad_7tist

pull the vac hose off the fpr and see if there is fuel in it. but def look into the secondaries for leaks

broken93

Your suggestions are very much appreciated, as is your time taken to read over the history of my problem.

The wiring harness is brand new... all of my sensors are reading normally on the PFC, coolant temp = air temp = ambient temp (prior to starting), TPS voltages are normal, and (yesterday) when I pressurized the fuel system prior to putting it back together, as I said, it held pressure just fine.

Mad_7tist, I have an adjustable regulator with a gauge on it. Idle pressure is around 43 psi, as expected, so I am not having fuel pressure issues. Somehow I am getting a LOT of fuel into the engine, and it's flooding... I have verified that the primaries hold pressure, and return to closed when the pintle is actuated (I clicked each one manually and it sealed on close)... the only thing left is the secondaries, or the secondary rail/o-rings/etc. They shouldn't be firing at all during cranking/idle, so perhaps the vibration is opening them or causing the o-rings to fail? I don't know...

broken93

Checked out the fuel system. No leaks. Fuel system is fine. Checked again for spark, spark ok. Checked fuel injector and ignition wiring again... ok.

While I had the UIM off I put some lube in the primary runners to help un-flood.

Noticed that if I cranked with the tb closed and the egi fuse out I saw about 8 mmHg vacuum; with the EGI fuse in, I saw almost no vacuum; disconnected IACV, solved that problem - fired for a few seconds, sputtered, and died.

It is consistently flooding (so much that I am seeing liquid gasoline coming out of the downpipes). I am fresh out of ideas. Flooding is probably just from cranking it so much.

broken93

Anyone?

J.S.J

I will say since you installed your old ECU its not likely a skewed PFC. I have seen Prom's fail within a General Motors ECM where the input sensors are reading fine. It's the computers calculations tables that are off.
I assume the engine has run fine at one time with the new wiring harness?? If not I would still keep that in mind. I would disconnect the secondary injectors electrically but it sounds like you have likely done this. I would like to see a reading of the injectors pulse with compared with a good running engine. Meaning you could connect a volt meter to a primary injector the voltage will just be a reference to compare to a good engine. If there are no leaks the pulse must be on far too rich. I would also assume the exhaust system is clear of restriction. It would have to be a bad restriction to cause this conduction.

broken93

I have swapped ECU's, yes, and haven't had much luck with either one.

The engine has never run correctly, with the exception of the one time that you are aware of when the fans kicked on and it ran smooth as butter. I did try grounding the thermoswitch to manually turn the fans on and try and start the car; didn't seem to make a difference. I have checked grounding with my DMM on several occasions and verified that all major grounding points are within 1-2 ohms of one another.

The exhaust is clear of restrictions (just downpipes, no muffler, etc). I also have nothing attached to the intake (incidentally, does the engine rely on vacuum in the intake tract for proper start/idle? My T/B is open directly to atmosphere for the time being).

I can repeat this procedure indefinitely:

- crank engine, engine floods, condensed/liquid fuel in exhaust, etc
- remove EGI fuse & plugs, unflood engine (fuel mist from plug holes)
- clean & replace plugs, crank, fires and stumbles for 2-3 seconds, dies
- floods again
- lather, rinse, repeat

The best luck I had today was after unflooding the engine, I replaced the plugs, disconnected the primary injectors, and cranked it at WOT (in an effort to ignite and remove remaining fuel residue in the engine). Re-connected primaries, and it fired up for about 4 seconds, promptly died, and flooded again.

Another hypothesis: I have four 850cc injectors, two of which I bought from a forum member. I don't recall which set I put in my primary rail and which set I put in my secondary rail (one of the two is the original set from the car).

Perhaps someone had their injectors bored to 1300cc, sold the car, and the chap I bought them from didn't know it, in which case I have 1300cc primary injectors

Although this is a known good motor I am going to do a compression test again at some point. Perhaps in the process of removing the engine and cranking it all this time I may have damaged something...

JSJ again thank you for your help on this one...

Mr rx-7 tt

iTrader: (5)

What is the compression on the motor?

broken93

I am going to try and get some solid compression motors as soon as I can find a ladder to get my compression tester out of the attic

Are there any tests for the CKP sensor (the outer one?)... or the ignitor (i.e. resistance/etc)?

I am thinking about "borrowing" an agilent from the lab and bringing it to my house - hmm what would happen if I got caught with a 25k oscilloscope

J.S.J

If you think the primary injectors could be 1300's that would explain the rich conduction but not the reason the engine seemed to run great when the fans came on. I guess there is one way it could happen. If the coolant sensor became shorted that would translate to about 300 degree F, and that would be leaner than a cold engine but its not likely. If the engine really ran good when the fans came on I would think there is still something a skew within the guts of the computer controls, sensors or wiring. As I talked about earlier, the computer will turn on the fans if a major input or output is skewed internally or externally with in the system. I have seen loads of cars with seaming good input and output sensor readings that ether wont run or run correctly. I can't speak for the PFC and how it reacts, however it may be similar. You know, you are now learning more about your car than you might want to but it may come in handy later on.

broken93

Well, I did rule out the obvious last night - I went ahead and got two large 4-gauge 24" ground straps (battery cables, but they have lugs on both ends), and replaced the UIM->Firewall ground as well as added a ground between the A/C compressor bracket mount and the chassis. There wasn't any change.

My course of action now is to:

- check the compression
- replace the e-shaft position sensor (the outer one) - I know the inner one works because I don't get an RPM reading without it
- replace the ignitor module
- replace the coil pack

In other news, WTB: working CKP, ignitor, and coil pack...

wait, wait... let's make that WTB: running RX-7

J.S.J

The crank sensors are interchangeable, so just swap them. Sure if you get a RPM reading its operational. I would not mess with the coils. There not going to over-fuel an engine. I don't see how the ignition module is an issue ether. Those units would cause misfires, radical scattered ignition but not over-fueling not to the degree you have. I assume the ignition is working after the engine stops running if so there is a reason you getting too much fuel. The injectors could be too large or the system is expanding the injector pulse-with. I would try to isolate which of the two you "could" have. I assume the primary injectors were fitted/installed using the plastic device I think its called an aspirator. I have herd and read flooding problems happening if that is left out. You could switch your injectors around, primary to secondary if there is any thought of the sizing. If the ignition system is skewed disconnecting electrically the injectors/fuel pump and running the engine using a can of ether or just a small amount of gas placed in the intake would give you about the same conduction, but if the ignition system is fine and the system is reasonably un-flooed placing some fuel in the intake would fire the engine and you may be able to modulate a small amount of fuel in the intake to evaluate the situation. You have racked you brain and bank account enough. I used to call this raw testing of subsystems. Isolate a suspected fault and test it in its raw form. If the engine runs with some gas, wd-40 etc your ignition system is fine. Gut feeling from 20 years of fuel injection work telling me this.

broken93

I get an RPM reading from the inner sensor (which has the ticks on the crank). I don't know if the outer sensor is working, which indicates a full revolution and allows the ECU to sync to TDC (only one tick). I still get an RPM reading with the outer sensor unplugged, so I will swap them; and if I get an RPM reading with the outer sensor in the inner position, then I know that they are both working.

Actually, no; I did some very careful machine work on my primary rail to properly fit the 850cc injectors so that they seal properly. I have verified (on multiple occasions) that they do indeed seal properly - I've pressurized the entire primary and secondary rails detached from the engine and experienced no leaks up to 60 psi static pressure. The only path for fuel into the engine is through the injectors.

According to the PFC, the PFC provides initial-cranking fuel enrichment on the order of about 15-20% duty cycle for about 10 revolutions and then backs off to 3.4% for continued cranking (I assume this is the compensated fueling point on the injector duty map for 850cc injectors and 250 RPMs). I would hope that what the PFC tells me it's doing with the injectors is what it's really doing.
As much as I hate to disassemble the fuel system for the 12th time, I think I will try this. If that is indeed the problem, I think I'd be pleased - I'd have a running engine and a free set of 1300 cc secondaries (hey, I could put 550's back in my spare unmodified primary rail - if I only had some 550's)...

I may give this a try, with someone helping out. I've got to be careful, though... I tried something like this once before on my other FD and set the damn wiring harness on fire

What I have been doing lately is disconnecting the primary injector harness (the way my FPR and under-the-UIM stuff is installed I can actually reach my hand under the UIM and disconnect the connectors). I've got to do a painful un-flooding procedure, involving pulling the plugs, disconnecting all four injectors, spinning the motor to clear the fuel through the plug holes, lubing the rotor housings, reinstalling the plugs, and cranking the motor with the EGI fuse reinstalled but the fuel injectors unplugged. Ceteris parabis, the engine will run for a bit on what fuel is left as it burns and clears it out, and I should have a nicely unflooded motor at which point to start troubleshooting - first, reconnecting the primariies and moving forward from there.

I keep forgetting, however, due to the large number of variables... I have to remind myself... this engine did run perfectly in this configuration for about 35 seconds, when hot (180 degF at the thermostat, 102 degC on the PFC) with the fans on, as you know.

I am off to troubleshoot some more... remind me to fedex you a pizza and a beer as a token of my appreciation of your help.

broken93

Update: no luck.

I have checked the ignition system; I have spark from both leading and trailing coils. I have checked the compression; I have 90 psi cold, throttle body closed; on all three faces of both rotors, so I would estimate 100+ psi warm, t/b open. Fuel system has been checked repeatedly, torn down and put back together again. I have no leaks. I re-adjusted the TPS to within spec, no change.

At one point the thing was massively flooded because I had a leaking secondary o-ring; I had the UIM off; I went to crank it to clear the flood, and forgot to unplug the EGI fuse (the injectors were disconnected, however), and the thing fired right up on the fuel that was left in the motor; unfortunately, there was no throttle body to prevent air from entering the engine, so it went right to about 5k before I shut it down (it was about 2 seconds of runtime - one big rev). At the time the TPS and Fuel Thermosensor were disconnected; but I don't consider those to be related. Other than that I have had no success.

So here's the question: I have compression, fuel, and spark. Why in the hell will the thing not start?

stevecod

before you start going crazy, find a way to fit your 550cc back in your primary and see if it works.

broken93

Well, tonight I put the 550cc injectors back in, using a stock rail so there are no leaks. I double-checked the fuel pressure. I am still not getting it started. Crank crank crank, flood, flood flood. Of course, I thoroughly unflooded the car before attempting to start it.

- I have good compression, as I mentioned before, around 100 psi on all faces on both front and rear.
- I have ignition. I have checked both the primary and secondary coils and the leads arc to ground with a bright blue spark and a click click click.
- I have the same symptoms on two different ECU's (stock and PFC).
- I have checked the CAS and CKP resistance, in spec.
- TPS has been carefully readjusted to spec.
- I can't test the ignitor without the test harness, but I have ignition...
- plugs are clean.
- wiring harness is good (brand new)

I have no idea what to do next, short of strip the car and start selling parts. Unfortunately, I'm almost serious. I've been working on this starting issue for more than two months now with no luck. Does anyone have any suggestions at all, or is anyone in the area with rotary experience (or perhaps spare FD parts) willing to stop by and scratch heads?

tiger18

iTrader: (1)

hi mate ive got you something silly to check,,,,but trust me.,,,,might sound dum,,,i bought my jap mazda,,,,same style problems as what you got,,,you know what it turned out to be,,,,
the crank angle sensors were plugged wrong way round (how stupid,,,car fired straight up after that) had the G and NE,,,plugged vica versa,,
some reason both plugs are the same.....just a silly thing,,,,but fixed mine

J.S.J

I am sorry to here all you have been through thus far. Since over-fueling is what stops you before you get started I might remove the coolant sensor connector and short the two circuits together. This might as it will in most systems give the ECU a real hot reading and surely would lean out the injector pulse. I guess it would also set a code and turn on the cooling fans as well. This isn't going to fix the engine but maybe just getting it running a bit could shed light on whats really going on. Ask you self (likely for the 100,000th time) what makes a system over-fuel. You have see fuel pressure and other basic readings/tests which look good but what else would cause the system to over-fuel or inject fuel at the wrong time. I would think a skewed or inaccurate rpm signal to the ECU could cause this, something grounding the injector circuits during cranking or related. I would want to see a noid light on the injector harness during key on engine not running and during cranking. You surely would not see a solid light. I also might connect the map sensor to a vacuum pump alone at a set 18inches or so during cranking and not connected to the engine vacuum. Something is throwing off the system and causing the over-fueling. I had a problem where I removed the upper intake and put some foam rubber in the runners. I ended up putting two peaces in the front side runner and only removing one before reinstalling the intake. Well, I had a one rotor engine for a while. I removed the intake 2 times and didn't see the extra foam rubber because it had been drawn down out of sight by that time. I finely got a glimpse and removed it. After that it was back running great. Lets assume the fuel is good which is in the tank as well. I had a jack *** tech put 1/2 gallon of diesel in full tank of a Park Avenue I was working on and there was just enough to keep it from running but you could not smell the diesel fuel or even feel its different consistency. That was a pain in the blank. John St John

broken93

JSJ, thanks for your help as usual.

Something I did discover which may or may not have been a source of the problem... in '93 there was a recall for the cooling fans which involved adding a jumper harness to the main engine wiring harness near the ECU. This spliced in the additional fan control box (black box bolted to the back of the ECU labelled "control unit" and also added the famous "fan mod" unused ground connector. My car had the original recall performed; and hence had this harness; it spliced into the main harness (via plugs) at two points, one on a blue plug, and one on a white plug.

Well, I replaced the wiring harness with a brand new one from Mazda. Mazda apparently (some time prior to producing the '94 model because I have heard '94 model harnesses are like this) integrated the fan control connector into the harness and did away with the recall splice harness. Well, I still had the recall sub-harness connected as well as installed the new harness, so I may have had some wiring problems. Remember how I said that everything ran great once the fans kicked on? Awfully strange coincidence...

Anyway, I have since corrected that, and gotten the splice harness removed and the wiring as it would be on a post-recall-redesign car. I haven't had any more luck, or noticed any major changes. I still have to unflood the car some more (already did it again once).

Does anyone know more about what was on that harness and what it would affect (it's hard to trace on the wiring diagrams since the sub-harness colors don't match the main harness colors)?

Well anyway, I'm done for today. Maybe tomorrow I'll get lucky.

dubulup

does sound like a weird coincidence.

x-05 and x-14 (white and blue) are in my car (1995) and I had the black box unit?!?!?!

how big are your primary ports? stock, small, mild, large, aggressive streetport??

where in AL.

J.S.J

I think I would check the harness layout as it relates to the sensor connectors. I think you said you have a commander and the readings look good but there are a lot of circuits not in the data stream you can't see. Maybe the ignition system. Sure you have spark but are the leading coils firing first etc. Yes another daunting task for sure. I would still look at the Throttle sensor as well as the map wiring. You want to know if each circuit has the correct placement in the connectors. John St John

tcb100

I agree with Tiger 18 - you have symptoms consistent with crank angle sensor problem.