Continuing ignition timing debate from the 416 thread.

[92b]

You have no idea doing sweep tests if the engine wants the whole curve moved up or down or parts of the curve moved up or down.

You can see it and I posted an example.

It’s about finding MBT, which is Minimum spark advance for Best Torque. It’s like guys want to use the maximum timing they can, even if it doesn’t make more power.

A person might get by with that but it certainly doesn't leave any cushion. I tell my customers " the least amount of timing that doesn't lose significant power"

90% of the nasty driving junk cars out there are that way because the tuning is wrong. But you can’t say something or someone is wrong because hate.

The cars that are that bad are not likely to be off by 1-3% on their timing for maximum performance. Maybe somewhere else in their tune.

 

[Newbomb Turk]

You can see it and I posted an example.

A person might get by with that but it certainly doesn't leave any cushion. I tell my customers " the least amount of timing that doesn't lose significant power"

The cars that are that bad are not likely to be off by 1-3% on their timing for maximum performance. Maybe somewhere else in their tune.

The least amount of timing that doesnt lose power is MBT. Im still waiting to hear how locked out timing can accomplish that.

 

[92b]

How do you know what the engine wants for timing at a specific rpm (4500 for example) vs any other rpm (3100 for example) if you don’t park it there and test? Sure you can ball park and ramp the timing through that range but how do you know you’re giving the engine “what it wants”? How do you quantify it?

The same way you see it at steady state. It will lose power at that rpm. The difference is you will be testing it in the same manor as it will be used. Read the superflow paper that turk posted if you haven't already.

 

[92b]

As TT5.9mag is saying, how do you get a curve on a new engine?

Same way you do when steady state testing. You start on the conservative side or your best guess based on past experiences. If you don't feel confident about the combination you are working with do a quick short sweep.

 

[92b]

I imagine if you want to an engine probably benefit some by having each cylinder have it's own curve.

We all would like to have everything perfect wouldn't we.

 

[92b]

And you can do that.

The issue becomes the time and the hardware to accurately sort that out.

I forget the name of the channel but it’s Ben Strader’s channel.

He did at least a couple of videos on in cylinder pressure testing or some form of cylinder measurement.

What they learned gained a fair amount of power on a very well tuned engine.

I actually looked into the sensor and related hardware and software and it was far too expensive for my customers.

It’s hard enough to get them to pay for two days on the dyno and the 300 bucks I charge to make the curve in the distributor match the curve the engine wants.

It’s all about money and time. I can do steady state testing and fit the curve to the distributor and that’s more than many dyno shops do, and I watch as many videos as I can find.

I’m not saying they don’t test for MBT but if they do they don’t show it.

I agree with nearly all of that.

 

[92b]

Not really.
An engine should go from 600 to 1400 rpm without enrichment.
Pump shot is needed only when the throttle position changes rapidly. It shows as a blip on WBO2.
Enrichment is needed for full load.

I think his comparison is to full load steady state testing as opposed to sweep testing on a dyno. Check out the super flow paper that turk posted in this thread to see what he is refering to.

 

[92b]

Ok, so you use your best guess to get a starting curve. I can get close doing that, but it will still not be what it really wants for MBT.

It might be. You won't know until you are done sweeping.

How do you determine what your next move is

If you re-read my post it says how I do it. I always go down in timing first. This is the safest way to do it. Find the least amount of timing that doesn't lose significant power first.

if you say raise the curve and it makes more power at peak but loses power at and around peak torque?

If you went down on timing first like I do, then now you know what the motor wants for timing at peak torque. I assume when you said " peak" you ment peak hp.

 

[92b]

Average power is what moves the car. In your 602 crate engine example that tiny percentage can be the difference between qualifying for the A main and coming through the heats.

There is no end to tiny differences. There is an end to my customers money.

Plus, you cant account for drivability on the dyno. Bill Jenkins called it the “nebulous factor” in his book.

You are right. Are you suggesting we not use dynos?

And the issue isnt being under timed in an area, it’s being over timed in an area.

I didn't see where someone was suggesting we should overtime engines.

 

[92b]

Like I posted before, how many guys would end up with a curve that needed 25 initial at 1k and 34 total at 6500? That’s what a 505 hp 395 inch iron headed small block wanted.

And I gave you an example of a motor that didn't loose power. To me what that says is it depends on the combination. You need to do a cost benefit analysis in the end.

 

[92b]

The least amount of timing that doesnt lose power is MBT. Im still waiting to hear how locked out timing can accomplish that.

Your wait is over, I already posted it.

 

[Newbomb Turk]

It might be. You won't know until you are done sweeping.

If you re-read my post it says how I do it. I always go down in timing first. This is the safest way to do it. Find the least amount of timing that doesn't lose significant power first.

If you went down on timing first like I do, then now you know what the motor wants for timing at peak torque. I assume when you said " peak" you ment peak hp.

I know how you do it. I get it.

If I have the timing locked out and the customer won’t pay for me to steady state test it I make a pull with the timing pretty slow, say 26 degrees. That gets me pretty close on the peaks.

Then I go the peaks, load it down and move the timing around and watch the numbers.

I still need to put a curve in the distributor so most of it is a guess.

The distributor in the 395 mentioned above wants 25 degrees initial at 1k and IIRC it’s only 28 at 4500 or close to it. So it’s pretty slow through there and the it gets faster from peak torque to peak power.

I didn’t plot his curve but I should have. It’s slow to peak torque and it’s faster from peak torque to peak power.

Again, I don’t know how you figure that out doing a sweep test.

 

[Newbomb Turk]

And I gave you an example of a motor that didn't loose power. To me what that says is it depends on the combination. You need to do a cost benefit analysis in the end.

Is that the engine you posted the screen shot of the curves?

 

[92b]

By the way, I appreciate the well thought out conversation we are having here. I know I'm not going to be right about everything or possibly anything but I know I will be better informed in the end. Even if better informed is contrary to what I thought coming into this conversation.

 

[Newbomb Turk]

Your wait is over, I already posted it.

No, you didnt. You need to explain how you can have the same timing at peak torque and peak power.

Unless the engine has a flat VE curve or close to it, there is no way that can be possible.

Now that I think about it, you could do a sweep and look at VE and build a curve based on that.

If VE didnt matter then there would be no need for a vacuum advance.

If I was younger and lived where you could buy hardware to do in cylinder combustion analysis I would do it. But it’s just too expensive to do when it’s hard to get guys to pay for two days on the dyno.

If it’s a roller cam I’ll let them do it in one day, but if it needs cam break in time and a spring change then it’s two days. And it’s a full two days if I have to break in pistons. There is no way you can do all that in a day. At least I can’t do it.

 

[92b]

I know how you do it. I get it.

If I have the timing locked out and the customer won’t pay for me to steady state test it I make a pull with the timing pretty slow, say 26 degrees. That gets me pretty close on the peaks.

Then I go the peaks, load it down and move the timing around and watch the numbers.

I still need to put a curve in the distributor so most of it is a guess.

The distributor in the 395 mentioned above wants 25 degrees initial at 1k and IIRC it’s only 28 at 4500 or close to it. So it’s pretty slow through there and the it gets faster from peak torque to peak power.

I didn’t plot his curve but I should have. It’s slow to peak torque and it’s faster from peak torque to peak power.

Again, I don’t know how you figure that out doing a sweep test.

If what you mean by "a sweep test" is a single sweep test then we are missing understanding each other. It will take more than one sweep test.

 

[TT5.9mag]

I don't think anyone is suggesting that you should intentionally over time an engine. But rather you may give up a tiny percentage by having it under timed in an area. To avoid the exta cost of building a custom curve for that tiny percentage left on the table. Not saying that is how I would do it for myself but for some of my customers that is the option they would choose.

No one intentionally over times an engine I hope. But locking out the timing and setting it at 34 (or whatever) IS overtiming an engine. That grossly overtime’s it at peak torque where it is the MOST vulnerable to detonation and is the basis for this entire thread.

 

[Mopar Sam]

The distributor in the 395 mentioned above wants 25 degrees initial at 1k and IIRC it’s only 28 at 4500 or close to it. So it’s pretty slow through there and the it gets faster from peak torque to peak power.

Are you using some sort of box to get a curve like that?

 

[92b]

Is that the engine you posted the screen shot of the curves?

yes

 

[Newbomb Turk]

No one intentionally over times an engine I hope. But locking out the timing and setting it at 34 (or whatever) IS overtiming an engine. That grossly overtime’s it at peak torque where it is the MOST vulnerable to detonation and is the basis for this entire thread.

I agree 100% unless something like an induction limited 2v class has a flat VE curve or close to it. Im not sure Thats even possible but it could be.

The closest circle track near me is 2 hours away. I haven’t spent any time going down there to drum up work.

 

[TT5.9mag]

The same way you see it at steady state. It will lose power at that rpm. The difference is you will be testing it in the same manor as it will be used. Read the superflow paper that turk posted if you haven't already.

I did read the superflow paper. The time aspect (cycles) is interesting to me and worth further Investigation for sure. I for one don’t think a sweep test narrows down the timing requirement as effectively as a steady state test. I don’t think you can be as accurate.

 

[Newbomb Turk]

Are you using some sort of box to get a curve like that?

I did that one with weights and springs. It took a bit to shape the weights and get the weight right to do it. Lucking I had a local company make up a bunch of weights for that particular advance mechanism so I can play with a couple of spare weights, get what I want and then get that shape and size on the factory weights.

Then I keep them as pairs with numbers etched on them so I can log them in a notebook. Then when I need a similar curve I can pull them out and start with those.

If someone needs an ignition or is upgrading I tell them to buy a programmable box. It is much easier and quicker to ride the dots than it is to screw with the weights and springs.

 

[TT5.9mag]

I agree 100% unless something like an induction limited 2v class has a flat VE curve or close to it. Im not sure Thats even possible but it could be.

The closest circle track near me is 2 hours away. I haven’t spent any time going down there to drum up work.

I do think when you restrict an engine with very limited induction it narrows up the VE table for sure (for lack of a better description) and probably artificially narrows the peaks, IE brings the peaks closer together.

What would it take to convince you to put a two barrel and a cast iron intake on your engine for some testing? We’d have good back to back data.

 

[Newbomb Turk]

I do think when you restrict an engine with very limited induction it narrows up the VE table for sure (for lack of a better description) and probably artificially narrows the peaks, IE brings the peaks closer together.

What would it take to convince you to put a two barrel and a cast iron intake on your engine for some testing? We’d have good back to back data.

lol I’d have to have a 2v manifold. I really need to head to the circle track next season and see what’s going on down there.

 

[92b]

I did read the superflow paper. The time aspect (cycles) is interesting to me and worth further Investigation for sure. I for one don’t think a sweep test narrows down the timing requirement as effectively as a steady state test. I don’t think you can be as accurate.

Not saying your wrong. Can you tell me why you think that?