Tuning Timing
Thread Starter
Joined: Jan 2003
Posts: 1,964
Likes: 0
Tuning Timing
I just got my wideband, and I wil eventually be tuning my car when its ready.
How do you guys tune the timing? Is there a certain amount to retard per pound of boost? I guess it would depend on the cfm the turbo is pushing. I read that on the dyno you advance until you detect knock and then back off a bit. I will be street tuning for the most part, do I need something like the J&S Knock Sensor.
Thanks for any input, and if anyone could recommend a link or book that would be great.
How do you guys tune the timing? Is there a certain amount to retard per pound of boost? I guess it would depend on the cfm the turbo is pushing. I read that on the dyno you advance until you detect knock and then back off a bit. I will be street tuning for the most part, do I need something like the J&S Knock Sensor.
Thanks for any input, and if anyone could recommend a link or book that would be great.
Honda-Tech Member
Joined: May 2003
Posts: 230
Likes: 0
From: Hamilton, Ontario, Canada
Re: Tuning Timing (b3435)
thats what I would do, advance until knock and then back off a bit. If your odb2 the ecu will retard the timing when it senses knock. So I guess in order to find this point your going to have to either get an AFC which has a knock sensor warning that reads off the ecu or find a way to interperet the knock sensor signal which I've read is only a microphone and the ecu interprets the AC voltage looking for a particular frequency associated with the pinging sound. Or ya an external knock sensor, this is what I'll have to do since my car doesnt have one to begin with.
Member
Joined: Dec 2000
Posts: 4,848
Likes: 0
From: Islip, CA, USA
Re: Tuning Timing (b3435)
j&s safeguard.
just take out .5 degrees per # to be safe if you have 9:1... if you have 10:1 take out .75 or 1 degree per #. thats to be safe on the street. you arent going to gain much power on 91 octane by adding timing... theres no point in risking it.
just take out .5 degrees per # to be safe if you have 9:1... if you have 10:1 take out .75 or 1 degree per #. thats to be safe on the street. you arent going to gain much power on 91 octane by adding timing... theres no point in risking it.
Thread Starter
Joined: Jan 2003
Posts: 1,964
Likes: 0
Re: Tuning Timing (94goldjungsr)
so I can just set the timing as you described and then there really isn't a need for the J&S correct?
BTW, I know I need to rtfm and I'm on my way to aem's site, but what does the ems do with the factory knock sensor? Can it retard if it sees knock?
edit: It does, just looked it up.
Modified by b3435 at 9:09 PM 9/19/2003
BTW, I know I need to rtfm and I'm on my way to aem's site, but what does the ems do with the factory knock sensor? Can it retard if it sees knock?
edit: It does, just looked it up.
Modified by b3435 at 9:09 PM 9/19/2003
Thread Starter
Joined: Jan 2003
Posts: 1,964
Likes: 0
Re: Tuning Timing (b3435)
Here is a good explanation from the aem forums
Found here
Originally Posted by blkmgk
Okay, after speaking to others and watching\observing and learning here's what I'd share and what I intend to put to good use on a dyno soon....
You want cylinder pressure to remain even across the RPM band in the engine at a given load. Too high and it will detonate\preignite\diesel, too low and you don't make peak power. At low RPM\loads the pressure is low and you must fire the misture earlier to get pressure up when the piston reaches TDC. As peak torque approaches pressure\effeciency have climbed and in order to keep pressure even without broaching into detonation you must retard ignition. As peak torque passes ignition again needs to be advanced to maintain high pressure. You end up with a reverse "wave" that ebbs at peak torque which is also peak effeciency I believe. However without some way of measuring cylinder pressure and a VERY high speed data collection instrument seeing measuring the pressure in the cylinder during an engine cycle is next to impossible. I'm told by someone who works in the diesel industry that's it's fascinating to watch engineers advance and retard injection on a diesel and see the cylinder pressure curves change - I think that's how they do it anyway.
So, how to do this without a lab and tons of money? Strap car to dyno, tune fuel to a good but somewhat conservative A/F, monitor the knock sensor. IF the knock sensor isn't showing spikes (internal log works best BTW) advance timing and make another pull - maybe even a partial pull to get the lower end tuned first. Did power go up? If yes, did it detonate? If no then you've made an improvement - did it lean out? If yes then it was a more effecient burn, pat yourself on the back and give it more fuel. Pull it again and get the A/F right. Advance timing, pull it again, rinse and repeat. When you see knock back off the timing or perhaps try a little fuel but only if power has continued to climb. If power stops climbing despite added timing then back off and perhaps fiddle fuel.
Realize there's a point of diminishing returns. You may reach a point where the additional spark doesn't add much power but puts you much closer to detonation - at that point I'll back off. I'll gladly sacrifice some power to keep an engine in one piece thanks. Also realize that the "93" octane gas you have today could turn out to be 87 next time you fill up and that the more power you make the quicker it'll goto hell when something goes wrong - be careful treading the knock\power edge. Same goes for "race gas" - more than one of us has gotten this with water in it or other issues and it will burn you if you aren't careful.
Where your timing ends up varies with each engine. One might have more carbon or a sharp edge in a combusion cylinder which could limit how much timing it'll take while another won't and will love more timing.
Lastly, when you look at the knock RAW you should see a progression of noise as RPM\power builds. Towering spikes are probably knock. In my case I'll have low noise and then a sudden spike even with low timing I think it's either a tank of crap gas (despite the 2gals or 103 I added) or that this is just random noise - I've not decided yet. I might let small spikes ride if I cannot completely eliminate them even spraying meth with rich mixtures just because they seem to be noise rather than knock. Before I 100% make this choise on MY car I'll try to run it once with almost nothing but race gas in it. YMMV but seeing what it looks like with GOOD gas and on pump would probably be a very good idea. Oh, and consider turning on knock response on both knock sensors (if you have two) just to be safe after your car is tuned.
I hope to be tuning my car using these exact methods Friday if a hurricane doesn't trash my travel plans Hope this helps, Sean and Justin feel free to throw in corrections please....
Oh, and one day I'll break down and explain how I intend to tune fuel via Boost Comp on the dyno
P.S. While tuning on the dyno turn OFF knock response so the ECU doesn't add fuel and toss out timing. You want it to do exactly what you tell it (carefully) in order to best gauge the car's reaction to your tuning changes.Okay, after speaking to others and watching\observing and learning here's what I'd share and what I intend to put to good use on a dyno soon....
You want cylinder pressure to remain even across the RPM band in the engine at a given load. Too high and it will detonate\preignite\diesel, too low and you don't make peak power. At low RPM\loads the pressure is low and you must fire the misture earlier to get pressure up when the piston reaches TDC. As peak torque approaches pressure\effeciency have climbed and in order to keep pressure even without broaching into detonation you must retard ignition. As peak torque passes ignition again needs to be advanced to maintain high pressure. You end up with a reverse "wave" that ebbs at peak torque which is also peak effeciency I believe. However without some way of measuring cylinder pressure and a VERY high speed data collection instrument seeing measuring the pressure in the cylinder during an engine cycle is next to impossible. I'm told by someone who works in the diesel industry that's it's fascinating to watch engineers advance and retard injection on a diesel and see the cylinder pressure curves change - I think that's how they do it anyway.
So, how to do this without a lab and tons of money? Strap car to dyno, tune fuel to a good but somewhat conservative A/F, monitor the knock sensor. IF the knock sensor isn't showing spikes (internal log works best BTW) advance timing and make another pull - maybe even a partial pull to get the lower end tuned first. Did power go up? If yes, did it detonate? If no then you've made an improvement - did it lean out? If yes then it was a more effecient burn, pat yourself on the back and give it more fuel. Pull it again and get the A/F right. Advance timing, pull it again, rinse and repeat. When you see knock back off the timing or perhaps try a little fuel but only if power has continued to climb. If power stops climbing despite added timing then back off and perhaps fiddle fuel.
Realize there's a point of diminishing returns. You may reach a point where the additional spark doesn't add much power but puts you much closer to detonation - at that point I'll back off. I'll gladly sacrifice some power to keep an engine in one piece thanks. Also realize that the "93" octane gas you have today could turn out to be 87 next time you fill up and that the more power you make the quicker it'll goto hell when something goes wrong - be careful treading the knock\power edge. Same goes for "race gas" - more than one of us has gotten this with water in it or other issues and it will burn you if you aren't careful.
Where your timing ends up varies with each engine. One might have more carbon or a sharp edge in a combusion cylinder which could limit how much timing it'll take while another won't and will love more timing.
Lastly, when you look at the knock RAW you should see a progression of noise as RPM\power builds. Towering spikes are probably knock. In my case I'll have low noise and then a sudden spike even with low timing I think it's either a tank of crap gas (despite the 2gals or 103 I added) or that this is just random noise - I've not decided yet. I might let small spikes ride if I cannot completely eliminate them even spraying meth with rich mixtures just because they seem to be noise rather than knock. Before I 100% make this choise on MY car I'll try to run it once with almost nothing but race gas in it. YMMV but seeing what it looks like with GOOD gas and on pump would probably be a very good idea. Oh, and consider turning on knock response on both knock sensors (if you have two) just to be safe after your car is tuned.
I hope to be tuning my car using these exact methods Friday if a hurricane doesn't trash my travel plans Hope this helps, Sean and Justin feel free to throw in corrections please....
Oh, and one day I'll break down and explain how I intend to tune fuel via Boost Comp on the dyno
P.S. While tuning on the dyno turn OFF knock response so the ECU doesn't add fuel and toss out timing. You want it to do exactly what you tell it (carefully) in order to best gauge the car's reaction to your tuning changes.Okay, after speaking to others and watching\observing and learning here's what I'd share and what I intend to put to good use on a dyno soon....
Thread
Thread Starter
Forum
Replies
Last Post
gilou bilou
All Motor / Naturally Aspirated
3
Oct 7, 2005 02:30 PM
