Ignition timing control

[mineral co]

Hello Folks,

This is the second topic that I have become very interested in after reading as much as I can stand on the major FSJ sites: upgraded ignition system.

It seems a "no brainer" to upgrade the stock system. Even going only to a Level 2 upgrade is obvious and appropriate (big Motorcraft cap, bucket, rotor, wires, plugs at ~0.045", coupled to Ford TFI coil). However, I really want to also do the MSD for the multi-spark operation, and anticipate a future upgrade to some form of fuel injection. This all means that there will be computors(s) involved somewhere so a goal should be to also get ignition timing control.

One version of the MSD 6a has timing control, accomodating the elimination of the centrifugal advance and, with a MAP sensor, can also eliminate the mechanical vacuum advance. I like this idea a lot!

Then, after researching this topic, I started on an FI education. My first efforts here will probably be a GM TBI DIY system from a donor 5.7 (my ol' girl has the 360). The intent is that the ECU be the 7727 or the 7427, simply because it seems to be very well supported and I don't need to be learning the idiosyncracies of odd-ball ECUs while also learning roll-your-own FI. I believe both these ECUs accomodate spark timing control (if I'm wrong on this, please correct me), but I haven't learned enough yet to know how they do that. All I have seen on that topic is that ECU will modify the timing, but not on how it goes about doing this. Are there parameter "bins", similar to the VE bins, that you configure or does it just run the timing up as far as it can and use the knock sensor to back it off? How does a person configure this?

Obviously, it would be better to have the ECU controlling spark timing if it's done properly. The MSD unit offers a lot of lattitude with configurable timing but having two distinct units that don't know what each other are doing is always less than ideal if one unit can truly handle both tasks, and that's the question: how does the ECU actually handle the spark timing task?

I want to get the MSD unit in the very near future and am willing the shell out the extra bucks to get the timing control version but, at more than twice the price of the regular version, I don't want to get the more expensive unit if the ECU can match its abilities.

Thanks from Mineral County!

Chris

[Stuka]

Honestly, I say go with an HEI and skip the MSD. You can easily use them with a TBI system. They are self contained, and work well. Plus they are cheaper.

CRT Performance has some great kits. Have installed several of them on V8's and 258's.


Sent from my iPhone using Tapatalk

[mineral co]

Thanks for the reply.

I really want to do the multiple spark, for reasons that go beyond just the GW. Does the HEI do multiple spark? I haven't seen that indicated anywhere.

How does the TBI ECU go about controlling spark timing? What user configurable parameters does it allow control over?

[Stuka]

By multiple spark I am assuming you mean on compression and exhaust strokes? Basically what a lot of vehicles do for cleaner emissions. The ignition module that comes with an HEI will not do this.

As for the TBI ECU, I will let somebody more knowledgeable (Blake or somebody) answer that.

[tgreese]

The regular MSD module (ie MSD 6A) has the capacitive discharge capability (multiple sparks) but operates like a "dumb" control module - ie, it is simply set off in synchrony with the distributor's signal.

You can control the 6A just like you would control the spark from a HEI distributor in your TBI conversion. There is nothing special about the HEI distributor, other than it works using factory parts with the GM TBI computer. You can do the same thing with a Motorcraft distributor (or just about any distributor except the Prestolite) and/or a different computer, like Megasquirt. The computer tracks the speed of the engine using the distributor sensor, and using all its inputs, decides how much advance to apply.

Read the section on "Ignition" (spark control) in the Megasquirt manual - http://www.megasquirt.info/ - there are a lot of links to follow, and not all the information will be relevant to what you want to do. Just hang in there, read a lot, and ask questions, and you'll understand what's possible.

hth!

[mineral co]

More reading and more thinking. . .

Conclusions:

1. The HEI does not do "multiple spark". This is not surprising and, from what I can tell, no other OEM does it either.

2. The ECU calculates all the timing itself. The only thing you have control over is the base setting.

From what I can tell, the ECU effectively acts like a sophisticated set of points to the HEI. It isn't just sending a trigger signal to initiate a spark, but is sending a pulse where the leading edge and pulse width are controlled and, inherently, controls the trailing edge which is when the spark is actually generated. This controls both timing and dwell.

This does not automatically classify the timing control as inferior to what could be done with something that a user had full control over. Whether it is inferior or superior would depend on how good a job it does of actually calculating the appropriate timing. One short coming that the system would have is that it has no way to adjust dwell if any coil other than the expected version of the HEI was used as there is no way to tell the ECU what the coil saturation period is. Since, with any less-than-fully-integrated system, this is also impossible to do, it's hard to consider this a glaring fault in the implementation of the HEI. Regardless of coil type, a particular coil's saturation time is a fixed interval if the starting charge is always the same on succesive charging intervals. That being the case, the saturation period can be determined (either mathematically or emperically), and that "period" parameter just hard-coded into the ECU firmware as the maximum dwell interval. This would work well for OEMs that will use the same coil in thousands of systems. Change the coil, though, and you may create a potential problem as the replacement coil's charge period will likely be different.

If the replacement coil's ideal charge period is shorter than what the ECU expects, then you run the risk of over-heating the coil. If the coil has a longer ideal charge period, the coil will not be fully charged when it is time to make the spark and the spark will be weaker than it could otherwise be. With the right equipement (an O'scope with a current probe), a person could easily determine if this has become a problem after doing a replacement, but most folks don't have this equipment and, even if you did, you still wouldn't necessarily know what a more suitable replacement coil might be if you did see a problem.

On the other hand, I would like to think that the big manufactures of aftermarket performance coils have taken a hard look at this and know what sort of coil charge period GM designed for and have designed their replacement coils to take that into account. I would be very surprised if this was not the case. To the end user, though, it means you should always keep a coil with the system it was designed for. For example, if you had the Ford TFI coil on your rig and then switched to a TBI system with GM ECU and HEI for fuel and spark control, you should probably also switch to a coil designed for the version of HEI you are using as there is no reason to think that the Ford TFI coil has the same ideal charge period as what the GM ECU is going to assume. The Ford coil may be perfectly matched, but how do you know without hooking up the 'scope and looking at the coil primary winding's current ramp?

Systems like the Pertronix points replacement units minimize these problems by turning on the current to the coil primary as soon as possible after the previous trigger and then limiting the maximum amount of current. This means you always have at least a farily well charged coil that won't flat burn up but, unless they are actually monitoring the current ramp to see when it starts to saturate, then it is still possible that the system may not fully charge the particular coil. I don't know if their system is that sophisticated or not, but it is irrelavent in this particular case because you can't use the Pertronix with the ECU "timing control" anyway (not without some additional complicated home-brew electronics).

I digress. Back to the issue of timing control. It seems that the user has zero control over dynamic adjustment of spark timing with the GM ECU, the 7747 and 7427 anyway (I would really like to confirm this, absolutely). So what's the likelyhood of the GM implementation being ideal for an engine and application that it was not specifically designed for? Anything I said here would be speculation, at best, as it would greatly depend on just how sophisticated the system really was. However, when it comes to making my own decision as to what to do, my perspective is that the system was no more sophisticated than GM decided it needed to be in order to work well in a specific application that would be repeated thousands in identical fashion. The fact that GM had so many different ECU versions, each targetted at specific models and production periods, supports my attitude that the ECUs were application specific and, therefore, there is no reason to think that a hard-coded timing advance scheme would be so sophisticated as to provide perfect timing when moved to an entirely different application.

Sorry about the long-winded post! In truth, it would have been longer but I lost the second half of it in a "wrong key pressed and the DSL is down (again!) situation.

I'll respond to the other new posts in a separate post.

[mineral co]

Hello Stuka,

I believe the system you are refering to is call "wasted spark", not "multiple spark". Hopefully, in my previous post, I made it a bit clearer as to what I am refering to as multiple spark.

Hello Tim,

The MSD system I am looking at is the "MSD Digital Programmable 6AL-2" (model 6530). This unit allows complete configuration of the centrifugal advance, as well as vacuum advance (with a MAP sensor).

Based on what I have learned about the GM ECU and HEI since I originally posted the question, the HEI control unit plays no part in the decision making process for spark timing. This is all done at the ECU so, as Tim stated, it should be possible to connect an MSD unit to the ECU so it IS possible to have spark control and a whopper ignition system. In this configuration, I could inhibit the timing control performed by this MSD unit without any problem, or use if to do some minor tweaking under certain conditions. Likewise, I could inhibit the spark control of the ECU and operate all timing functions from the MSD though, surely, it would not respond to certain conditions the same way that the ECU would. With a knock sensor, for example.

However, something potentially insidious caught my attention.

If I interpreted everything correctly, the ECU generates a positive pulse to the HEI unit, the spark occuring at the falling edge (end) of the pulse. Using the ECU as a trigger, the MSD unit would have to use the "points" input for the trigger signal. Devices that have a "points" input generally trigger on the rising edge of a pulse so, in this case, it would be triggering on the start of the ECU pulse, not the end! . If this is indeed the case, the spark would be occuring well in advance of where it is supposed to be! Since the pulse from the ECU is indicating the period of the pulse, and not the dwell angle, there is no easy way to compensate this out. It would require the signal from the ECU to be inverted before being passed onto the MSD (or any other device getting its input signal from a "points" input) in order to work correctly.

With no way to anticpate what the timing SHOULD be under any given circumstances (because it isn't user configurable), how would anybody have any reason to suspect a problem of this type? The net effect would be that the timing was either too far retarded at low engine speed if max advance was manually adjusted into the distributor at high speed, or too far advanced at high engine speed if basic timing was manually adjusted into the distributor at idle. Additionally, there could be noticable jitter in the timing as different engine parameters changed.

This will be my new research subject until there is a definitive answer.

[mineral co]

Always learning something new. . .

When using the "points" input, the MSD unit is triggered by the rising edge of a trigger pulse, as would be expected in any true points ignition system.

The GM ECU, with 7 and 8 pin modules, does generate a positive pulse to the module that intends the spark to be at the end of the pulse, the falling edge.

So you can't directly interface the ECU to the MSD without something to invert the signal.

As it turns out, though (and not surprisingly), I'm certainly not the first person to recognize this problem. MSD has a very simple solution (which might have been more obvious if I would take a break from trying to cram all this stuff in my head in too short a period of time). They go ahead and feed the ECU output into the 7 or 8 pin module's timing control input and let the module generate its normal signal to the (no longer attached) coil. The signal to the coil is inverted inside the module so that the coil is driven just as it would be by a set of actual points. There is no coil, but the module doesn't know and doesn't care.

It is the module's coil signal that is used as the points input to the MSD. Sort of the long way around, but I can see how it would work just fine.

So, yes, the MSD will work with the ECU, and either will implement a form of spark timing control. But the question remains: how does the ECU go about implementing timing changes? Without the source to the installed code, it may well not be possible to know this if there are no user configurable parameters controlling it.

This may be a good reason to view an FI project as a two, three, or more, step process. Start with the GM ECU, get it running well, try a couple different approaches to timing control, and then likely transition to MS for the real hard-core stuff which would likely lead to, ultimately, MPFI and then to the Holy Grail: SPI.

Thanks for your posts, guys. It always helps me to better think things through.

[tgreese]

Brevity will get more responses ...

[mineral co]

I'm not fluent in brevity (as you can tell). It's a character flaw.

I'm going to go rebuild the carburetor now.

Oops!

Sorry.

[husker77]

I'm reading his post with considerable interest, never know what you need to know.

[Mr. Goatman]

Wow!! Interesting topic. I like your thought process. I can't help but to think that this would be on level with installing a 24th century nuclear fusion powered warp drive engine on a Mercury rocket. If your interest is in electronic timing control and you would like fuel injection the TBI system has been done many times and works very well. It can be used with stock box, HEI, or MSD ingition if you like. There are always ways to argue theoretical gain, and there is major cool factor in your desire, but I don't think you'll see a performance gain. If you did, it certainly would not offset the price. I would upgrade the ignition, get the MSD box if you want it, and then go with a TBI and spark control system.

[mineral co]

Thanks for the responses guys!

I can't help but to think that this would be on level with installing a 24th century nuclear fusion powered warp drive engine on a Mercury rocket.

That gave me a good laugh! But I'm going to try to not get that carried away.

I, respectfully, have to disagree with the notion that MPFI and SPI don't add much and are used only to get cleaner emissions. Granted, at least where the majority of auto manufacturers are concerned, it is the requirement for cleaner emissions that has forced them to implement these systems, but that is not the extent of what they (can) do. My first argument would be that we now routinely have naturally aspirated cars that make several hundred horsepwer and are capable of getting amazing mileage and are reliable enough that a manufacturer will not only give you a warranty for it but you can expect to drive it well past 100,000 miles without any problems.

This was certainly not the case back in the '80s and TBI was already commonplace then.

Certainly, MPFI and SPI are not solely responsible for these improvements, but they are an important part of the system.

I've also been a bit disturbed by some writings I've seen on various boards that suggest that induction system considerations are less important with FI than with carburetors. Huh? What? The physics of fluid dynamics doesn't change just because you're squirting the fuel in rather than sucking it in. Nothing has changed about how the air gets into the engine. Everything that was important in the induction system with carburetors is just as important with FI.

Without going into all the potentially nausiating history of it, suffice it to say that Bosch mechanical FI systems were being used on production German cars in 1954 (and non-production racing cars a few years earlier). These early systems went even one step beyond SPI and used direct chamber injection, not unlike mechanical diesel injection except that the fuel was not being ignited as it was being injected. The effects were terrific and, with the DCI system, the torque curve projected noticeably higher into the rpm band.

These systems didn't stay in production for very long, the last one I'm aware of went out of production in 1963. This was not an inexpensive system, and in the later half of the fifties the new models of top range cars were introduced with mechanical SPI and the mid-range models had mechanical MPFI. The port injection didn't have the full range of the DCI, but was close, and it was less expensive to build (slightly). Primarily, though, it was more reliable because the injectors weren't sitting in the chamber and living in that harsh environment. By 1965 even the mechanical MPFI was dropped and all FI car models used mechanical SPI. (Note that none of these Bosch production systems were TBI.) There was no enviromental or mileage legislation in place during any of this.

By the late '60s there were many European manufacturers using mechanical SPI. Porsche, Mercedes, Alfa, BMW, Aston Martin being a few examples. These systems were expensive, and these manufacturers wouldn't have put them on their cars if there wasn't a comparable benefit. An example, and the only one I can think of right now where there are published numbers for both the carb'ed and MPFI version of the same engine, is the Mercedes 220S and 220SE models of the early '60s. The 220S had two two-barrel carbs (four barrels of carburetion on a factory 2.2 liter engine!) and the 220SE (the "E" meaning "einspritz", or "fuel injection"). The 220S engine was rated at just over 120 HP and the 220SE engine was rated at 135 HP. Now, that may not sound like much, but that is still a 10% increase in peak measured horse power. Plus the 220SE had somewhat better mileage, as well.

How much would you be willing to spend to add 10% more HP to your engine? And, if you were willing to do it, would you expect to get that increase while also improving your mileage? Bear in mind that this is with MPFI, not SPI. SPI would do even better. Also, bear in mind that this was with a mechanical system, not an electronic one. Just like a distributor with mecanical and vacuum advance, or with the typical carburetor, these are mechanical systems. You cannot tailor them as closely to "ideal" as you can with a fully electronic system, and that is the one reason why everything on modern production engines is electronic.

But, you say, the TBI is electronic so why go to MPFI or SPI? Good question.

What is the one word that comes up repeatedly when discussing fuel systems? Atomization.

Fuel must be atomized before it will burn the way we want in an engine. It seems a little unnatural at first to think that you could put liquid gasoline into an engine and have it not burn, but that is indeed the case. Any gas you introduce into the engine that does not remain atomized until the ignition point just goes out the exhaust, wasted. It is vaporized (atomized) fuel that burns, not liquid fuel.

Atomized fuel will hang in turbulant air for a while. Atomized fuel in relatively still air will start to condense out to liquid again fairly quickly. Temperature has a lot to do with the rate at which this happens. It happens much more quickly at lower temps. Now think about the path that the fuel takes into a cylinder from the carb or TBI injector. It's a relatively long path, especially in systems designed for lower revving, higher torque applications.

Flow through the system is stuttered. While an intake valve is open and a piston is descending the air into the cylinder is flowing and will be turbulant. Then the intake valve shuts and air flow stops (ignoring the effects of momentum for the time being). All of a sudden the air is relatively still and now it is much easier for the atomized fuel to start to condense out into liquid in the intake path. That perfectly adjusted source of atomized fuel is no longer perfect. The fuel that has condensed out is now totally useless to us. Both our potential power and mileage suffers. The system has to be set rich in order to compensate for the lost atomized fuel in the intake tract.

Again, temperature has a LOT to do with how fast the atomized fuel condenses out. We all know that the ideal air/fuel ratio is about 14.7-to-1. If I set up a carb to provide that, why the heck do I have to have a choke to get the engine to run when it is stone cold? You are right! There really is so much gas that condenses out, even in that relatively short distant and time, that the only way to get an ignitable charge is to put ten times as much fuel into the system. And all that extra fuel goes right out the exhaust, wasted. That is why modern systems make such a great effort to warm up the intake tract and engine as quickly as possible. Note that warming them up to "operating" temperature only reduces the problem, it does not eliminate the problem.

Hmm, what to do? What to do? Hey! What if we don't let atomized fuel just sit in the intake tract? What if we introduced the atomized fuel only after the air is already moving into the cylinder? Then substantially less of it can condense out and be lost. We could use that un-lost fuel to make power!

SPI.


Thanks for indulging me,

Chris