One for the techies.... (This is in no way an advert )

Been thinking about this for a while and havent had a chance to try it myself.

I think I have come up with an idea for a new control algorithm that can be incorporated into an ecu to help reduce turbo spool up and increase throttle response.

This can only be added by an ecu manufacturer as it will require software modifications to the control program.

Normally, ignition is mapped with suitable advance to give maximum power within safe limits.
However, doing this takes energy away from the exhaust turbine as the gas expansion happens sooner and is used mainly to push the piston down.
Mappers will often reduce the ignition advance in the spool up areas of the map to a comfortable average point where the there is a split between gas energy used for both.
I.E. This reduces pushing power on the piston but increases the gas expansion into the exhaust turbine thus spooling up the turbo quicker.

Anyway, at these points in the map, these values are fixed.
I.E. If the revs and boost (load) do not change then the ignition value does not change.
(I exclude closed loop knock advance out of this but you would have disabled this anyway at these points )

So on to my new algorithm idea ...

The method is only enabled when accerating and in the spool up/low boost areas.

The algorithm basically alternates in various sequences a number of different ignition values for 1 mapped ignition point.
Lets say 3 for example.

The first is the normal ignition value originally used. (I.E. the compromised value)
The second is an advanced value optimised for maximum piston pushing !
The third is retarded value for maximum gas flow out the exhaust port.

You cant normally use the second or third values above as either the turbo will not spool up very fast or as in the latter case, prolonged use will make the exhaust valve melt or cause major detonation issues.

The values outside the normal are could be calculated from the original after some experience.

The alternating ratio used and the duration of it depends on a master calculation and is done very fast and perhaps changed every ignition in some circumstances.

This calculation takes time, previous history, current load demand, BSFC to determine the heat load passed through to the turbo and in the cylinder.
EGT and knock sensor could be used to trim this.
This isnt a pure calculation based on current events.

FYI, I have done work like this before on controlling 3 phase motors and knowing how hard you can drive them by back calculating heat load and estimated time to failure/overload for variable speed drives (VSD's).

My basic idea relies on the fact that an engine can actually tolerate DET and intermittent ignition advance/retards outside normal limits for short durations.
E.G.If you heat up the exhaust valve you must allow it to cool
But all this happens very quickly !
Once you can quantify this in real terms as a heat/energy load history I think there are gains to have.
This would need to be done for both the cylinder and turbo.

I hear people screaming that DET is bad for an engine....Not always true.
Nearly all modern cars deliberately run on and over the detonation limit in order for closed loop ignition advance to work !!

Now add to all that fuel control too, it gets comlicated.

I have written a C program modelling a YB engine based on all my mapping experiences and my conclusion are that there are gains to be made.

Unfortunately, there is only one one to find out for sure and how much

The heat load calculation is way too complex for a L8 (sorry Karl/Stu )

Something to think about.....My head hurts now.

 

IMO, once you start putting the extra heat in, you won't be able to get rid quickly enough and it will snowball.

Or, it will be forced to return to the compromise values and you won't have any useful gain.

 

Steve,

You are right to a point, but in terms of engine cycles, you accelerate hard for shorter periods than you would do holding on to max power.

The calculation would know this and would limit its use depending on the energy dissipation used for the piston/turbo in relative history.

 

Si, your a clever clever chappy !!

jb

 

Simon, my own idea along similar lines is "transient timing" where currnetly you have enrichment based on throttle angle change, also have retard and quite a lot of it, would be great for spool up IMHO.

done this way its so simple that I dont know why its not in aftermarket ECU's already, but I guess there must be a reason that I have overlooked!

 

LOL Thanks


A good idea chip.
I assume ths would be time base controlled to keep it simple like you said.

Although I "think" one of the yank ecu's does something similiar to this already

I was trying to be ultra clever and let the ecu work it all out without mapping. Just for the dumb mappers

 

Yes I would have it time based, with a start value and a decay over X seconds to zero.

so for example 15 degrees of retard and drop back over 2 seconds


That would be an option in the map, same as accel enrich is currently



Devilishly simple and potentially quite effective dont you think?

 

PS

We (my mates company, I just do spec work for him as a hobby) have an external device that can do this, but its not looking financially viable.

 

Cool !

Sadly, I have many good ideas like that but thats the always the problem.

Unfortunately, everyone expects you to be brilliant in all aspects of business, as I have found out the hard way LOL
It takes money to make money ..D'oh !

I will stick to designing ecu's for waste gas powered generators ... shit is good LOL.

 

We've got the hardware all working, and the software done to a certain extent, the problem is that to make the hardware is hundred of hours of soldering, and to production-ise it is 100K or so upfront for the first batch.

Not sure if that will happen or not, im just enjoying playing with it at the moment (although my T34 didnt enjoy 40+ psi of boost at a standstill, LOL)

 

PS

And yes before you ask, it does have a boost override in my design so that if you reach boost before the 1.5 seconds or whatever decay time, it cuts back to the normal map instantly

Ours even works with L8 and every other stanard ecu

 

Sounds like fun.

You could productionise that for far less than you think !

It took about 4k for my ecu, and thats not mates rates !

Of course, it all depends on your target price !

Can put you in touch with someone ...

 

http://www.cmd.bristol-uk.com/index.html

We used these aaaages ago...

 

this post needs Stu!

 

chip a t6 can do transient ignition!

 

I forgot about that one.

Its more like chips explanation... Mine is quite unique LOL

 

im just not sure about how much of a gain is to be had in your idea simon!

 

That price is for the first 500 units, no point doing it unless doing it in a big way.

 

Yes, I thought T6 and the Life stuff might both be able to, but can it do a transient ONLY up till a boost condition is met?

Would be interesting to play with it then and see what gains you can make if any

The DOWNside to it of course, is that during the lag, it will be SLOWER than before, so although the lag my last less time, its possibly still not enough of a benefit to make it worthwhile.

 

i cant remember the conditions to the ignition correction and its not on this pc,, ill try and install it again and see what it is!

 

Ive got the T6 software somewhere, I really should just have a look

 

.....plenty other tuners use this site aswell.........

 

True, but Stu is without a doubt the best at explaining things in terms even stupid muppets like me and Porkie can understand

 

..........still makes NO sense to though.....derr.....

 

Sensible input from anyone is welcome

 

Gareth, my virtual simulation says around 3% on a T4 500 bhp spec car which is the model I am using.
Parameters set on the cautious side.


Not much I know...but it may not work in the real world or it could be the next big thing PMSL
As an example, dont forget in F1, milliseconds matter !

A few years ago a race was won with a difference of only one full load cossie injection pulse width (20 ms) LOL

 

....damn.........i was going to talk about the size of pigs testicles.......nevermind......

 

PMSL @ Pete