supercell

Hello everyone-

I have been trying to track down the source of this problem for a while now, but have made very little headway. So far, I am just making mental notes of all of the perceived issues, and plan to tackle the troublshooting in the next week or so (as weather permits). Here is the primary condition:

Car idles around 1200-1300 RPM with no load. With any electrical or A/C, P/S load, the idle is a little lower, at around 1000-1100 RPM. When idling at this high level, the car sounds like it is running very rich - it pops and vibrates more than normal, but not like it is going to die. Here is where it gets strange. If I tap the gas pedal (sometimes more than once is needed) the idle will drop right to 750 and the car will run smooth as can be, just like nothing is wrong at all. It will not get down to this level by itself, though - I always have to tap the gas to get it to do this.

The car also surges while driving on small residential streets, which usually means I have to keep the car in say 3rd gear for a 35mph zone. Parking lots are no fun at all and usually require a lot of clutch work. If I don't the car will start bucking badly.

The two may or may not be related, I am not sure, so I am including both for clarity. Also, in my underhood voyages, I have checked every vaccuum line in sight, and replaced the T/B gasket and elbow O-ring when I replaced the plugs and wires a few weeks ago. A look at the fuel filter indicates that it has been replaced somewhat recently - the paint on the filter looks new and only has a little dust on it while the hoses it is attached to are pretty filthy. The throttle stop screw looks like it has been tinkered with - there is wear on the edges of the jam nut, right where a socket would grab it. MAP sensor looks good - hoses are ziptied in place on it.

So here is my line of questioning- is it possible that someone improperly adjusted the throttle stop and caused this, or does the fact that the idle returns to 750 when I tap it negate this? I was going to start testing with the TPS to see what values it is giving the ECU. Does this seem logical? Could the TPS being adjusted improperly cause the problem instead?

I can't seem to find much on adjusting the throttle stop - if I adjust it, should the throttle plate be completely closed against the bore, or is there a predifined gap that needs to be set with a feeler gauge like older models? I don't want to do this unless I can determine it has really been messed up. I also checked the fast idle cam to make sure it was getting out of the way when fully warmed up, and it does.

I guess that's about it for now....anybody have any ideas along these lines, or maybe a instant cure? Thanks for listening, rotary brethren!

Jon
'93 MB Touring
'85 Silver GSL-SE

KZ1

yeah, sounds like it could be dashpot, the adjustment i n the manual i s a little fuzzy. this is something I woudl like to know as well. But, I don't think it is your problem, I have had mine completely disconnected, and didn;t notice anythign that bad. How is your throttle cable adjustment, also maybe check operation of cold idle cam.

Eric Michel

Since your also having problems at part throttle also I would think the problem is more likely the throttle position sensor or maybe your map sensor is malfunctioning although I have never heard of one going bad. The throttle position sensor actualy has two pots built into it , one for low range (idle& part throttle) and a wide range for like over 30%. The sensor is very difficult to adjust, the shop manual only talks about using a special service tool, you might have to have it checked/adjusted at a shop/dealer. Has the check engine light ever come on? Did you check to see if there are any error codes?

Hope this helps, good luck,

Eric

lone wolf

The bucking might be solve by grounding the engine. I use to get bucking at low speed until I added a ground to my engine.

supercell

I have had no check engine light, and there are no codes on the ECU now either. My thinking is along the lines of someone tampered with the throttle stop adjustment, which in turn knocked the TPS on-idle condition out of whack, and now the engine thinks that part throttle is being applied all the time, and richening the mixture/advancing timing to compensate. Perhaps this is why it will settle down when the gas is tapped - maybe it is already really close to being at the closed throttle (idle) position, and can get there by 'jiggling the handle' so to speak. Throttle cable slack looks good, fast idle cam looks good - disengagement is happening just as it should. I also should mention that once 'tapped down' the idle will adjust itself properly for engine load, so I think that the ISC is doing its job as well. (assuming it stays closed until load is applied like every other FI 7 I've owned - ISC/BAC- is there a difference in operation?) I plan on checking the TPS narrow and full band values this week sometime. Hopefully that will be a good place to start....then maybe back to the drawing board if this is not conclusive. Anybody have any other possibilities? the fact that it can (and will) return to normal idle suggests to me that this isn't a LIM leak or vaccuum leak.

supercell

Update:

I checked the TPS narrow and full range values through the whole sweep and they are good. I also yanked out the elbow while the car was still hot and checked the primary throttle plate to bore clearance, and it will pass a .10-.15 feeler gauge wiper at fully closed position, but nothing more. Now I'm back at square one again, I suppose....has anyone seen an ISC stick partially open? I know the BAC/ACV on my GSL-SE was gunked up really bad when I first got it.... I had to completely clean the whole thing out to get it to work properly again. Anybody have *any* leads/ideas I can chase down on this? I have never seen anything like this before, and I thought the SE's idle problems were interesting when I first got it. Thanks for any and all help!

Eric Michel

I have been runnig a Haltech for the past 2-3 years but a long time ago I had a very similar problem, Im trying to remember how I fixed it. Here's something that might help, the fuel pump has two speed settings at idle and low load the voltage/current is feed through a resistor, above a certain load point a relay closes sending full system voltage to the pump. I had that resistor burn up on me, it caused all kinds of problems with the idle and low load operation, once you got above the load point and the other relay kicked in it ran fine. If I remember correctly the resistor is mounted on the fire wall under the cruise control unit (if you have one). I think the shop manual has a voltage level check you can do for the fuel pump also. If thats not it you probably do have the throttle plates misadjusted, from what I remember that can also cause strange idle problems. Now I remember I played with those adjustments trying to correct for the fuel pump problem after that was fixed I had to go back and readjust them (actually RP diagnosed the fuel pump problem with me, thanks Ari). If some one has already messed with your throttle plate positions you probably cannot make it any worse. You might want to get advice from a pro but here is my 2 cents on how to adjust them. The long vertical screw is the stop for the two upper plates, normally dont mess with it, if to late set so they are completely closed ( the screw just touches the stop arm). There is a third idle screw valve under the throttle body, (with the elbow off you can see it) try to adjust your idle with that valve completely closed (fully clockwise). the last adjustment is the horizontal short allen head screw that controls the lower single throttle plate, adjust for a smooth idle. You may need to drive around and readjust it 2-3 times before you get the proper setting, the ISC may temporarily mask an improper setting.
These adjustments are based on my own experience, I have no idea if this is right but it worked for me.
As some one has already suggested an additional ground wire from the upper intake manifold to the chasis is also a very good idea.

Hope this helps,

Eric

supercell

Thanks for the advice, Eric. Fuel system might be the next best place to try. If I take the vacuum tank off (the top one) right after parking and stick my nose as close as I can to the injectors, I get a slight smell of fuel. Is this normal? I looked all around the injectors and pulsation dampener as best as I could, and there is no wetness there at all. I wonder if maybe leaky injectors could be causing this? Can the fuel system cause the car to idle high? I thought that only extra air entering the system system was the only cause of this. The car is running rich by the look of the plugs, and today I noticed that the idle was hunting. If I pressed the gas and held it at any RPM, the engine would drop a couple hudred RPM and come back over and over until the pedal is released - totally repeatable while the hunting was there. Does this indicate that the fuel pressure is inconsistent, or is this another (unrelated) problem?

Eric Michel

I don't think I have ever smelled the intake right after shutting down the engine but I think it is possible to have a little unburned fuel left in the intake, I wouldn't worry about the smell. When the engine hunts/surges as you described I have found that indicates a very lean condition. Cars that run rich (unless extremely rich) drive and idle smoothly, stink, snap, pop, shoot flames, foul there plugs in short order and load the engine oil with gas (smell your dip stick. From what you are describing I am even more suspicous of your fuel pump, it sounds like you are running lean at idle and light loads, the majority of carbon build up on your plugs occurs under high loads, with the situation I have described you could easily be running lean at idle and part throttle and then be running rich under full load. While fuel mixture can affect idle speed, not to the extent that you have, I am thinking some one played with your throttle plate settings trying to correct for this other issue. Does the car run good under boost? what mods do you have? How many miles? do both spark plugs show similar carbon build up? if so its probably not an injector although having them flow tested is a good idea in any case.

suganuma

my car has been doing the same thing.....or was
the idle hunting and the low speed bucking, etc....sounds like we may have the same problem - i cant figure it out either.

i keep the car around tx for the winter - next time im home i may setup an appt to get it checked out @RP - maybe chris has seen the problem before or can diagnose it better than me.

the fear i have after what has been said now is that i thought that tthe searching idle and all was a rich condition - i know that i was running rich. i have the power FC and tried to mess around with idle settings with no success.

i talked to chris aobut it one time briefly and he just said "its probably an adjustment somewhere" ....nothing i didnt already know, but he was in a hurry. he seemed inquisitive on what might be the problem though. he told me he didnt think it was anything bad and shouldnt hurt the car so i continued to drive it the little that i do.

i did some work where i had to disconnect the battery and then roll the car and inadvertently reset the ECU when i pressed the brake pedal. i started the car and my idle was magically back at a solid 750 for the whole day.....but only that day.

after that things got worse. now the car doesnt even want to hold an idle. where it used to surge from 900-1300 or so it surges much lower - down to 500 and then sometimes dying. last time i was home i pretty much had to keep gas on it to keep it alive. i wasnt able to really "get on it" and clean out any carbon deposits due to some suspension problems that i need to fix as well the next time im home.

the fact that all of this time at idle i may have been running lean...maybe very lean and then rich (i know im running very rich under full throttle b/c my injector duty measurements are all out of whack - theyre very high for close to stock boost levels. 10 psi yields injector duty in the 90% range)
anyways.....this all makes me fear that i may have blown my engine or cause some sort of permanent damage.
however, not knowing clearly the variety of different characteristics or what the engine does when blown im not for sure. maybe someone could shed some light on this or even if it would be possible to blow the engine from this situation.

i will shortly be converting to haltech and having RP do some tuning and hopefully fix this problem. i hope its nothing major.

sorry for the short novel and thanks in advance for any help yall can give.

supercell

Ok, Here's yet another update:

Tonight on a hunch, I tested voltage at terminals A and B on the MAP with a voltmeter. I didnt want to mess with warming up the car and idling, so I just turned on the key and tested and the voltage measured was 4.95v between the two. According to the workshop manual, this reading is off the charts, somewhere over 29 in/Hg of boost. Granted, the book also says to do this with the engine running at idle, so I am reserving judgement at the moment. Eric or anybody else- do you know if this value is high, or is there something else in the ECU perhaps that prevents this voltage reading from being accurate unless the engine is running for sure? Anybody willing to measure their known good one under similar conditions?

Eric - the car boosts up to ~10psi on primary turbo, somewhere around 5 at transition, and second turbo is unknown, presumably not boosting at all - but I don't have the roads around here to try this out very well either. I'd like to get this other problem solved first before I push it real hard. So to answer your question, yes the car runs very well when I am accelerating, but downshifting is almost completely pointless with the high idle, and mid throttle driving without extreme care will nearly snap my neck due to surging/bucking. I think your call on the lean condition is right on the money. The top of my downpipe glows cherry red if the car sits at idle for very long. As far as mods go - it is nearly stock....I have an aluminum AST, M2 downpipe, and magnaflow catback. The spark plugs were evenly coated between rotors, and the leading ones were significantly more worn than the trailing. I replaced these along with the wires a month or so ago. My exhaust tips still collect carbon, but not nearly as much as they did before. After changing the O2 sensor, replacing the downpipe, and doing plugs/wires, it did significantly reduce the build-up from before. From what I was told from the previous owner (car has had two before me) the engine is original at 84k miles.

supercell

Hey Nic -

Please let us all know what the diagnosis is at RP - I would be very interested to hear what they have to say about it after laying hands on the symptoms. I live in rural AL, so I have no hope of taking my car to a shop for diagnosis or work unless I want to risk driving it a minimum of 3 hours to Mazcare in GA. (or one expensive flatbed job!) All I can do is relay the symptoms to anybody who will listen and hope I give accurate descriptions!

Did you by any chance check your TPS and MAP values? I am not familiar with the Power FC yet so I don't know if it tells you these values. Take a look at the above post for my MAP readings - god I hope the problem is this simple!

Eric Michel

Other than the severe surging I do not think that running lean at low loads will damage your motor, it's the high load conditions you need to worry about. I dont have a stock map sensor anymore so I can't help you with that. Did you check the fuel pump circuit yet? I really think that may be your main problem. As a test find a way to engage the high load relay all the time, the car may run a little rich but if your surging problem goes away you found your problem. I am going out of town for the rest of the week, so I wont be able to respond for the next 2-3 days, good luck hunting.

supercell

Eric -

I tested the MAP again with the engine at idle, and realized that the book's image does not imply one way or another which connector is being diagrammed, so I tried the other terminal and got the normal voltages. I tried again with the engine turned off, and the voltages are correct, so the MAP checks good. In thinking about what you have said, I wonder, if the idle has been raised to 1300, but the TPS and MAP report the engine at idle, wouldn't the extra air lean the thing out for sure? I think I may be understanding this a little better now. I suspect maybe my compression may be low, and the previous owner couldn't keep the car running at (normal) idle with the A/C on (car came came from FL), so he raised it compensate. My boost gauge never indicates vacuum below 16-17 in/Hg - ever. Idle vacuum is usually at around 14-15 in/Hg. I just wish I could explain the accelerator tapping and idle returning to normal thing. I think I'm going to have to (GASP!) take it to the local Mazda dealer for a compression test - assuming they even have the right tester anymore, or know how to use it! I tried testing it myself before, but I now think maybe I took out the wrong check valve, and thus don't trust the readings.

Sorry for the novel - If you think the car is running lean for reasons other than the above, how do I run the fuel pump in HP mode all the time? (and does this have any bad side effects?) Let me know when you get back - I'll give you an update then. Thanks for your help!


Jon

Eric Michel

Jon,

The good news is if your compression was low tapping the throttle would not correct the problem, so its probably not that. You can find info on the fuel pump circuit on page F-100 & 101 of the shop manual. You can bypass the resistor temporarily by pulling out the fuel pump relay and shorting the contacts. The relay is in the box/relay strip mounted by the front bumper, shorting the two connections closest to the bumper will bypass the fuel pump resistor (see attached pics). The pump draws several amps so make sure you have a good connection (at least 14 gauge short wire with some good flat blade crimps maybe).
According to the manual this two stage pump voltage helps maintain proper fuel presure and improves the reliability of the pump. Bypassing the resistor will slightly increase your fuel presure at idle and part throttle, causing the car to run a little rich at these points. You do not want to permanently leave this but for test purposes it will be fine. After you do this your idle may still need to be adjusted but you should get a signifcant improvement in your part throttle driving.

supercell

Eric-

Ok, I see the relay you are talking about. I have pulled it and fabricated the jumper using 10 gauge wire with crimped on blades as you suggested. I will try this out tomorrow. Were there supposed to be attached pics on your last post or were you referring to the shop manual? Not to jump the gun here, but if you are right and the pressure is too low, what do you think could be the problem - Faulty filter, pressure regulator, clogged primary injectors, bad pump?

Also, on the other stuff - a couple of nights ago, when I tested the MAP with the car idling (untapped; at 1200-1300 rpm) I turned on the A/C briefly and it killed the engine. That is the first time that has ever happened, but I don't use the A/C very much, so its not like I have a good baseline. I also did not bother restarting it and trying it again with the idle at the proper level (tapped). All I can say at this point is, when it is idling high, it runs like a cammed V8 (this must be the lean condition) , but as soon as I 'tap' the pedal and the idle comes down, the car runs smooth as silk. Weird stuff. Do you think maybe the lean condition would have caused it to die like that with the A/C on? Also, do you think that this fuel pressure problem is in any way related to the high idle, or do you think I need to do a full adjustment of the throttle stop and AAS screw to get that part out of the equation? If it helps any at all, the throttle stop screw is screwed in to the point that the bottom of the screwdriver slot is even with the stop nut. Thanks again for all your help!


Jon

djantlive

Do you have a good ground on the battery? Bad ground on the battery can cause bucking. Try grounding a thick wire.

supercell

Ok, I have just tried the jumper to keep the fuel pump in high output mode, but this did not make any difference at all other than make the exhaust smell more like fuel. The downpipe was still glowing orange at idle as well. I also tried a second temporary ground wire from the battery to the chassis, and this did not make any difference either. (tested this separately) I did find that if I place load on the engine like the A/C the idle would return to normal and stay there even after I turn it off, but as soon as I use the gas again, idle returns to its normal high. BTW, the A/C did not kill the engine today.

This thing has got me baffled.....I'm wondering if maybe I have a very minor vacuum leak somewhere. It looks like I still have the original vaccuum hoses under there - almost all of them are rock hard. Anybody ever see an actual improvement in idle vacuum by doing the hose replacement? (assuming you didn't find any hoses completely disconnected)

Jon

an_juan

check the idle control solenoid....it might be stuck a bit open.

legendr35

My car was bucking like hell in low RPM ranges and I had to give it lots of gas in low speeds like you said cuz it will stall if i don't, but all this was fixed when i bought a new universal o2 sensor by bosch. its only 20 bux, maybe you should give it a try.

supercell

Thanks guys - I am looking into the ISC thing now. I replaced the O2 sensor when I put on the downpipe last month, so I am pretty sure it is still good. I have a feeling this may be in the ISC system somewhere, so I am troubleshooting as many sensors that tie into this system as possible right now to eliminate the electrical possibilities before I have to pull the manifold to check/clean the valve's mechanical parts. If ihave to do this, I think I will also pull the AWS to clean it too....could just be carbon/oil deposits builds up on these after many miles and gunks them up good - my '85 SE's BAC valve (similar in operation to the ISC) was absolutely filthy when I first got the car. There was thick gunk all over it, preventing it from opening and closing properly - and this was a non-turbocharged car! Please keep the ideas coming everyone!




Jon

Eric Michel

Jon,

Did you drive the car with the fuel pump in high load mode? Still bucking? An easy way to check your ISC just unplug the electrical connector for it and use the idle air screw on the bottom of the throttle body to set the idle. You will not have load compensation but if the ISC is your problem the idle should become stable accept for load variations. The accelerated warm up solenoid could also inject air into your system, it gets air from that hose connected to the throttle body elbow tube. To rule that out simply stick a bolt in the hose and cap the elbow hose connection. The third air possability is the Secondary Air Injection System, that is not easy to check and you will definately have to remove the upper intake (see pages F119-F122). Does the air pump clutch engage at idle?
When I have my idle set to lean it does stall when the AC comes on. I tried to attache some pics for you but they were to big and sounds like you have a maual or you are using the online one so I just referred to the pages. Did you put an extra ground from the Upper intake to the chassis?

Good luck,

Eric

supercell

Eric -

Thought you dropped off on me! Well, I tried the fuel pump thing, and it didnt really change things while driving either. I have been studying the operation of the ISC, and have noticed that the ISC kicks in a little every time the engine comes down from 4000+ rpm to emulate an anti-afterburn valve. Assuming this thing is gunked up, it seems very likely that the problem I'm having would occur when the car is decelrated to a stop, then I have to tap the gas to make it close enough to idle properly. When I tap the gas, I have found that it will not work if I rev up to ~4k rpm, but the idle will settle down usually when it is tapped up to 2000-2500 rpm and then dropped. I'm going to take it out driving today to warm it up and then check the idle with ten and gnd jumped to eliminate the E/L unit, A/C and P/S idle compensation systems from the picture. When I took off the throttle body to do my plug wires, I cleaned alot of carbon build-up and general 'gunk' out of the throttle plates. Seeing as how the ISC and AWS frequently allow the intake air a way around the throttle plates, I am thinking those must be extremely filthy by now as well!

I have a factory workshop manual, which is what I have been using to try and understand the idle system. What's funny about it is that I am finding there is some interpretation necessary to understand it on some points. It truly is the bible of third generation cars! I tried the extra ground temporarily, but I was not satisfied with the way I had to attach it, so I removed it after the test.

I'll let you know the latest when I get back and have tested the idle again.


Jon

supercell

Sorry, forgot to comment on the air pump thing - yes, the air pump clutch is engaged at idle.

Cutaz35

Sounds to me like your secondary throttle plate is not closed properly, check to see if you need to clean the throttle plates, and possibly adjust the sesondary plates slightly more closed, that will usually keep it from hunting.Good luck..... Jim