Backfire through intake EFI & Carburetor

[svocapri]

I have a freshly rebuilt 302 in my 85. It has a rebuilt set of Dart Windsor heads with a fresh valve job and springs. The Cam is a Factory HO roller cam. Ever since I put this engine together it has had a problem with backfiring through the intake below 2500 rpm. It does it when I open the throttle as if it doesn't have enough accelerator pump shot.

The car is running a Megasquirt fuel injection with a Holley 4barrel throttle body injector atop a Weiand Stealth intake manifold. The Megasquirt and intake was on the car before the engine rebuild but didn't have the backfiring problem with the previous engine. In troubleshooting the problem, this is what I have done...

• Added acceleration enrichment until the car would stumble from being rich on acceleration
• Loosened the valves until they were clacking thinking I might have had a valve too tight
• Adjusted the distributor phasing thinking the spark was jumping between cylinders
• Installed a Holley 4010 600CFM double pumper which was on the car before the Megasquirt

None of these things have been able to alleviate the problem. When I datalog with the Megasquirt it does show the engine going very lean as the throttle is being opened but then quickly goes rich to between 11-12 AFR. This is true both with the Throttle body injection and carburetor.

Can anyone offer me some troubleshooting advice? I have exhausted all my troubleshooting knowledge.

[Mustang-junky]

Verify that you have the correct firing order and that the initial timing is set right.

Jess

[Haystack]

X2. Timing or wrong firing order will cause a back fire.

[qikgts]

Look at cam timing in addition to ignition timing...

[brianj]

Look at cam timing in addition to ignition timing...

This. Verify TDC on the piston with cam events at the rocker, as a quick way without having to rip the motor apart.

[erratic50]

I agree on cam timing and ignition timing being likely off. Leave the advance on, advance timing until it just speeds up then back a slightly. Ignore your timing light, make sure it starts OK. See how it runs as timing marks can get way off causing bad readings.


I had a cracked distributor cap make my car do that one time. Have also seen bad rotors and worn out distributors cause similar problems.

Also - on a hunch - try replacing your coil to cap wire.

check your advance springs and weights, etc,

[cb84capri]

What are the differences between this engine and your last?

Cale

[Walking-Tall]

Phasing? What's in it for a distributor? Where's the initial timing and how much and when is mechanical advance all in? If it's got vacuum advance, how much does it advance, and what's it connected to? Judging by it mattering not between throttle body injection and carburetor, it sounds like a too-much ignition advance off-idle, at low speed ("... below 2500 rpm.") issue to me, which could cause misfires and/or backfires, that an air:fuel ratio measurement would show as being lean (even a rich misfire will be indicated the same, an important reminder that people ought not forget that wide-band oxygen sensor measurement is still only a tool), because all sorts more oxygen than fuel goes out the exhaust with either of those two scenarios...

[306gt]

Take a compression test. Just to verify the valves are closing. If the compression test proves everything is in good mechanical shape, Then you may have ignition problems. Is the firing order correct? If the ignition timing at idle is too far retarded backfiring will occur.

[11secGT]

Ditto like the others; timing,leaking valvejob,vacuum leak, firing order or wrong cam firing order-vs. order @ plugs/dist.

[cb84capri]

So Julian... have you made any progress on this?

Cale

[svocapri]

Well, I finally have made some progress. I re-wired the distributor pickup to the Megasquirt with a higher grade shielded wire, and I re-wired the ignition module to get the high current wires away from the distributor pickup. I was thinking that maybe there was some electrical noise causing my problem. I also swapped my coil to a MSD blaster. None of these fixed the problem.

I had a friend come over and we put the Holley throttle body back on. When I went to check for fuel leaks we found a leak between the injector pod and the throttle body. Four new o-rings fixed that and then things started to improve.

We drove for about an hour and tuned the fuel map. It was lean between 1500 and 2200 RPM, so that was contributing. I also found out that the recommended 0.2 to 0.3 seconds of acceleration enrichment is way to low for a throttle body injected dual plane manifold. I bumped the acceleration enrichment up to 0.6 seconds and things started to work much better.

As for the carburetor not working, I have no idea why and I don't have another engine to test it on. I am going to say something must be wrong with it, but both accelerator pumps have fresh diaphragms on them and squirt fuel like they should when the throttle blades are opened.

I am still working on dialing in the tune, and for those who know Megasquirt, I am going to try the enhanced acceleration enrichment to see if I can get the throttle transitions smoother.

Julian

[Walking-Tall]

FWIW, all of the advice above by everyone of basic mechanical and ignition things to check still apply. Ignition or cam timing that's off (making the explosions come up out the intake valve/s and the induction system), that's not ideal, can indeed present as an air:fuel mixture problem. Wide band oxygen-sensing ability, in whatever form, is only a tool that's as good as the information it gets, and ignition happening at the wrong time, or rich misfires, WILL read false as lean with wide band oxygen sensing. Clue #1 is that the engine does the same thing with either fuel injection or a carburetor that both seem to indicate that enough or more than enough acceleration enrichment is being provided or you have amply made provision for. Verify true cylinder #1 TDC valve and ignition timing, that the distributor is stabbed in where it should be, that you're getting whatever initial timing you think you are with your timing light or however you're checking it, and you're likely to get past this seemingly brain-busting issue.

[erratic50]

Verify your cap and rotor are in good shape. A cracked cap presents as a major ignition problem

[svocapri]

Mike, I agree with what you are saying, and I am not convinced that everything is fine just yet. I rented a compression tester but unfortunately the universal spark pug adapter it came with wouldn't fit it the spark plug hole of the Dart heads.

I used a piston stop to verify my timing mark when I put this engine together and I degreed the cam during assembly. I set the cam two degrees advanced to help compensate for the large cylinder head chambers and 8.9 to 1 compression. This engine was built from scrounged parts to be a daily driver fun car, so being easy to drive and spin the tires ranked higher than top end horsepower. That being said, the heads were redone with a fresh valve job, new spring and retainers, and new roller rockers. The short block got new bearings, a fresh hone and new rings.

Something I didn't mention earlier is that I had run the carburetor with the secondary linkage disconnected and it didn't have the backfiring problem. I had similar results disconnecting the secondary linkage on the throttle body injection.

Finally something I find strange is the way the VE table is shaping up. It is calling for over 100%VE between 1400 and 2000 RPM. It has another high VE area between 3200 & 3600 rpm.

[svocapri]

306GT, I wanted to follow up on your retarded timing comment. The two pictures are my initial timing table and what I am currently running. The more retarded table was based on a table I pulled from an 87 mustang computer. The engine definitely acted too retarded on that one so I bumped the timing advance across the board to the higher table. I don't have a full set of timing marks on my flywheel just a TDC mark so I am going to add timing tape and verify the advance at higher rpm's.

What is the consensus on the more advanced timing?

[Walking-Tall]

Mike, I agree with what you are saying, and I am not convinced that everything is fine just yet. I rented a compression tester but unfortunately the universal spark pug adapter it came with wouldn't fit it the spark plug hole of the Dart heads.

I used a piston stop to verify my timing mark when I put this engine together and I degreed the cam during assembly. I set the cam two degrees advanced to help compensate for the large cylinder head chambers and 8.9 to 1 compression. This engine was built from scrounged parts to be a daily driver fun car, so being easy to drive and spin the tires ranked higher than top end horsepower. That being said, the heads were redone with a fresh valve job, new spring and retainers, and new roller rockers. The short block got new bearings, a fresh hone and new rings.

Something I didn't mention earlier is that I had run the carburetor with the secondary linkage disconnected and it didn't have the backfiring problem. I had similar results disconnecting the secondary linkage on the throttle body injection.

Finally something I find strange is the way the VE table is shaping up. It is calling for over 100%VE between 1400 and 2000 RPM. It has another high VE area between 3200 & 3600 rpm.

Well, backfiring in the intake tract requires air, fuel, and an ignition source. It's that ignition source, the plugs inside the combustion chambers, having an open path through open intake valves, that I'd suspect to be a condition that is existing, that should not. In my mind, that can only mean either intake valves still open at the wrong time when the spark is happening when it's supposed to, or vice versa. Since I believe you're saying you verified #1 TDC and camshaft position upon assembly, I would investigate ignition spark potentially happening too much advanced, during a part of the intake phase instead of during/after the compression phase.

I want to say that secondaries not being in the picture would mostly decrease the available air in this scenario of backfiring.

I'm not familiar at all with the software, and am not understanding this graph. Those VE percentages are results, or inputs? The left column is 15.0-97.0 KPa, so that's pressure, 2.2-14.1 psi... so this is fuel pressure variations that vary solely, but seemingly all over the place, per RPM?. I guess what I'm asking is what the heck is this graph displaying and telling us?

Regardless though, that's only one (injection) of the two (carburetor) air:fuel sources you've tried that both result in backfiring, so I still would focus on and suspect incorrect valve/ignition timing. I'd pull the valve cover, set #1 to what's believed to be TDC (re-verifying #1 true TDC if not positively certain) of the compression stroke, and verify that both valves are closed, and then look at where the distributor rotor is pointing. It should be somewhere at or near the point for firing #1... but if it's a good amount past it, that would explain #1 firing early enough to backfire out the intake...

[svocapri]

I'm not familiar at all with the software, and am not understanding this graph.

Mike, the graph is the volumetric efficiency table from the Megasquirt. The left column is the manifold pressure, and the bottom row is the RPM. This table is used to calculate how much fuel to inject into the engine. For any given RPM and manifold pressure, the ECU looks in the table and finds the volumetric efficiency (percent of the cylinder that is filled with air). It then injects enough fuel to meet the AFR requested given the pounds of air it calculated from the VE table.

Those VE percentages are results, or inputs?

This table would normally be considered an input, but during tuning the computer compares the oxygen sensor reading to the commanded AFR. When there is a difference it goes back and adjusts the VE at the point in which there was a discrepancy. It does this numerous times while the engine is running until the commanded AFR matches the oxygen sensor reading.

The table I have shown is after numerous iterations of the tuning cycle and therefore is somewhat of a result. The part that confuses me about this table is why the engine needs so much fuel between 1,400 & 2,000 RPM to get an acceptable O2 reading. I also believe that the O2 reading are correct because the engine was breaking up before those VE cells were increased. The engine now revs through those areas clean and smooth with power whereas before it would fall on its face.

For those who like me are used to reading a vacuum gage, I made this conversion chart from KPA to In HG.

[Walking-Tall]

Gotcha, Julian

You read my mind, lol, because I was thinking a table for fueling for vs RPM would/should be engine load, or vacuum levels. Great chart converting KPa to in-Hg, thank you for that.

Regardless, get on verifying the mechanical stuff, because I think that's where the issue is!

[erratic50]

If you're carrying a lot of vacuum 1400-2000 ask yourself what bearing that will have on timing advance down there. I would say in normal carbureted conditions due to vacuum it's going to be aggressive. Advancing the cam to favor intake only increases this when HOs tend to carry a ton of vacuum down low already.

Pull the advance and see what it does. Identify how to limit how quickly the advance comes in or reduce base timing advance.

You may very well be seeing fuel demand increases as you come into your torque and horsepower production rev ranges.

Factors such as swirl in the head can cause much more rapid burn meaning less timing or more fuel or both in those ranges is needed for a smooth runout across the rev range.

with that fresh valve job, how far up into the head is the valve when it's closed? Again that can generate lots of swirl.