Starting up a project that got postponed due to a shoulder injury a while back, rotator cuff, but better now and trying to get my TBI project completed.

I pulled the TBI system out of a 1989 Chevy 1500 with a 350 and installed in my 1976 Jeep CJ7 on an AMC 304. Using a 127747 ECM with a dynamic EFI flash board so I can just learn the VE tables and flash without having to get prom burners and such. Just seemed like a better approach for a newbe.

Jeep start starts up and runs good, but at idle it stumbles at times and sometimes just die's.

At idle the BLM is 128, so is INT, I notice the O2 sensor readings are all over the place jumping up and down from 50 to 900 MV, sometimes the O2 sensor will lock up at 900 mv, and than just drop down to 50 and start bouncing around again. I am using the WUD(whats up display) in the EBL flash software to monitor the O2 and other sensors.

The fuel pressure is set at 13 LBs and holds steady at throttle. I had to replace my fuel return line as it was only 1/4 inch and caused me back pressure and unstable fuel pressure, now that is fixed and the engine still stumbles at idle.

So any thoughts for this confused Jeep owner? whats up with that O2 sensor readings and no error codes on the ECM or could this possibly be normal?

The O2 sensor is mounted in one of the header collector reducer flanges to the exhaust system, it is a heated O2 sensor as it is some distance from the engine. The O2 sensor is a Borg Warner AFS-74

I am running in frame headers into 2 1/2 inch dual exhaust.

I have a 350 throttle body mounted on an adaptor plate on top of an edelbrock performer manifold. Everything else is stock AMC 304, but I am running an MSD billet distributor with vacuum and mechanical advance locked out, the magnetic MSD pickup is driving the EST module. I am controlling spark and fuel with the ECM


Jim

 

Fully warmed up?

NB02 sensor should be cycling 100-800 or so. Normal. Lock up at 900? dont think that is normal. Maybe it is falling out of CL but then again it is heated and should flip nicely CL at about 165degF coolant. When in CL you should see BLM changes and more so INT changes.

I assume you are using stock EBL.bin with necessary changes. What changes did you make?

With key on no start the CE light should blink on - off - then stay on. That is test the .bin is OK. Do you have a CE light? you need one for sure!

Set idle RPM high to get going. I would set in .bin at say 650 rpms in drive(190F)with 5 steps IAC but then use the throttle stop screw to set manually to say 850 rpms just to keep engine running. Is the IAC functional? do visual check on IAC.

Are injectors spraying nice conical?

 

Yes fully warmed up, I have a 180 degree thremostat in my 304, as I am down here in South florida that seems to work for me, is that to low for a GM TBI setup? Jeep never really goes over 185 degrees. Large radiator, seperate transmission cooler.

I do see BLM and INT changes while driving around. BLM typically is around 132-135.

I have an AMC 304 with a TH400 automatic, so I started with a bin for a chevy 305 with a TH400, Bob at Dynamic EFI sent it to me. I also have the ECM downshifting the TH400, I used to have a switch on the gas pedal, now the ECM controls the kick down solenoid with a slave relay.

I made the recommende changes per the Dynamic DFI instructions like TBI, 8 cylinder, calculated the BPC and installed new BPC, disabled EGR as my manifold is non-EGR, enable charcoal canister, I as using an early 90's Caddy solenoid operated charcoal canister. All the recommended stuff. Nothing more.

Can't check the CE light, not at the Jeep at the present time.

IAC seems functional, did the intital IAC setup. Have a rebuilt Throttle Body. As I have a 180 degree thermostat, 190 most likely will not work for me. Did not set up 5 step IAC, didn't make any changes to that setting, I guess I will have to go look what it is set to after I find where to find that.

The injectors spray a nice conical except at idle, at idle the spray diameter really reduces to almost a drip.


Jim

 

Soulds like all is OK cept:

you need a CE light. I suspect one of the "outputs" from ECU(pins) need to be wired to simple bulb/lamp. Howell engineering sells one I believe. they specialize in Jeep BTW. I believe mine is number 20 but I have a diff harness same EBL ECU. Maybe it is 20. that female coupler has 1-24 pins and other is 1-22. circuit 419? But again I am 84 Vette.

IAC steps stock are most likely higher than 5. 5 is recommended at idle.

Drips no good. Bump FP to 16 and check drips at idle. Requires you reset BPC. Are injs rebuilt? FIC Injectors will do that and ultrasonic CLEAN them. If unk I would do it. they also offer flow match and determine flow # if not sure.

 

OK, I do have a CE light, I guess I call it a SES light. Yes it is an LED and all is OK with this LED. I have it mounted in the dash. I added two LED's, one for the SES light and one for the fuel pump, I wanted to know if the fuel pump had power at all times. I am going to add a kill switch and roll over switch to the fuel pump as soon as I get this idle issue worked out. I sure do not want gas flowing on me after a roll over in the CJ.

I had the fuel pressure up to 15 lbs, and no change. I got this throttle body as a rebuilt, I looked at the injectors, but they do not have any markings on them to identify what they are. Maybe I need to send them to FIC to check fuel flow and clean. I guessed at the injector flow when calculating BPC. But the engine runs flawless until idle, really nice performance other than at idle.

Do you think this poor injector performance at idle is causing the engine stumble?

Jim

 

Possibly. If dripping that is not good. Should be conical spray. Check the SES. Key on, should blink on - go off. then on solid. If not then your .bin is not right.

 

Sounds like there isn't enough injector offset compensation. Check the injector PW at this time. Once it starts to go below .8 to .9 msec the injectors don't always open. It may also be that the ECM is switching into async mode injection. Which when at idle doesn't work out.

If the PW is getting small (.8 -.9 msec) then increase the Injector Correction Offset table values by 15 - 20%. Leave the value at 0 V at 0. Note that doing this will also cause low PW areas (light load) to richen up some. So the VE table may need some additional tweaking.

But the main idea is to get the injectors under control at low PW's so that idle works out.

If it is going into async mode during idle there are the Async Transition Lo and Hi terms to change that. Along with a Min Async PW that can be used to change the characteristics.

RBob.

 

Rbob:

OK, thanks for the suggestions, I will give that some attention this week-end. Thanks to all for the advice, this idle issue has been beyond me, and I surely needed some expert advice to get me on the right track.

There have been a couple of CJ TBI conversions I have noticed on the Jeep forums that didn't work out just because they could not get the idle worked out. They went back to carburetion. So I thought I would try here where you guys are so experienced with TBI operation. Might just be an AMC engine thing, if I can get this worked out maybe I can help some other CJ owners that have struggled with TBI conversion idle issues on some of the CJ/Jeep forums.

Jim

 

Before you make calibration changes, warm up the engine and take a 2 to 5 minute data log of the troublesome idle. Set the log aside.

Then after making whatever changes to the calibration to improve the idle make another 3 - 5 minute data log. If the idle is still troublesome please send me the logs and I'll take a look at them. Actually, I wouldn't mind looking at the logs either way.

Sent the update out, let me know if you don't get it.

RBob.

 

OK, so I was looking at the current bin in tunerpro and checking everything once again, while checking BPC which calculates to 118 for my application, BPC under constants in TunerPro says BPC-Base Pulse Filter and is set as a pecent from 0-100.

I do not see any other constant BPC or BPW, so what am I doing wrong here. I am sure I set it to 118 several months ago with no problem, but now the text edit box comes up as a percent, not a value

Never mind, found it BPC vs VAC table. Locked myself out thinking it was a constant. I will log some data today.

Jim

 

Here are two log files, both with the engine warmed up and idling, the first file "304 first run no changes to inj offset table" which is the engine idling CCP vacuum removed from TB and capped, transmission in park.

The second file "304 first run no changes to inj offset table 1" is as above but with the transmission in drive, idling. The engine after a brief few stumbles stalls out. You can see it loose CL and drop out.

When just idling it does stuble but remains running, when in gear it tends to stall out, bit once on the road runs just fine.

I never saw the async mode while idling, and the PW stayed above 1msec so I did not make any changes to the injector offset table yet. Thought I would let you take a look and maybe make a recommendation on how I should proceed.

Thanks for the help

Jim

 

I believe the BPC table for non VAFPR users should have same values in for all VAC areas.

I recall I idle around 1.4 msec PW around 14.00/1 on WB currently. B4 I made changes in tune to help idle quality I think I was seeing changing PW (.80 to 1.4) cycling surging up and down with idle set at 900 rpms neutral

I would give the TSS aturn to get the RPMS upto give you an opportunity to tune the idle. With engine stalling/dying it makes it had to get a handle on things. Check the IAC steps you are seeing in datalog during idle. I cat see the log here at office.

Interesting to see what RBob sees.

 

Looked at the log files. It isn't a short PW problem that is causing the stalling.

Try setting open loop idle mode (it is an option flag) and see if that clears up the stalling. The INT and BLM is dropping during idling, maybe the O2 sensor is a bad. I don't think that this is the issue, but only takes a couple of minutes to try it.

If anything the idle should smooth out as some of the idle being erratic is from closed loop prop gains. Use the WUD Analysis graphing and you can see where the idle surges a little each time the O2 swings lean.

Another thing that may help steady the idle is to reduce the idle SA compensation. There are two tables, idle high and idle low. Try cutting the values in the tables in half.


I am running an MSD billet distributor with vacuum and mechanical advance locked out, the magnetic MSD pickup is driving the EST module.

This may be the cause of the stalling issue. Is the EST module a GM unit?

After setting the base timing (BYPASS connector open) does the timing then advance once the BYPASS is closed and the engine restarted (timing mark moves toward passenger side)? If it retards then the mag pickup leads are reversed going to the ignition module.

With the BYPASS open is the timing at the damper steady? Does it look like the timing light flash sometimes drop out?

I wonder if the mag pickup signal to the ignition module is dropping out at slow engine speeds. How well does the engine crank to start?

The ECM needs the EST pulses in order to fire the injectors. If pulses drop out then injector firing also drops out.

RBob.

 

I am running an MSD billet distributor with vacuum and mechanical advance locked out, the magnetic MSD pickup is driving the EST module.

This may be the cause of the stalling issue. Is the EST module a GM unit?

Yes, same part as original in the 1989 Chevy 1500, already swapped out the EST with a new one. I originally had the 1989 Chevy 350 1500 TBI distributor machined to fit the timing cover on the 304 and shortened to fit my AMC gear, had the same problem with that original distributor.

After setting the base timing (BYPASS connector open) does the timing then advance once the BYPASS is closed and the engine restarted (timing mark moves toward passenger side)? If it retards then the mag pickup leads are reversed going to the ignition module.

Yes, timing advances around 20 degrees. The 304 would not run when set at 0 degrees base timing, I had to set the base timing to 10 degrees to get the engine running and not being so retarded. I guess difference between the Chevy 305 and the AMC 304 base timing. Tried reversing the mag pick up leads, ran like crap.

With the BYPASS open is the timing at the damper steady? Does it look like the timing light flash sometimes drop out?

No, timing is steady

I wonder if the mag pickup signal to the ignition module is dropping out at slow engine speeds. How well does the engine crank to start?

Cranks right up really fast, never did that with the carburetor. No issue when running above idle, really runs good with great acceleration.


Jim

 

I looked thorugh some data logs for a stock 305 SBC with 350 injectors. Idle in gear at 550 RPM steady as a rock. Same injector PW's as your in gear log (1.7 - 1.8 msec).

Back to the injectors where you state that they almost drip at idle.

at idle the spray diameter really reduces to almost a drip.

They shouldn't do this. There should still be a decent cone spray from the injectors. I recall looking at them on the stock 305 and thinking that they sprayed a lot of fuel. The eye integrating the pulses into a constant stream makes it look this way. Some dripping from the center of the injector is normal. The film builds up and drips off.

You can use a timing light to stop action the injector spray. Will need to move the pickup to another plug to see the other injector (alternating firing scheme).

The unmarked injectors do throw a flag. GM injectors are always marked.

I would get a set of GM injectors from a 305 or a 350 engine and try them. The 305 injectors are 55 #/hr and the 350 injectors will be either 61#hr, 65#/hr or 68#/hr (vehicle & year dependent). I wouldn't worry about getting any particular ones. As long as they are stock GM injectors.

Even JY injectors which can usually be had inexpensively will work. They rarely go bad or clog. Change the BPC to suit and install them. See how they spray at idle along with idle quality.

RBob.

 

If you need injectors, we can help you out. We have "rebuilds" and new andy size you need!
SouthBay Fuel Injectors

 

OK, I have some old throttle bodies in the shop, I will pull the injectors and see what I have and send them out for a re-build or core for some new ones. They are not ready for use, but will cut the cost somewhat.

Thanks for the help, I have been at this for a while and hope to get it worked out soon. Just the idle, everything else is just fine.

Jim

 

OK, I think I might have made some progress. I set the loop idle mode to open as Rbob suggested and it cleared up the stumble at idle, purrs like a kitten at idle. So I guess this means the ignition and injectors are OK? Maybe O2 sensor?? Any thoughts? The O2 readings while in closed loop in the WUD jump up and down between around 220 and 990 mv constantly, in the open idle mode the O2 sensor reads a steady 980 mv.

Jim

 

I would expect idle quality would have improved OL as no sensor feedback. A/F will change based upon coolant temps.

02 voltage changes I believe are normal. The swings can be tightened up on swing points and the swing over a different median voltage effectively changing the CL A/f.

 

I would expect idle quality would have improved OL as no sensor feedback. A/F will change based upon coolant temps.

Ronnie, help me understand this, I would expect the engine to run better with the ECU making real time changes to the A/F, rather than the engine running on the fuel tables alone. If th eengine runs good in the OL mode, than the fuel tables must be close to being right on I would think.

02 voltage changes I believe are normal. The swings can be tightened up on swing points and the swing over a different median voltage effectively changing the CL

What makes the O2 reading swing like this, if in CL I would think the O2 readings would be constant as the ECU makes changes to keep the A/F close to 14/7. Help this novice understand how this works.

Jim

 

OK, I did some reading and now I understand the bouncing O2 readings, it confirms the CL mode and is the ECU constantly richening and leaning out the A/F around the magic .45 volts (14.7 A/F). So this is normal and how the system works as O2 sensors are not that precise.

I still do not undertand why my engine would idle better in OL than in CL, I guess I need some help here.

Jim

 

Found this description of open loop mode suggesting some engines run better in open loop idle becasue they need a richer mixture than most.

Open loop mode differs from closed loop in that the O2 sensor is ignored and the engine can be managed to run at A/F ratios other than 14.7, usually richer or lower than 14.7. The ECM controls the fuel injectors without getting any feedback that the calculated fuel delivery rate actually matches what the engine received. Without feedback, the loop is open, hence the term "open loop". A good example of open loop is when the engine is first started on a cold day. It requires a rich mixture to start a cold engine since a lot of the fuel doesn’t reach the combustion chamber. This is because a portion of the gasoline doesn’t vaporize and pools inside the manifold until engine heat vaporizes the fuel. Another reason that engines run in open loop when cold is that O2 sensors don’t work until they reach about 6008 F, so it takes a few minutes in cold weather for them to begin functioning. Open loop is sometimes used at idle conditions since some engines idle better with a rich mixture. In open loop, the ECM commands an A/F ratio that is determined from a table of A/F vs. engine coolant temperature. The open loop A/F is also adjusted to run richer as MAP increases.

When looking at the analysis mode of the logs I posted, I notice the graph for the O2 sensor shows readings with significant drop outs at times, this O2 graph even though in closed loop mode looks ragged, is this a normal looking O2 graph?

So I am thinking I either have a defective O2 sensor or my particular engine AMC 304 needs a rich mixture at idle requiring the use of open loop idle mode.

Jim

 

CL idle I believe may be done for emmishions reasons. Add a larger cam and larger injectors, OL idle is improved over CL. Other factors contribute to idle issues. asynch vs synch idle and proportional gains also need to be addressed. Search that.

My ALDL logs of idle show the swings as you state. My WB logs can be smoothened for mA/F and are. My pillar mount A/F is set to show frames maybe .25 sec apart? shows a more stable A/F at idle. My prior goofy A/F gauge was like a blinking xmas tree! back forth.

Point of interest. I idle OL and tables show for coolant at 170-200 d F 750-950 rpms "all flattenned" to around 13.5/1. Yet I idle on WB at 14.7/1? Exhaust is very clean. No vac leaks! Maybe the WB is being tricked? Seeing fresh air from other end of ext pipe(reversion) or same at intake(reversion)? with a 114 LSA dont think that should happen. Is my BPC wrong at that VAC? Maybe?

 

Ronnie:

Take a look at the picture I have attached showing the analysis graph using the 1/4 mile performance tool that comes with the DynamicEFI EBL Flash Board. It is not the O2 voltage swings I am concerned about, it is the drop out between swings. Anybody using the EBL Flash system take a look and tell me if you see this type of O2 sensor drop out while idling in closed loop mode out of gear. Maybe Rbob can take a look and say OK or not. I just have not seen a normal engine running with this tool, so I am trying to determine if this is a defective O2 sensor or not.

Jim

 

It drops to .20 Volts? From .90 volts. I believe that is functioning NB02 sensor. I never looked at mine that closely. I will look at my logs this weekend. I do know for fact my NB used to fall out of CL(headers) especially when cruising with no load. a heated NB02 fixed that. My swings I tightened up about 10% and moved or enrichened about 10%. .bin shows stoich at 14.3 I recall.

 

jf, what you are seeing for the O2 swings is normal for closed loop operation. What I didn't like about it was that the engine RPM surged each time the O2 swung lean. Also, try reducing the idle compensation SA tables. It may be that the two together are causing the rough or surging idle.

Or at least between changing the two (OL/CL and idle SA comp) you can get the idle to steady down.

I previously mentioned that the lean surge is probably from proportional gains. When it may be caused by the injectors. The strange part is how the engine just stalls. A tuning issue usually has increasing surging with the engine eventually stalling on a down swing.

To better understand the fueling and why the O2 swings see the Tuning Guide Book sticky on the DIY_PROM board. In chap. 5, section '8063/'8746/'7747 there are a couple of papers on various areas of the ECM.

RBob.

 

With the BYPASS open is the timing at the damper steady? Does it look like the timing light flash sometimes drop out?


Rbob Ihave a 350 TBI and it seem like my flash drops out,what does that mean?

 

No, my timing is steady with no drop outs.

I put the idle in OL and ran several long VE learns to get the fuel tables tuned in down to around 700 rpms, that seemed to help somewhat on CL idle. Than I put it in CL idle and ran several VE learn's now that it was not dieing on me everytime I tried to get going from idle, that cleared up the surging, but still have an occasional drop out. Still some work to do but getting better. I think I will do a compression test this week just to insure nothing has changed internally. Vacuum test showed normal engine operation.

Jim

 

Finally got around to doing a compression test, not perfect but cetainly nothing that should cause any serious issues. Here are the results:

1 140
3 137
5 131
7 140
2 141
4 130
6 129
8 140

So on to more reading and changing some additional parameters.

Jim

 

OK, so the idle problem is solved. I pulled the distributor(MSD Billet) and found one of the gears on the reluctor ring was just touching the magnetic pick-up core. After resolving that problem the 304 idles perfectly in CL or OL. Thanks to all. The combination of the distributor issue and the untuned engine certainly confused me a bit, but in the end it all worked out.

The MSD Billet Distributor was new and added when I converted over to TBI as the Mallory Unilite I had was not compatible with the EST, not a magnetic type pick-up.

Jim

 

Thank you for letting us know what solved the issue. The good idle is important.

RBob.