Engine "popping" at high RPM???

[Mechanicaldan]

Hello all,

I'm still at it. I just can't believe how much I've learned about tuning engines over the last semester, and I'm trying to finish this.

Tonight, we were on our dyno and tuning. We just installed some used 14 lb/hr Bosch injectors (10 lb/hr injectors max out at 8500 RPM), and have been getting our fuel map set up starting from about 4000 RPM. We were able to map all the way up to 9000 RPM, when we started getting a "popping" sound. The fuel pressure is at 45 psi. The plugs look a little light, but our oxygen sensor is reading in the same place it has all night at 12.5:1, or about 800-850 mV. We pulled the injectors out and ran the injector test on our ECU, and all the spray streams look fairly constant. I can adjust the fuel + or - 30% with the lambda sensor giving me a reading anywhere from 10:1 to 14:1. The engine pops the whole time. There is a pop about every second or sometimes a few pops per second, with no distinct pattern. We are running heat range 8 plugs gapped at 0.030". We are running the stock coils. The ignition is set at 40 degrees advance. Adjusting the advance + or - 5 degrees has no effect either. Even when I put the ECU in closed loop, and watch it make fuel corrections, I still hear popping.

Is it maybe lean popping? Maybe backfiring because I'm to rich? Why do I have a correct reading on the lambda sensor? Could I be maxing out my restrictor? I've done calculations that say I should only be at around 0.5 Mach. Anyone have any suggestions? What other information can I provide to help you, help me?

Thanks,

Dan http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[AZWildcat]

Dan,

The more important question is what did the dyno graphs look like? Audible misfires are plainly obvious on a dyno graph, and unaudible misses will have 90 degree corners on the graph. Constant audible misfires will look like a heart monitor. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Depending on how your graphs look and how your exhaust is set, it's possible you're hearing a little after burning. F40 Ferraris are notorious for doing that. 12.5:1 may be on the rich side for you motor promoting the afterburn. If the graph looks smooth, I wouldn't worry.

Dennis Walsh
http://www.wheelkinetics.com/hre/supra/suprax.jpg

[Dan Deussen @ Weber Motor]

Can you adjust the dwell (charge) time of your coils in the ECU software? Maybe you don't charge the coils long enough at high rpm. Connect an o-scope on the trigger wire (the one coming from the ECU) of your coil and measure the charge time. If you have a multi-channel scope you can connect the crank trigger to another channel to have a reference of the charging and firing of the coils in relation to the crankshaft position.

Keep us posted how things go!!!

Dan

[Kevin Hall]

Is your exhaust tight, or are there any small leaks. This may make random popping noises, as air can come in through the holes, and cause the excess fuel you are adding to burn. Even 14:1 is a bit rich. 14.5:1 is about right, but whatever makes the most power is best.

Just a thought. Good luck!!

Kevin Hall
University of Saskatchewan
Engine Guy in Need of Help

[Mechanicaldan]

The dyno that we are using simply gives us a load cell reading in pounds. We have to manually multiply the lever arm to get ft-lbs. We are writing down numbers manually to get power curves.

The popping is a problem because each time it pops the lambda sensor swings back and forth, making tuning difficult. Maybe I should just pull the exhaust collector off and look for flames coming from the muffler? The exhaust leaks are possible, but why only now at high RPM? We have had the header on an off a few times, and had lots of backfires trying to get the engine to start in the past.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Mechanicaldan]

I believe the dwell time is set at 3000 microseconds, the factory default. There is a note that suggests this, unless we know we need to change it. We can change it, but it applies to the entire map. We can adjust it from 1000-9000 microseconds. So, we might not be burning the entire mixture at high RPM? Any suggestions? Change it to 5000 microseconds?

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[This message was edited by Daniniowa on April 30, 2003 at 04:28 PM.]

[Dan Deussen @ Weber Motor]

3000 should probably be an okay value for the stock coils. Most high performance coils actually use higher charge current with lower charge time. It might just be something to try to change in order to eliminate the possibility of this being the problem.

5000 microseconds is definitely too much to try if you are running waste spark (spark every crankshaft revolution), since at 12k rpm your charge time would take a full revolution and there would be no time to fire the coils. I would suggest trying 2000 and 4000 microseconds and just see what happens.

I would still hook up an o-scope and take a look at the coil and injector signals from the ECU to make sure they are not cutting out for any reason.

Also keep track of the battery voltage and make sure it is being charged by the alternator.

Another possibility that I could think off is that you start to get air distribution problems at high rpms and that you are starving a cylinder and the unburned gas is causing the backfiring.

The cheapest and most powerful tool to detect those kind of problems are k-type thermocouples mounted in each exhaust header about 4-5 inches downstream of the port. This should give you a pretty good idea how even the combustion between the individual cylinders is.

Keep us posted.

[Angry Joe]

Are you using a Hall Effect sensor? Is it inside the engine case? We had a similar problem that basically acted as a rev limiter above 10-12k. Turning the sensor closer to the crank helped the problem, though it did not completely solve it on the dyno durning extended pulls. I might try moving it closer still if it reoccurs in the car.

Thanks to Brian Lewis for helping us with that one by the way, I NEVER would have figured it out otherwise...

Lehigh Formula SAE

www.lehigh.edu/~insae/formula (http://www.lehigh.edu/~insae/formula)

[Mechanicaldan]

We are running an equal length intake manifold, having a conical plenum where the primary tubes connect to the bottom in the same order as the cylinders are fired, hoping to keep the air constantly moving from one plenum to the next without every really stopping or having to make major directional changes.

We have an infared temperature sensor, and all the primary exhaust tubes have been running around the same temperature. 700-800+ depending on the dyno load. The temperatures are taken at about 3" from the head.

The crank sensor is the stock Kawasaki sensor. We had a 32-3 tooth wheel laser cut to the same diameter as the stock wheel, which it bolts to. The stock wheel has had the teeth removed. The sensor is spaced over by one washer thickness, which is about the same thickness as our toothed wheel.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Mechanicaldan]

The voltage regulator is not connected. The engine is running from a large Interstate car battery, and it is continually being charged at 12 amps from a battery charger.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Scott Harsila]

Are you sure you have enough resolution on the crank sensor? At 32-3 and 12k you might start to smear the signal. Hook up an oscilliscope to the hall effect crank signal and make sure you have a nice saw tooth looking signal with a regular pause for the missing teeth. Rev through the RPM range and make sure the frequency changes smoothly without huge jumps or voltage loss.

[B Lewis @ PE Engine Management]

Hey Dan,

If you are making power and the popping starts at high speed under a load, I would bet dinner you are just dealing with exhaust leaks. Our experience has been this occurs more at higher speeds because of the increased heat in the exhaust.

Also, try not to get caught up in tuning for a specific A/F ratio. What you want is power, regardless of the mixture. Just use the O2 as a reference.

Good luck and see you at competition.

Brian Lewis
Performance Electronics, Ltd.
www.pe-ltd.com (http://www.pe-ltd.com)
"Complete Engine Management Systems"

[Moeye]

Well our engine made a couple pops at high speeds tonight on the dyno. Well, it wasn't really the engine, but it was the chain flying off.

The first pop was fine, then the second pop came a split second later. Sparks flew and a bearing decided that being two pieces was better than one. Fortunately, nobody got hurt.

A typical night in the dyno cell at U of I:

A: Whoa, why is the power suddenly dropping off?
B: Choking?
A: No way, too low for choking.
B: Hmmm, let's do it again and see what happens.

[Richard Lewis]

Same thing here... a nice intake backfire tonight grenaded our intake manifold, blowing it off the engine and apart. Causes are still under investigation...

-------------------------
UVIC Formula SAE Team
http://members.shaw.ca/drax77/Formula%20UVic%20Sig.jpg
http://uvic.fsae.ca

[Mechanicaldan]

We'll be hooking up an oscilloscope tonight. Here's another thought. Our intake is designed to make maximum power right where we are starting to get the popping. How much of an increase in fuel would be likely?

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Mechanicaldan]

Well, we ran the engine again tonight, and same thing, still popping. We did get an oscilloscope hooked up, and the signal is clean all the way to 12,000 RPM. We can actually rev the engine freely to 12,000 RPM without any problem, but once we put a load on the engine, pop, pop, pop... We took the exhaust off tonight and did see that we have the black, sooty signs of a few small exhaust leaks. Seems the nuts and bolt have loosened a little from the vibration. We'll be getting a new gasket tomorrow. I have not adjusted the dwell, but do intend to if the exhaust is not the problem.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Mechanicaldan]

Well, I think I got the popping to go away, but now I have some BIG backfires.

I increased the dwell from 3000 microseconds to 3500, then to 3800, then to 4000 microseconds, and I believe the popping has gone away. Now the headers glow red fairly quickly, and I'm getting BIG backfires. I can tell I'm too rich. OUR oxygen sensor or setting for it's scale must be off. The plugs look dark at 12.5:1 (850-900 mV), and the engine doesn't clean up until we get down to 14:1 (700-750 mV) or so.

Can some please tell me what millivolts I should be looking for when tuning. The air/fuel ratio seems useless.

Right now, I seem to do the best just tuning by ear until the engine sounds smooth. No popping, no backfiring.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Charlie]

Don't trust a standard O2 sensor. They are not very accurate. If you look at a voltage vs. A/F graph it is like an on/off (rich/lean) switch, not a good measuring tool.

If you have glowing headers (really badly, it's fine for them to glow a bit in steady state tuning) then try advancing your timing to lower EGTs.

-Charlie Ping
Auburn University FSAE 1999-present

[Kevin Hall]

Don't worry too hard about A/F ratio. 14:1 is a bit richer than stochiometric, and should give decent economy. Even 15:1 can be run with reasonable safety. Whatever makes the power should be what you run. If it sounds like crap at 12:1, then lean it out some more. If it makes bigger power at 13.5:1 than at any other mixture, run that.

As for EGO sensors, they are alright, but you cannot have any exhaust leaks. The spark plug test is better than EGO in general. The EGO is there to give a bit of regulation to your MAPS, because your dyno isn't setup in the rain, or at 35*F, or at 110*F. Make the MAPS work, and it'll be your little buddy on the track (or your worst nightmare if it's not working).

What kind of timing are you running at this high RPM. We are running 42* from 10-12000. This seems to make decent power (65 hp) with a crappy intake. Big backfires and gklowing headers usually indicate that you need more timing.

Good luck!!

Kevin Hall
University of Saskatchewan
Engine Guy in Need of Help

[Scott Harsila]

I second the suggestion of being way too rich and the engine misfiring due to that. The O2 sensors will spike lean as soon as it sees a misfire and all that free oxygen flows down the pipe. The excessive rich mixture could be causing continued combustion in the headers causing them to glow red. I might try a -20% trim with a set of new plugs, do a quick run with or without load (your choice) and pull the plugs and do a plug reading. If they are still black, less fuel, if they show white then your know that you took out too much. Sometimes you have to make extreme changes to make sure you are in the ball park. In general if you can smell raw gas, you know you are way off. But a little disclaimer, I am not there and cannot see the computer screen of your ECU, so I am not liable if you smoke your motor. Just some suggestions.

University of Washington
Formula SAE 2001-2003

[Mechanicaldan]

Timing is at 40 degrees advance by 8000 RPM, and stays there until redline.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Engine Team Leader

[Kevin Hall]

That seems to be enough timing. Lean 'er out 'til she runs better.

Good luck.

Kevin Hall
University of Saskatchewan
Engine Guy in Need of Help