Phasing Of Spark And Injection

[steve@cornwall 100]

Where is the optimum place in the compression cycle to introduce fuel if injected solely for that particular cylinder? Was just something I was thinking about as with batch firing, it seems to me that fuel is introduced four times to every spark? Not any particular relevance but interesting.

[welshpug 1,657]

you mean in sequential injection/spark applications?

 

no idea specifically, I'd imagine it varies with engine speed and load, so its quite a complex thing to get right and its a balance of power vs emissions/economy

[Sandy 191]

Batch fire is usually once every two revolutions or on every revolution, often referred to as two stroke mode. Semi sequential should do the same, but with 2+3 cylinders 180 degrees out of phase with 1+4.

 

I've been doing a lot of development in this area over the last few years, which has taken a large amount of time and money, so I'm not going to give away what I've learned in detail; but synchronising injection with induction for each cylinder, has the obvious benefit of giving equal conditions for each cylinder, balancing the way each performs better than batch (batch conversely creates four different fuelling conditions). Spark wise the main benefit is in reduced coil duty at high RPM and losing the softening of spark energy at high cylinder pressure, which in most cases is hard to detect and virtually irrelevant anyway with modern wasted spark ignition systems.

With equality between cylinders and injection angle control, you can manipulate the mixture behaviour quite significantly and the effects can be startling, although the real power benefit mostly negligible; even with engine dyno precision, I've struggled to see any power benefit in most cases, that tends to rely on how well the rest of the package is working IMO, ie if the injector location and pulse tuning are presenting problems, injection timing might mitigate that, but in a well rounded engine, no problem to solve equals no benefit. What I'm not afraid to tell you that I've learned though, is how you can't always rely on the Lambda reading at high load, when you're playing with this; it will deviate from the actual cylinder mixture significantly in certain situations around the sweet spot.

The main benefits I've found to sequential, have been in manipulating the feel of the engine and finding a slightly higher level of driveability, but that's largely irrelevant to low-medium spec builds and for the time and cost involved in throughly testing and mapping to each engine, there are much more productive areas you can focus on.

 

Here's one of my highest spec 2-litres with full sequential and coil on plug, to illustrate how sharp and flexible an extreme engine can be... https://youtu.be/CHa6dgZIbco

And here's a similarly powerful 2 litre Vauxhall engine I mapped on wasted spark, on the same event; to illustrate the difference. https://youtu.be/jFJAIIJz_Ws

Edited December 12, 2016 by Sandy

[petert 586]

I agree that total power difference is negligible, but the difference comes in torque and thus throttle response. By fiddling with the injector phase angle at lower revs you can definitely see torque gains on the dyno. Motec and Autronic have always been sequential fuel only. No option for multi-point, batch or semi-sequential. What does that tell you?

[welshpug 1,657]

that they dont cater for old school low tech engines ?

 

not sure that is correct anyway?

 

the tu5j4 saxo I service for is on motec without a cam sensor thus running batch fire and standard wasted spark coilpack.

[petert 586]

Or Motec make low spec ECU's for the UK market?

 

They probably changed their product to cater for a greater market, as they were definitely race only orientated originally. Haltech have always overed the option.

[Biggles 65]

Hope this is clear enough to read (from some work I did a couple of years ago for an OEM) - top line is torque, x axis is injection timing, 3500 rpm, WOT, MBT spark, fixed lambda, fixed cam timing - gives you an idea of the effect - also might give you an idea about the amount of time involved to investigate it properly given the number of variables involved :

[Sandy 191]

Interesting, is that from TDC firing or overlap?

[petert 586]

Absolutely, and then think about what would happen if you added the variable of either a 250cc or 500cc injector to that data.

[Biggles 65]

Interesting, is that from TDC firing or overlap?

Not quite that simple I'm afraid - the chart is plotted against the variable in the ECU (which is related to the inlet cam timing). On the chart, the dotted vertical line at 192° for the injection parameter equates to an EOI of 421° after TDC firing. (Blame the Germans for making it so complicated.)

[Sandy 191]

Thanks, that gives me an interesting correlation. Obviously race vs production style engines is a big difference, but there is some consistency.

 

I only worry, in the context of the OP, that readers might get hung up on this, when so many other engine spec aspects will give them more benefit for their money.

[steve@cornwall 100]

All very interesting reading. (If a lot is above my head it's certainly still interesting) My initial thought was that fuel/air was being wasted by batch injecting then I realised the exhaust valve wasn't open on the non firing cycle. It became one of those musings whilst walking the dogs in quiet woods so I asked the question. It all led me to wonder how engines that have been around for many years (like my fiat 1.2 8v) have managed to become both spritlier and more efficient over the years, with the most noticeable changes being to the engine management.

I even took a look at the old kugelfischer system as I remember my old BMW going like a rocket (before exploding) when it leaked air into the fuel system.

[petert 586]

It amazes me to see pics of GTi6 engines on here which have been converted to TB's, minus the crank sensor. Given they're street cars, rarely driven at max injector duty cycle, I can't help but think they're missing out on something.

[welshpug 1,657]

its almost always down to budget, £500 gets you a batch fire S40, you need an £800 S60 to go sequential, and for what gain? add to that most mappers say you don't need it.

[Sandy 191]

A lot of mappers don't make a thorough job of batch, so I wouldn't hold out much hope of them making a good job of sequential! I wouldn't be snobbish about batch Peter, a well set up batch fire/wasted spark will be very nearly as good, the vast majority of my work is on batch/wasted, with very happy results and excellent fuel consumption; I'd take the Pepsi challenge with most of the sequential mapping from other I've seen!

As Meirion says, cost is a big factor, paying for the big increase in time to have it mapped through properly and the physical aggravation of adding a cam sensor, reliable trigger and CPPs to engines that don't have it or even with alloy pulleys on the GTi6. There are posters on this thread with batch/wasted engines I've mapped, I'd be interested to know if they'd like to pay the ~£750 difference to have had sequential instead.

[Biggles 65]

Based on the amount of dyno time to do sequential properly (which I'm absolutely sure you do), i'd say £750 is remarkably cheap. We spent a couple of weeks on the dyno doing the work above and it was by no means complete - we learned what we needed to learn in order to pass it on to the guys who'd do the production cal - which is fine if you're going to spread the cost over 500,000 units - not quite so fine for a handful.

Would I have it on my engine (and pay the £750) ? I have absolutely no complaints about my batch fire / wasted spark setup. It's the driver who is the limiting factor not the car so TBH the £750 would be far better spent addressing my deficiencies rather than anything to do with the car.

Edited December 15, 2016 by Biggles

[Sandy 191]

That's the rough additional cost over what you'd pay for the batch fire set up on a new build, which I do at cost basically on my high spec engines, because there are further reasons I prefer the next ECU up on them. To do it retrospectively is more expensive, complete change of ECU, loom and the installation of cam sensor, trigger and coilpack(s). The first one of my race engines that had it, took a week on the engine dyno to map, which I did off my own back as research and thankfully the dyno owner was very sympathetic! The second one took 4 days, but I've done many since quicker, including some extreme bike engines and the last one I did a few weeks ago, only took two days including some exhaust manifold development. Thank goodness for DTA, it would not be possible to test/map so many variations so quickly on the other systems I know well. The guy I was doing the last bike engine with was a Pectel stalwart and was astonished how quickly I was able to refine the maps on DTA and we found about 5% power increase on that one, through twin injector blending and injection phasing alterations, by far the most increase I've seen and as ever, race fuel was worth nothing at all over V-power.

[petert 586]

It's certainly interesting marketing, starting off at the ECU price. For example, an entry level ECU such as the Haltech Elite 550 supports multi-point, batch, semi-sequential and sequential for the same price as an S40. Because most ECU's are sequential here, it's just part of the accepted setup and mapping procedure these days. $1500 would be a very long and thorough dyno session. Definitely top end of scale. Double that for a fully emissions tested, certified and locked ecu. Sure, there are many race cars around running multi-point or batch, and I fully accept that differences in tyres, suspension and driver can fully outweigh any gains made with sequential.

 

I'm not sure when wasted spark entered the discussion but it would be interesting to see the difference between an engine mapped batch/wasted spark and then done again as sequential/cdi.

 

Also, we define batch fire as injectors firing in two even groups. Multi-point being simultaneous firing of all injectors. Are we talking about the same thing?

[GLPoomobile 958]

A lot of mappers don't make a thorough job of batch, so I wouldn't hold out much hope of them making a good job of sequential! I wouldn't be snobbish about batch Peter, a well set up batch fire/wasted spark will be very nearly as good, the vast majority of my work is on batch/wasted, with very happy results and excellent fuel consumption; I'd take the Pepsi challenge with most of the sequential mapping from other I've seen!

As Meirion says, cost is a big factor, paying for the big increase in time to have it mapped through properly and the physical aggravation of adding a cam sensor, reliable trigger and CPPs to engines that don't have it or even with alloy pulleys on the GTi6. There are posters on this thread with batch/wasted engines I've mapped, I'd be interested to know if they'd like to pay the ~£750 difference to have had sequential instead.

 

Over the years that I've read posts on here and followed peoples projects, one thing that has amazed me with the cars that have moved to ITBs and aftermarket management is how unreliable the setups appear to be. Those who have managed to run for a few years without issue seem to be an exception, and it almost seems the norm that owners run in to some sort of issues, whether it be component failure, air leaks, or with the mapping and so on. But it often leads to much head scratching and further refinements being needed. That's the impression I have picked up anyway. It's an impression that has stuck with me as back before ITBs became the defacto choice compared to carbs, one of the "selling points" was that ITBs should be fit and forget by comparison, with carbs needing regular fine tuning. The reason I'm waffling about this is that when I read Sandy's post above it just made me think how a complicated sequential map would have the potential to throw more fuel on the fire for these people who run in to reliability problems! I am of course referring to the type of owners such as members on here who run them as roads cars or track toys, not full blown race cars with extensive engine development.

[welshpug 1,657]

mpi is just a phrase for an engine with more than one injector, i.e not spi, a glorified carb with an injector.

 

batch fire is all injectors firing.

 

semi sequential is alternating pairs.

 

 

its about £300 for a straight non sequential map.

 

it would be cool if there was an ecu at a lower price point that does the job, Haltech certainly isn't a new name to us but by the time it gets here its nearer a grand (£1000) than the £500 an s40 is.

[petert 586]

No, batch fire is two pairs or groups, in order to reduce fuel rail fluctuations. Jetronic is an example of multi-point, not batch.

 

Batch can be done without a cam sensor and shouldn't be confused with semi-sequential.

Edited December 15, 2016 by petert

[petert 586]

it would be cool if there was an ecu at a lower price point that does the job, Haltech certainly isn't a new name to us but by the time it gets here its nearer a grand (£1000) than the £500 an s40 is.

Just ask. I'm a Haltech dealer.

[Sandy 191]

It's certainly interesting marketing, starting off at the ECU price. For example, an entry level ECU such as the Haltech Elite 550 supports multi-point, batch, semi-sequential and sequential for the same price as an S40. Because most ECU's are sequential here, it's just part of the accepted setup and mapping procedure these days. $1500 would be a very long and thorough dyno session. Definitely top end of scale. Double that for a fully emissions tested, certified and locked ecu. Sure, there are many race cars around running multi-point or batch, and I fully accept that differences in tyres, suspension and driver can fully outweigh any gains made with sequential.

 

I'm not sure when wasted spark entered the discussion but it would be interesting to see the difference between an engine mapped batch/wasted spark and then done again as sequential/cdi.

The cost of the ECU alone is only the start and there are cheaper entry level sequential ECUs than the DTA S60, but they are inferior in most cases in other areas, such as spark angle accuracy and resolution (mentioning no names, but a a well known system comes to mind!) Also another system that includes injector angle in its spec list, but operates 90 degree steps 2D! The loom will need to be custom made in most cases at additional cost, the coil packs, the physical cam trigger if non original cams/pulleys are being used, before you consider the extra "dyno" time, by which I mean engine dyno, I would not describe rolling road mapping as accurate and consistent enough to properly evaluate the effects of injector angle 3D. For engine dyno time with a mapper that really knows what they are doing, you're typically looking at £500/day+ and it will need to be trimmed in the car after.

I fully agree that ultimately sequential becomes superior, but only if it's done properly and I think you're sending the wrong message about batch fire, making it sound inferior, when it's barely any different in most cases. I doubt very much Colin Satchell could shave anything off his hill records by going sequential and it certainly runs much more clean and crisp than virtually all the high spec sequential equipped cars I've seen from others.

 

Wasted spark vs CPP goes hand in hand, IMO, because in most cases it will be part of the same choice as batch vs sequential. One of the first high end bike engine jobs I got, was sent to me because a very well known mapper couldn't get it to go through 11000RPM (14000 red line). It had coil per plug, but was running them in pairs because it had been wired incorrectly. The coils were saturating because of the duty involved and spark accuracy/strength was being lost. I identified and sorted the wiring issues, set them to fire cylinder order and it was free to rev out as a result. This is one of many issues, that probably wouldn't ever get noticed on a 205 engine of any spec, but of course the issues of coil duty and path of least resistance for the spark energy exist at the very least at theoretical level, if not measurable. Reliability is surely a benefit of CPP, but I won't forget when Sarty came to me for mapping years ago and one of his LS1 coils signed out before we put it in the rollers and there was no practical way of sourcing another in time to do the job; had it been on a typical Peugeot or Ford wasted spark coilpack, we could grab another in seconds!

[Sandy 191]

GLP- I think the issues you describe are partly about installation and partly about hardware. Installation is responsible for most of the problems I see and I try to convey as much as possible, the issues that need to be addressed to make a reliable ITB set up. The basics such as wiring reliability (avoid adapter looms on old wiring, custom looms should be properly made with adequate wire gauge to be physically durable, terminations solder tipped to resist corrosion, wiring adequately protected and supported, power feeds substantial etc); inlet mounting and sealing, ie avoid thermal insulator plates, use adhesive sealant such as Loctite 5660, ensure o-rings are lubricated and correclty fitted, throttle cables should be nylon lined and carefully fitted, with a positive stop for the pedal, to avoid straining the cable nipples. The problems with hardware run deep though, most aftermarket ITBs have problems with wear and linkage sag, which combined with the slight changes that will occur within the engine, mean an annual or thereabouts check and adjustment of the physical set up and mapping, is advisable.

[petert 586]

... and I think you're sending the wrong message about batch fire, making it sound inferior, when it's barely any different in most cases. I doubt very much Colin Satchell could shave anything off his hill records by going sequential and it certainly runs much more clean and crisp than virtually all the high spec sequential equipped cars I've seen from others.

Parry's 205, one of the fastest FWD cars in Australia, my engine & ecu, runs multi-point/wasted spark. Tuned by Phil Armour and ranks 5th overall on his Under 2L NA wall of fame.