Tuning ignition / timing / spark without dyno

[Onime]

First of all, I'm new to tuning.

I want to tune my car without dyno. I know why dyno is better and I know that on dyno, I can get maximum power possible.
I know that it is easy on dyno to test engine on every possible load and so on.

But, still I want to tune my car on the road. I understand how to adjust VE table, but this question is about ignition.

My idea of tuning ignition on a road is:

(First of all VE table should be adjusted at least on basic level, before start adjusting ignition)

1) At the begining of tune retard ignition on entire map, especially in high boost region.
2) Then get good knock detection headphones, and go ride a car.
3) At the time of ride step by step advance ignition at every possible engine load, until I hear knock.
If engine starts knocking, then I can retard ignition with 2-3 degrees for that region on the map.

As this is general question about adjusting ignition table, I don't want anything specific from you, just tell me if this is right direction,
or maybe there is somthing important I should know?

[racingmini_mtl]

Using knock to tune the ignition will not give you the best ignition timing. It will only give timing that will not give knock for the testing conditions. There is no guarantee that it will not give you knock under different conditions.

But more importantly, it has absolutely nothing to do with optimal timing; some engines will never knock even with wildly off timing and others will knock way before the best timing.

Jean

[slow_hemi6]

Tuning a turbocharged engine by feel will cost you a lot more money than tuning it on a dyno in the long run. I can't fathom the fact that someone embraces efi and is tech savvy but still wants to tune like they have a carby and dissy on a flathead ford.
All car manufacturers have been dyno tuning since they were available. Plenty of pictures out there of the Chrysler engineers with the max wedges and hemi's on the dyno. I have seen Chrysler 413 long ram dyno test sheets dated 1959. There is a reason for dyno tuning even with the resources of a factory R&D team.

[E4ODnut]

Onime,
I feel your pain. In some situations dyno tuning can be prohibitively expensive and one has to settle for the next best thing. You didn't state the details of the engine you are tuning, that can make a difference. As was mentioned, tuning a turbo'd engine that way is kind of risky because when things go wrong they tend to do so quickly and quite dramatically.

Jean,
I'm a bit surprised by your response. I agree that tuning for knock may not result in optimum timing, depending on your particular engine, but if your engine knocks before optimum timing then the engine is knock limited and there is little you can do to improve it, assuming that you have already tried all the tricks you can to reduce detonation. Perhaps my thinking is wrong, if so please set me straight.

Knock tuning may not result in optimum spark timing, but it should result in safe spark timing.

[racingmini_mtl]

Jean,
I'm a bit surprised by your response. I agree that tuning for knock may not result in optimum timing, depending on your particular engine, but if your engine knocks before optimum timing then the engine is knock limited and there is little you can do to improve it, assuming that you have already tried all the tricks you can to reduce detonation. Perhaps my thinking is wrong, if so please set me straight.

Knock tuning may not result in optimum spark timing, but it should result in safe spark timing.

If your engine is knock limited then knock tuning is safe. If it's not then you can set timing with way too much advance and then you put a lot of strain on the engine components in addition to having less power and efficiency.

And how do you know that your specific engine is knock limited? Unless you know from other methods that your specific engine setup is knock limited then you can't rely on using knock. And if your engine is knock limited, you're either not using the right fuel or the components you used to build your engine were poorly chosen. There are exceptions and extreme conditions but being knock limited should not be the normal state of the engine.

So knock detection should not be used as a tuning aid but as a protection mechanism for exceptional circumstances.

Jean

[vagman]

+1

[Onime]

Using knock to tune the ignition will not give you the best ignition timing. It will only give timing that will not give knock for the testing conditions. There is no guarantee that it will not give you knock under different conditions.

But more importantly, it has absolutely nothing to do with optimal timing; some engines will never knock even with wildly off timing and others will knock way before the best timing.

Jean

Let define what is optimum spark technically.

Following is my opinion reality may be different.

Optimum spark is that satisfies following two requirements.

1) Pressure in cylinder is far from TDC, this way we get more mechanical leverage on crankshaft. (Law of the lever)

2) Peak pressure in cylinder is maximum possible for given mixture without knock.

These two limit each other. We cannot satisfy both with 100%. Let say we have spark angle so that, it creates maximum
possible peak pressure in our cylinder, this satisfies second requirement, but does not mean it will create maximum power,
because we don't know how it satisfies first requirement. It may be better to retard spark so that peak pressure is built
little bit later, which will decrease it, but we get it farther from TDC. This satisfies first requirement more. Yes maybe
peak pressure is decreased but it is on better angle and piston pushes crankshaft more effectively.

So satisfying one from these two requirements more, leaves second one less satisfied. But somewhere between is optimal point.'

The approach I have described in my initial post may let you find out spark angles to satisfy second requirement, but not first one.

What else I'm missing or maybe I'm completely wrong?

I feel your pain. In some situations dyno tuning can be prohibitively expensive and one has to settle for the next best thing. You didn't state the details of the engine you are tuning, that can make a difference.

Yes fu**g situation...

Engine is BMW M44,
4 cylinder, 85mm bore x 83.5mm stroke
lowered 8.6:1 CR on stock internals
target boost is about 1 bar.

I can't fathom the fact that someone embraces efi and is tech savvy but still wants to tune like they have a carby and dissy on a flathead ford.

I do not want, I'm forced, it will be hard to explain why.

[racingmini_mtl]

2) Peak pressure in cylinder is maximum possible for given mixture without knock.

That is not correct. You can have localized hot spots, localized high pressure point, bad fuel,... that will cause knock. Those have very little to do with peak pressure.

Knock does not give you the information you think it does. It only tells you there is uncontrolled ignition.

Jean

[billr]

And there are sort of two different kinds of knock... if ignition is too advanced then the F/A charge starts burning at the plug but doesn't propagate smoothly, at some point the rest of the F/A all ignites simultaneously (detonates) resulting in the "knock" violence. However, the ignition advance could be well below that point, yet a hot-spot in the chamber could cause pre-ignition of the F/A charge before the spark occurs, effectively "advancing" ignition in an unpredictable manner. I'm not sure, but I think that pre-ignition is what the boosted-intake folks really worry about.

[vagman]

You(Onime) have to do some search about what "MBT" and "Spark hook test" are.

[Onime]

OK, afters arguments posted here I decided to go on dyno within 2-3 weeks, see how it goes.

Knock does not give you the information you think it does. It only tells you there is uncontrolled ignition.

Yes Jean, looks like I have incorrect view on this part.

yet a hot-spot in the chamber could cause pre-ignition of the F/A charge before the spark occurs, effectively "advancing" ignition in an unpredictable manner.

Yes I think it can happen for example when you have bad fuel, running lean in high boost and intake air temperarue is hot, otherwise on right CR and boost it should not happen.

You(Onime) have to do some search about what "MBT" and "Spark hook test" are.

Tnx for giving right direction, I found few good resorces just started reading.

[kaeman]

I understand your issue, but correct timing for every engine is different.. the correct timing for your engine is designated by the components you have chosen to use in your build, the head design, the compression ratio, the top of the piston design, ie flattop, domed, dished.. all effect correct timing and will all be effected by the choice of fuel and additives in the fuel you may be using, I did some reading and found that the manufacturer of my heads say they require 36 to 38 degrees total advance for best power. I emailed them and gave them the specifics of my engine and they gave me recommendations for my timing. I am currently working on my fuel map and then going to add timing control via the ecu to the mix so I can add timing to what the manufacturer recommends... I believe that I have read somewhere that optimal spark advance gives your engine max cylinder pressure at 12 to 15 degrees after tdc on the power stroke. I would recommend reading matt cramers book on tuning and building efi systems.... he has lots of valuable info....

[ifreaknik]

I'm tuning my car by the seat of the pants and here's how it was done. I found a forum that host the same cars that I do, I found a guy with a build that is similar to mine then I reached out to him. I asked him a simple question . . . "What's your ignition map read at peak torque?"

He told me 33 BTDC at 4500RPMs. Knowing my idle is at 10BTDC - I scaled the numbers up and adjusted my VE accordingly. Then I went in and turned on my assist like AE and etc. General knowledge about 3D maps allowed me to smooth the map out and at this time I have no issue or knock. #BeastMode

I would NOT suggest you tune the the car based on KNOCK for the reasons you have already listed. Each motor is different and your ign-time may be different. Make sure you know the variables to each equation like fuel type (e85, 93, etc). Missing details like injector size and plugs could get you on a knock ride much faster then you anticipated so play it smart.

[cygnus x-1]

The way to tune ignition on the street is the same way you would tune at the drag strip. You do a few runs at full throttle with data logging turned on and look at the results with Megalogviewer. Make some adjustments, do a few more runs, then compare results to see what improved and what didn't. Keep doing this until you are satisfied with the results. All the while it's a good idea to have some sort of knock monitoring to make sure you don't advance the timing too far if the engine is knock limited. Some will reach best torque before knocking, others won't.

C|

[ifreaknik]

The way to tune ignition on the street is the same way you would tune at the drag strip. You do a few runs at full throttle with data logging turned on and look at the results with Megalogviewer. Make some adjustments, do a few more runs, then compare results to see what improved and what didn't. Keep doing this until you are satisfied with the results. All the while it's a good idea to have some sort of knock monitoring to make sure you don't advance the timing too far if the engine is knock limited. Some will reach best torque before knocking, others won't.

C|

sounds legit - any advice for working with the ML??

[vagman]

Here is a good video what is the proper way of finding the correct ign. timing. All other mentioned methods are far away from the good measurement practice.
https://www.youtube.com/watch?v=35bZTek ... e=youtu.be

[ifreaknik]

Here is a good video what is the proper way of finding the correct ign. timing. All other mentioned methods are far away from the good measurement practice.
https://www.youtube.com/watch?v=35bZTek ... e=youtu.be

Watched the video and that doesn't seem applicable with out a dyno and a MoTec. However the case it is cool as crap none the less.

[dontz125]

Here is a good video what is the proper way of finding the correct ign. timing. All other mentioned methods are far away from the good measurement practice.
https://www.youtube.com/watch?v=35bZTek ... e=youtu.be

Watched the video and that doesn't seem applicable with out a dyno and a MoTec.

You really aren't paying attention.

[E4ODnut]

That was an excellent link!

It confirmed my on the road tuning in that torque falls off quite quickly after ~13:1 AFR leaning, but hangs on quite a while richening. As for timing it also confirms that there is a relatively wide range of timing that will change little in torque, as long as you are not knock limited.

Thanks for posting that.

[242ATL]

So that's what a spark hook test looks like. Thanks for the link!