spectre6000

iTrader: (1)

I've been having a slight hesitation since I did a bunch of non-engine related work. That work involved the ECU, CPU, dash, etc. so there were plugs and electrical bits that share a circuit with the engine, but nothing more. I changed the plugs and wires today to rule them out, and because it simply and clearly hadn't been done in a good long time.

So here are the symptoms:

• Hesitation in high RPM situations.
• Only after the secondaries come in.
• If I stomp it, it's very lurchy and cuts out quite a bit.
• If I roll in slow and smooth and let it climb gradually, there are no issues at all until around 6K and everything is super smooth to that point. (<= this is the kicker)

Assessment:

First reaction was the 5/6 ports. Any issues with them are seemingly negated by everything coming in nice and smooth if I roll through the transition gradually. Since that rules out air, we're left with fuel and spark. New plugs and wires take care of some of that potential, but there's obviously still coils and whatever that thing is called that replaced the distributor; neither have been tested yet. The injectors are new as of maybe 300-400 miles ago along with the fuel filter and pump strainer, so they're pretty safe. The engine ground under the manifold was cleaned, DE greased, and made happy when the injectors were done.

Electrical connection between the fuel injectors and whatever tells them to fire (ECU?) seems suspect, and I guess I should test the coils for whatever specs they have (I seem to recall there being a test spec in the FSM related to the ECU plugs).

Anything else I should check out when I get around to fiddling again next?

j9fd3s

iTrader: (3)

fuel filter? i assume its been changed but didn't see it.

um next step might be to make sure the secondary injectors have a good connection, electrically, if the connector is in bad shape it'll do exactly what you're describing

spectre6000

iTrader: (1)

When we did the injectors, clokker brought his little jar of DE grease, and it was used on the injector plugs. What do they connect to on the other end? What is in play that I'm not accounting for? I still haven't fully wrapped my head around all the extra/unnecessary electro-doodads that Mazda thought were a Good Idea™ in the 80s.

EDIT: I just looked back at the time stamps of your post and my edit, and it's likely we were busy simultaneously and the email just took the scenic route through the ether to my inbox. Either way, accounted for.

j9fd3s

iTrader: (3)

yours goes like this;

battery -> main fuse -> main relay -> egi fuse (or maybe BTN?) -> resistor block -> injectors -> ecu -> ground.

the resistor block lets Mazda use low impedance injectors with smaller current injector drivers, later cars just have high impedance injectors. i have seen the resistor block break, its right under the airbox, looks like a ballast resistor for a points ignition.

the other trouble source is the injector connectors, but if you're 100% sure they are seated right, probably not it.

spectre6000

iTrader: (1)

obviousDS strut tower??also DS strut tower? what's a BTN?under air boxobviousobviouscase to body.

j9fd3s

iTrader: (3)

the main fuse, EGI fuse and BTN fuse are all in the little fuse box on the drivers strut tower. main relay is kind of in the corner by the brake and clutch master and the trailing coils

spectre6000

iTrader: (1)

I back probed all the wires at the ECU, and came up with a few things worth further investigation. One, my AC blower motor seems to have a short. I imagine this is from when we got that slushy snow a few months back and I had water running into the cab... Bummer. AC doesn't work yet anyway, so I might just unplug it for now.

The more pertinent anomaly is the relief solenoid valve. It has something to do with air intake it RPM above 3500, and it's reading battery voltage regardless of... anything. Sounds like a short. Does this sound culpable for the symptoms I'm experiencing?

It looks like all of those solenoids are identical, and I seem to recall reading somewhere that they're interchangeable. They also seem NLA... I'm about to go pull it off the car. Maybe I'll swap it with the EGR solenoid (at least the plugs) to see if the reading follows. Are they reparable at all?

spectre6000

iTrader: (1)

Here's a thought I just had! When I replaced the injectors, I discovered that the EGR had been deleted. NBD, it still passed emissions with flying colors. There are four of those little solenoids in a row that I think are identical. The first is the relief solenoid valve, the second controls the EGR. Since it's missing, I move them around, and now it's testing out. Pretty sure that a lack of an EGR more or less precludes the need for a solenoid to control it, so barring anyone having any suggestions to the contrary, I think I'm back to full operation!

I'm about to go mess around a bit with the AC blower (probably just unplug it since the AC compressor won't engage) and see if I can take care of that short (and likely parasitic draw), then take it for a test drive.

spectre6000

iTrader: (1)

Well... I systematically unplugged every plug related to the AC blower, then the AC system, and the short never went away... Ideas?

Actually, I misread that... The reading for plug 1E is <2.5V with the blower on and the AC on and ~12V with the blower on and AC off. The AC doesn't turn on all the way, so maybe that's just telling me the problem is in that direction. The switch turns the LED on, but the compressor clutch does not engage. Mayhap that's what I'm seeing. Disregard that.

spectre6000

iTrader: (1)

Swing and a miss... Swapping them may have made it test out at the ECU plug, but in practice, there must be some difference between them. It went from lurching when in the 5/6 ports to just lurching no matter what. Going to try to figure out what the other little solenoid guys are for and see if they can be swapped around at all...

j9fd3s

iTrader: (3)

the ECU is on the ground side.

if you're looking at the wiring diagram, power is at the top, and ground is at the bottom.

i'm not looking at the diagram, but i do know that the ECU is the ground for the AC system. this is so the ECU sees the ac being on, and can then compensate the idle. the blower fan also needs to be on. i'm a little hazy but there is also a low (or high?) pressure switch, so if there is no freon in the system it won't turn on.

spectre6000

iTrader: (1)

What it would be compensating for is the extra load from the compressor. Since it doesn't engage, there's something that won't drop that voltage there. Low priority at the moment.

What I need to figure out is the vacuum solenoid issue...

spectre6000

iTrader: (1)

The swapping theory didn't pan out. The part numbers on them are identical, but it went from stumbling above 3500 RPM to stumbling... Everywhere. I didn't try to get it above 3500 RPM, just making a logical assumption. That likely means I have two bad solenoids. The EGR removal wasn't exactly done with care (it was just left open), so I imagine the EGR one is pretty gross inside, and that's what's causing my issue there.

Any other ideas? They don't seem serviceable (internal rivets hold the stupid things together...). I've put out a smoke signal on the local rotary FB page, but it looks like I'm either going used off eBay, or stupidly expensive new... Open to suggestions.

j9fd3s

iTrader: (3)

those are the same, the ECU is the ground.

the relief solenoid is one of two that controls which way the air pump air goes in the air control valve (ACV). the other is the switching. the switching, switches between port air, and split air. the relief chooses either venting the air pump air to the air cleaner (actually the relief silencer, atmosphere), or to the switching valve .

with Mazda's two light TPS setting method, you are literally looking at the outputs of these two solenoids, at idle one solenoid is on, and the other is off. not sure which.

spectre6000

iTrader: (1)

Am I correct that they have something to do with controlling air flow when it gets into the 5/6 ports? The issue I'm having is intermittent stumbling at high RPM. If I go into it slowly and smoothly, it's generally fine until around 6500, if I go in full throttle, it stumbles starting around 3500 RPM. I can obviously keep my foot out of it, but mechanical condition is priority number one. Compounded issues can mask each other and I might miss that something more devious is afoot if I don't have this sorted.

j9fd3s

iTrader: (3)

nope. the 5th and 6th ports open from exhaust backpressure, so there is a little hose to the split air pipe, to the cat.

stumbling with load sounds like a fuel delivery issue

spectre6000

iTrader: (1)

I can now say with ABSOLUTE CERTAINTY it was the boost sensor... It must have started dying slowly a while ago, because holy wow it's all back! Why Mazda thought to call a device on a naturally aspirated car a "boost sensor" is beyond me... But that was it all the way.

spectre6000

iTrader: (1)

According to Mazda:

"Boost sensor" "Maintains constant -96mmHg (3.78 inHg) command" AKA Vacuum sensor...

Way to be descriptive with the part naming. Personally, I would have gone with "headlight caliper" to really throw people off, but I wasn't in Mazda's engineering (translation?) department in the early 80s...

spectre6000

iTrader: (1)

After yesterday's revelatory change in performance with the new boost sensor, I went to run some errands today, and it's back to where the boost sensor must be dead. It was an S5 boost sensor, but it sure worked great there for a bit! Hopefully it's something simple and I don't have to go chase electrical gremlins for months... Overly and unnecessarily complicated electrical systems is why I hate newer cars...

spectre6000

iTrader: (1)

I back probed the plug on the boost sensor, and there were two brown wires with black stripes that had a very nice clean ground connection, a brown and red wire (looks like it goes to the ECU) that had no connection to ground, and then the other two wires had some resistance between them and ground, but ultimately read conductivity. It was kind of weird the way they did though... Set to the beeper setting (beeps if there's a connection), touching the probe to either of the other two wires would cause it to beep very quickly, then go silent. This is all car off, key off. Searching manuals to try to find some more informed way of testing this crap.

If I read the wiring schematic correctly, the brown wire with the red stripe goes directly to the ECU. Ergo, back probing the boost sensor at that wire at the plug should give me an identical reading to what I would get at the ECU. I set up the multimeter, turned the key to the ON position... And got a perfect reading. Spec is 3.5-4V, I got 3.55. Not sure which end of spec that is, but at 100 mmHg it should read 2.5-3.5V. I don't recall what the old boost sensor read at the ECU, but it was either very close to ~12V or 0V.

When I messed with it on Saturday, I replaced the bum relief solenoid with a new one (never did test it with the multimeter, but it made no difference compared to the old one). Since there was no change, the guy I was getting the parts from said he'd seen a bad boost sensor cause similar symptoms. He had one in his stash, we plugged it in and it was revelatory. Night and day. I drove the car home grinning ear to ear.

The next morning I moved the car around the parking lot, and it was just as happy. The next time I started it though, it was back to driving like crap. I'm still having CPU issues, but looking at the schematics they don't seem obviously related... Really stumped.

j9fd3s

iTrader: (3)

section 4A, or B i think.

the black/white wire is switched 12v, for only the S4.
brown with black is the sensor ground.
brown with white is the 5v reference voltage
and then brown with red is the signal.

if the sensor is electrically plugged in, and you blow and suck on the vacuum nipple. the signal voltage should change

spectre6000

iTrader: (1)

If by 4A or B you mean manual sections, then I've been all over it. Also the wiring diagrams.

I tested the boost sensor as suggested, and it all checks out. It was out of spec at the ECU Saturday morning. Saturday afternoon, I put a new boost sensor in and it was back to awesome. Sunday morning crap again, now the boost sensor is checking out perfectly.

On the plus/minus side, there's been some change that might be indicative of something progressing... It went from being down on power and intermittently losing power at high RPM, to being awesome everywhere, to back the way it was before the new boost sensor, and now it's worse than it's ever been. I took it for a drive just now and it's down on power, occasionally losing a little more power at low RPM and fully hitting a wall at the 5/6th UNLESS I ease into it. If I'm gentle on the throttle, I can still get it to rev smoothly all the way up to 6K or so.

So in the air/fuel/spark equation, I'm at a loss.
• Could be not getting fuel... The progressive nature of it could indicate a fuel pump that is steadily failing or maybe since the fuel regulator is electronic somehow for some reason that probably made sense once upon a time, maybe that's failing.
• Could be spark. I haven't looked at timing in a while, but with my air cooled background I'm usually overly meticulous when it comes to timing (even with my diesel truck). Plugs and wires are new, coils test out. The only unknown is how the electrical stuff controls advance and whatnot.
•Finally, it could be air. Changing the boost sensor MIGHT have temporarily fixed something, but it could just be that some other unrelated thing that's actually the culprit was accidentally jostled into behaving. Looking at the wiring schematic, it's pretty tightly coupled with a number of other sensors. Since the symptoms are sensitive to throttle position and only occur under low vacuum, and the boost sensor may or may not have had any effect, it could be some electrical thing related to how they measure load between the TPS and the various air flow/vacuum sensors...

I'm stumped and it's stressing me out bad.

EDIT: Also, and this could just be me being paranoid, but maybe it's worth something, it might have smelled a bit hot after a longish highway drive on my way to get the parts Saturday. The guy I was with thought he might have smelled it as well, but it wasn't really all that strong and we couldn't really be sure it was actually coming from my car...

j9fd3s

iTrader: (3)

fuel regulator is analog. there is a vacuum line to the intake, so fuel pressure across the injector remains constant. there is a solenoid in there that is used on hot starts only, to interrupt the vacuum signal, and thus raise the fuel pressure.

i'd bet that spark is fine, i know there are a lot of "check your timing" threads, but the thing doesn't move unless someone moved it.

the TPS, boost/pressure/map/whatever you wanna call it, and a couple other sensors all share a ground, and there is a common 5v power too.
i assume that the ground on top of the engine was tightened/checked when you had the intake off? its reachable with the intake on, although you need a swivel, 12mm

it is odd that its intermittent.

j9fd3s

iTrader: (3)

what happens if you put the old pressure sensor back in?

spectre6000

iTrader: (1)

Are you suggesting then that it's not likely related to the fuel pressure regulator?

Agreed.

The ground on top of the block was removed, sanded, greased, and replaced. I'm bad about being forgetful, and there were other people involved, so if anything I would have double checked my double checking on that one. I might even have partially reassembled the thing, then disassembled it again just to be absolutely sure I didn't forget to tighten it. It tests out good with the multimeter.

Additionally, testing the 5V reference signal at the boost sensor, it was 5.00V on the nose.

You're telling me...