self tuning efi

What efi system do you have ? Comsync uses the Bosch 4.0 02 sensors and most other systems use the NTK 02s and I can tell you from experience ( atleast on a turbo car ) the Bosch 02s read leaner then the NTK don't know if it's the system or placement of the sensors . We run around 4.0 to 4.3 hope this helps

 

. I would say if running the comsync efi start out at 3.6 on a blown motor and read the plugs

 

Rossele,
I would agree with Troy on our system but, I am not familiar with the Holley system. Are you running a mag or coil pack? Are you running 4 magnets or 10 degree wheel or 60-2 wheel. The more aggressive you want to run your timing the fatter I would make the AFR. 3.6 AFR should be safe but, every head, bore, stroke setup is a bit different on what it likes. Since you don't have EGTs unless you have a Racepak, I would start as safe as possible read the plugs and lean it out or increase the timing (be careful with this). I think the Holley only has one O2 sensor. Put it in the hole that you are most likely to burn up (6 or 7 or what ever it is). You really need 8 O2s and 8 EGTs if you want to safely dial it in. I know this is not a precise answer but, it is a safe one. Again, I am not that familiar with the Holley system other than I know some of it's weaknesses (not trying to bash it either). How much data do you get off the Holley on what it did for each cylinder firing?
Blake

 

Hello. We have 8 coil packs from holley. The crank has the 4 magnets with the hall effect sensor. Aem cam sensor. There are 2 ox sensors. We started with 3.8 a/f and are up to 4.3. Pretty close on the tune. We just turned the boost from 20-30. I noticed your comment that the ox sensors are not great for alcohol? We should be getting the egt sensors. We are at 34 degrees for timing.

 

Rossele,
I think you are about right (without seeing your data). I have found that the best power is between 4.0 and 4.4 on some motors, again it completely depends on your heads, bore, stroke and intake temperature. Going from 20-30lbs boost though, I would fatten it up and pull some timing out and then look at the data and read the plugs. Do you know your intake temperature? That will affect what you need to do. 500lbs an hour per cylinder seems a bit light for 30lbs of boost to me (I like to be safe). Are you using the IGN-1A coils? What is your dwell and what gap are you running? Does the Holley have timed injection on the ports? Sorry, I am not that familiar with the Holley. I have a few people convert over because they said the Holley was burning up the coils. Not sure if that is a real problem with the Holley system or not, it is hard enough staying on top of our own stuff.
Anyway, I like the procharger system. Just curios, Is it a direct drive or belt drive? We had a few issues with crank sensors with the direct drive. As far as the O2 sensors, if you are running 4.3 and 34 degrees of timing, I don't think you will have too many problems with the O2 sensors and they should read very accurately. That changes when the timing comes down and the AFR goes down and you start to get raw fuel coming out the port then the O2s become much less reliable.
Where do you have the magnets set on the crank around 40 degrees BTDC? It is usually best to have the magnets as close to the highest timing you want to run. Take everything I said here with a grain of salt because again, I am not that familiar with the Holley system, I think (not sure) they are a time-based system and we are a synchronous system and they act differently.
Blake

 

We have the holley coils. Ramping the dwell from 4.0 low rpm to 6 up high. Without the use of learn it is a little scarry taking it for a pass. I did not know the ox sensors were that slow. It will be in th 6.7 in the quarter mile. Should I keep the learn on for a pass or 2?

 

Rossele,
I am not familiar at all with the learn function on the Holley but, if you are scared to make a pass without it then by all means don't. I am assuming the "Learn" function will adjust your fuel table, if so try to setting your fuel fat and let it lean it out rather than going the other way. There is a Comsync Top Sportman out there running an F3 blower that runs in the 6.10s all though in the high teens and in the 20s in hot weather. You should do well. I would pull some timing out and lower your AFR until you are very comfortable with how your plugs look. You probably don't need to run 6ms dwell on the top either 5ms should be more than plenty but, again, I am not familiar with the Holley system and the IGN-1A coils can handle 6ms.
Enjoy your new setup,
Blake

 

Blake could you expand a bit on this? Thanks.

 

TOL,
Time based EFI uses a "time based" clock to tell the ECU when to sample the data, there are usually 4 or 8 priorities levels. say your total cycle time is 100ms or .1 seconds then every sensor and every calculation done on those sensors will be completed 10 times a second. This is how the ECU from the 90's and early 2000's work. There are priorities in that 10 times a second. If their are 8 division and you have a high priority sensor such as the MAP it will be sampled 80 times a second, throttle might be a second level or 40 times a second, MAT might be 20 times a second and Engine temp 10 times a second. This sampling rate never changes no matter what the RPM the motor is running. So, at high RPM the MAP sensor of all the sensor readings might be used for 2 or 3 cylinders in a row and next level sensors for 5 to 10 cylinder and so on. Time based EFI allows the use of slower processors and ADCs and therefore are much less expensive and the code is much, much easier.
A synchronous system takes data synchronous to the crank position (same position every single time). The Comsync (COMpletely SYNChronous) system samples all the sensors necessary and does all the math for the cylinder it is working on every single cylinder firing. So, the faster the RPM the more data it is taking and processing. This make the math a much more intense because the sample rate is always changing but, provides much faster and more accurate control of the fuel and timing than a "time based" system. The Microprocessor and ADC parts have to be much faster usually 10 to 30 times faster in order to do this and that costs more. I have tested the Comsync system to over 32,000 RPM and there are no problems but, the coil packs can not go near that high of an RPM but, this allows me more processing power to do other things and have plenty of time to spare.
Blake

 

Thanks for posting Blake. I know exactly what you are alluding to in your post, and I am with you 100%. That's why I felt I needed to nudge you to expand a bit for others so they can also read/understand/learn.

Using time based control on an engine that has a fast rate of change of RPM (up or down) is a bit of a guessing game at times, just based on the approach. Do you still prefer these days the 60-2 toothed wheel approach for your stuff, or what about a precision angle encoder on the crank instead of the wheel/sensor? Any benefit to the later?

A V8 engine that changes revs at maybe 500-1000 rpm/second is not that hard to manage based on time. One that changes revs 1800-2500 revs, or more, per second is a whole different animal.

 

TOL,
I use a 10 degree wheel for V6 Even or Odd fire, V8s and for V12s. It doesn't work for V10s but, over the years I have found it to be about the best for high RPM engines and for engines that have to have a smaller crank trigger wheel (which is not a good idea) and for starting high compression engines where the crank almost stops approaching TDC during cranking. I like the idea of the 60-2 because with good software you can start the engine without a camshaft sensor but, they are more prone to noise because of the smaller teeth and weaker signal. We can maintain about .1 degree accuracy on timing with the 10 degree wheel and the signal is much more robust. The accuracy actually goes up with higher RPM. Also, it is impossible to have accurate timing with a 4 magnet wheel, I don't care what they say. I think an encoder would be great but, the ones I played with were not very noise immune during ignition and they require a lot of wires to get one real-time without a delay. I am not sure how durable they would be over time with the vibrations in the front of the engine and they need to be attached dead center on the crank. That is not very practical on a lot of engines. Being a digital and precision guy, I love hall-effect sensors but, in reality they are not nearly as reliable as the VR sensors and you have to go with what works.
Caring onboard O2 to feed into the engine during the run would be ideal but, just not practical at all. so, some settle for Nitrous-Oxide or Nitro-Methane
Blake

 

When you reference a 10 degree wheel , is that a 36-2 wheel? Pm me your number, I'd like to talk to you about a system for a twin turbo car.

 

boostedgt500,
I am new at this, I think I sent you a PM. Not 100% sure though.
Thanks,
Blake

 

Much better wheel then the Holley

 

The new phone book is here! The new phone book is here! Well it feels like that, WWW.COMSYNCRACING.COM is now live on the web. There is still work to do but, you can get some ideas on what we have done and what is possible. Please take a look and let us know if you can think of any improvements that can be made to the website.

 

Best news in a long time one of the baddest systems on the planet don't miss out

 

If anybody is in Pomona this weekend stop by the Jegs Top Dragster and ask for Ryan (573-690-6083) he can show you the Comsync efi system they have on their car.

 

Comsync EFI congratulates Mike Coughlin for his win in Top Dragster at NHRA winter nationals in Pomona today.
Blake