I'm trying to improve the fuel mileage of the new tow pig, and I figure that the timing curves could probably use some work. Engine is a '71 Ford 400, 9:1 compression ratio, 2-barrel carb, all stock.
I've played with recurving a little in the past, but I want to know more.
Are you an expert? Take me to school.
Replace with megasquirt and modern spark control?
Then just tappy-tap-tap on the keyboard and the magic pixies will take care of the rest!
There's also Megajolt as an ignition-only spark control module.
The suggestions of using MS or even just MJ have some really good reasons.
Mostly because they add some very helpful degrees of freedom for tuning the timing.
With a distributor- you have two choices- one to change the advance for engine speed and one to change the advance for vacuum. Both are not all that great at being able to optimize the spark advance- as they are "tuned" to prevent knock at high temps and the vacuum advance is limited by it's capability. AND because of the nature of the beast, both have some natural scatter in their operation that is also compensated for in the original tuning.
When you go to digital spark control, the base timing that a distributor can supply is a good base line to assume worst conditions. So for a sub 100F day, you can add a few degrees of spark as the ambient air cools down. You can also tweak the timing based on engine speed a lot easier (so the sub 2000rpm will be much better). You can also advance the spark a lot more confidently at lower MAP's. Having compensation for engine temp and ambient temp added into the more consistent base timing control is a great thing.
All of that will allow you to pick up considerable thermal efficiency by getting the combustion phasing closer to ideal.
For what you are doing, I'd very much look into trying the MJ set up. It's pretty simple, and should be able to use the original distributor (keep the advances in place, so that it will do a better job pointing at the correct throttle. OR you can get a simple coil pack set up that will work great.
The effort that you will have to put into modifying your distributor to accurately do a better job is going to be a lot- especially given the number of distributor experts that are out there (which are getting less all the time). And comparing that to adding a simple MJ set up.... Well, all I can do is suggest to look into both the above suggestions.
BoxheadCougarTim said:There's also Megajolt as an ignition-only spark control module.
Woah, that's pretty snazzy! And for less than $200, it looks like it would be a half hour project to get going. Neat.
Pretty sure that if doing this with nothing but weights and springs, you would need a distributor curve machine.
I ain't seen one of those in a long time. Also, I have a feeling that if optimizing for fuel, you are going to be mutually exclusive of acceleration.
To those that know better than me about distributor curves, I am guessing that there is a "cam" effect witht he weights that can be put into play as the weights move or is it a straight linear change based on RPM and vacuum? A cam would open a lot more options for designing an advance curve.
In reply to NOHOME :
There is some tuning with the weights and springs for the WOT engine speed spark curve, no doubt. There's also mechanism range, too, that's part of the design.
The vacuum advance, I would suspect, is more linear change vs. manifold vacuum.
But the general mechanisms are separated from each other, which is why one needs to plug off the vacuum advance when setting the base timing.
And some cars don't even have vacuum advance (like my Alfas). I'm sure I could gain a BUNCH of FE if I had that in my Alfa.
Saying spark accuracy and distributor in the same sentence is a mouthful.
Oxymoron even, like military intelligence.
At best a distributor is the tail wagging the dog.
You want a perfect identical spark for each cylinder that fires.
A crank trigger does this best then you can run Coil On Plug and have no moving parts and full control.
Nothing to wear out, higher voltage because the coils have more time to rest, and timing accuracy that no distributor can equal.
Spark accuracy is vital when you are dancing with detonation, especially when detonation feeds on itself.
Systemic spark variation compromises your tune and eats up any safety room you leave and sweeps usable power onto the floor.
+1 for MJ/MS. This will give you vastly better control over timing AND better reliability than a distributor. You can also add a spark-cut limiter if you don't already have a rev limiter - it's better than no limiter and it's quite a spectacle 
NOHOME said:Pretty sure that if doing this with nothing but weights and springs, you would need a distributor curve machine.
I ain't seen one of those in a long time. Also, I have a feeling that if optimizing for fuel, you are going to be mutually exclusive of acceleration.
To those that know better than me about distributor curves, I am guessing that there is a "cam" effect witht he weights that can be put into play as the weights move or is it a straight linear change based on RPM and vacuum? A cam would open a lot more options for designing an advance curve.
I really want a distributor machine. Maybe sometime I will stumble across one or make one myself. I think it would be an ideal 'retirement business' in about 30 years when the only people still running and needing distributors setup are the ones that really really want them and therefore the services are needed. The distributor has always been one of my favorite mechanical engine control items.
That said, I'm pretty sure there are cam profiles you can mess with, to an extent. So for example if you wanted the first 10 chunks of extra 100 RPM to add 1 degree timing each and you wanted the next ten 100 rpm chunks to add 2 degree each, it could be done. but the more extreme the "cam" gets the larger it needs to be or the more chance of mechanically binding up.
I think the mechanical advance only is probably not too hard to understand and also pretty consistent if everything is lubed appropriately. Vacuum advance however, opens up an issue because engine vacuum fluctuates wildly based on like 100 different factors, and it CAN wag the dog (vacuum increases, spark advances, engine speed increases, vacuum decreases, spark retards, engine speed decreases, vacuum increases, etc), so there also needs to be a damper in the system.
The wagging I refer to is the daisy chain of mechanical mayhem that the distributor is at the end of.
Plus the wobbly ass distributor shaft itself.
Even if a distributor didn't wobble there is not much resolution found in a 1" diameter wheel.
The 6 to 8 or more inches of diameter found in a crank trigger offers accuracy in multiple orders of magnitude better than a distributor.
NO DISTRIBUTOR can ever come close to the accuracy of a crank trigger. NONE!
The only reason to leave a distributor in a motor is to drive the oil pump.....
I converted a mechanical FI motor to carbed operation and used one of those "modules" for ignition control. It was a great solution and I would do it again.
bentwrench said:The wagging I refer to is the daisy chain of mechanical mayhem that the distributor is at the end of.
Plus the wobbly ass distributor shaft itself.
Even if a distributor didn't wobble there is not much resolution found in a 1" diameter wheel.
The 6 to 8 or more inches of diameter found in a crank trigger offers accuracy in multiple orders of magnitude better than a distributor.
NO DISTRIBUTOR can ever come close to the accuracy of a crank trigger. NONE!
The only reason to leave a distributor in a motor is to drive the oil pump.....
Yeah, I get that there is slop in any mechanical system, and that slop can be a problem too. Keep in mind I am not arguing that distributors are better than electronic crank fired ignition, just that I have always liked distributors.
Not sure I understand your claim about the diameter of the trigger (distributor or sensor wheel) making a difference in resolution. If the diameter is smaller, sure, you have less circumference (bad), but the outer edge also moves slower (good). Assuming both turn the same RPM, it's a 1-1 trade off, every extra inch in circumference is an extra inch of distance your sensor has to cover in the same time. So sure there would be realistic limits that you wouldn't want microscopic trigger elements nor would you want a 16 ft diameter trigger wheel, but you should be able to work with anything in between. If instead you are just saying that a mechanical device will not be as accurate as an electronic one, then I agree with you.
https://www.msdperformance.com/products/ignitions/street_and_strip/parts/6530
With this you lock your dizzy at the max advance then dial in how much retard you want for a given RPM.
If you get one I have a laptop with the software on it that you can have. I would totally have one of these if I had a carburetted vehicle.
I have nothing to teach about timing curves, but you say its a stock 1971 engine, so I assume it has points. Every vehicle I have had with points I have replaced the distributor with an ebay HEI distributor, and have always had great results. Better mileage, more power, smoother running, and more rpms at the top end. They eliminate the maintenance needs of points and use the same electronic bits that gm used for about 40 years so parts are easily available. They can usually be had for less than $100 new. There are also aftermarket kits out there to replace your points with electronic ignition while keeping your original distributor and cap. I haven't personally used these except on tractors but they worked well there.
In reply to gearheadmb :
I have already replaced the points with an Accel electronic trigger, which is identical to the Mallory/MSD piece (the instructions are the same, word for word), so the points are gone.
In reply to everyone:
I'm a little surprised that there hasn't been a full-on distributor expert who popped up and immediately delved into a dissertation on spring rates and tension, slot widths, and vacuum advance adjustment. I'm beginning to wonder if this question might get more traction on the Classic forum.
Thanks for the help, though. I'll look into/consider the full electronic options.
I have had some time with a distributor dyno (Sun brand, cont remember the model) and got to do a little work with custom curving, but not enough to be considered an expert by any stretch. I also built a static advance test rig so that I could tweak the cams and weights to get different total advance before playing with spring combinations to tune how fast the advance came in. This was all for Alfa four cylinder engines.
I typed a nice huge response to this and my browser crashed.
I have some suggestions
More later
snailmont5oh said:I'm a little surprised that there hasn't been a full-on distributor expert who popped up and immediately delved into a dissertation on spring rates and tension, slot widths, and vacuum advance adjustment. I'm beginning to wonder if this question might get more traction on the Classic forum.
Thanks for the help, though. I'll look into/consider the full electronic options.
because its better to go to a programmable ignition solution rather than compromising the ignition curve for efficiency and potentially reducing the power output.
You can alter the curve with different weight springs and different weights, ideally you'd have a distributor machine to do this, realistically you COULD do this on the road with enough sensors and monitoring tools installed with a passenger to watch. At that point, you'd likely have spent more money than a programmable ignition solution would have cost.
Sometimes these older skills die off, because they aren't needed, not because people don't care enough to learn them.
Curtis said:I typed a nice huge response to this and my browser crashed.
I have some suggestions
More later
I look forward to later, when I get to see more. :)
You've already taken the best first step by ditching the points. Did you also remove the resistor wire and run new 10ga wire to the distributor? Most points systems drop the voltage down to 9v or so. They do this to improve the reliability of the points contacts, so when you switch to an electronic puck, its always wise to give it full voltage.
While I agree that fancy crank triggers are more accurate, you're not trying to re-invent the wheel.
Two possibilities. An MSD ignition system will help, but MSD (and other similar boxes) are held back by liability and CARB/EPA regulations to some extent. They can't have you put a box in and put a hole in a piston. Having said that, they are a great upgrade.
Other possibility is to just tune it yourself with good old fashioned analog mechanics.
Take whatever you know about timing curves and toss it. There is no such thing as "best." Phrases like "initial timing" need to be tossed out the window. Initial timing is a benchmark. It is only a reference point based on the distributor's other activity in the curve. It's not that your car needs 10 degrees of initial timing to work, its just that setting at 10 degrees means that the mechanical and vacuum advance curves will yield the correct advance during operation on the road. The only reason for setting timing while at idle is because its a little tougher to do at 65 mph with the hood open.
What you want to shoot for is an idle timing of whatever works well, then tailor the curve to suit the rest. You'll want a mechanical advance that starts coming in around 2800 rpm, so getting a spring/weight kit and following the directions you should be able to get darn close
You should get an adjustable vacuum canister. If you don't, get one. Chances are, your canister is currently hooked up to manifold vacuum. Put it on ported instead. Old school distributors used manifold vacuum to reduce advance during cranking, but then advance it after starting for idle quality. This was important with old-school closed-chamber heads and 10:1 compression, but not necessary anymore. Once you get past about 5% throttle, the vacuum signal is the same anyway.
Most distributors give about 24 degrees of mechanical advance. You can fine tune this on your own, but I would bet a few bills that your best total mechanical advance will be around 34-36 degrees. That suggests you set initial at 12 and call it good. But not so fast. Now you should play around with detonation thresholds at cruise and part throttle. Disconnect the vacuum line. Mash the throttle (do a little drag run) and listen for detonation in the 2000-2800 range.
If you're getting detonation in the 2000-2800 range (before mechanical comes in) when set at 12 initial, you have to back off the initial to 10 and find some way to increase the mechanical advance range to get 26 degrees of total mechanical (to maintain your optimum 36 overall)
Conversely, if you don't have detonation there, you could be leaving efficiency on the table. So you advance it to 14 initial to try... no pings. So you advance it to 16 initial and get pings. So you back it off to 14 initial... but now you would have to limit the mechanical advance to only allow 22 degrees to keep the total at 36. Above all, trust yourself. You might find that 18 or 20 initial offers no detonation at WOT in the 2000-2800 range. Trust it. Your goal is to find detonation at WOT and then back off a bit. (a little side note: During these tests, its best not to exceed about 3500 rpms if you're more than about 14 degrees intial or 38 degrees total. If the mechanical advance comes in and you're suddenly doing 5000 rpms with 40 degrees of advance, bad things can happen)
Fortunately most Ford dizzys let you change total advance pretty easily. There are little roll pins in the advance plate with holes marked with something like 12L and 13L. These are advance numbers in cam degrees. If the pins are in the 12L holes you'll get 24 degrees of mechanical advance. Don't mix and match. Make sure the pins are in the same holes. The way the geometry works out, using one 12 and one 13 will not give you 25 degrees, it will give you a strange plateau in the curve.
Now, here is where the vacuum part comes in. Once you add ported vacuum, the canister will add advance proportional to how little you use your right foot. This increases advance during part throttle cruise. Put the vacuum hose back on and do a similar test. Its best to test this by cruising several roads at various speeds. From a steady speed, roll on the throttle quickly. Don't floor it, just enough to not make it downshift (if it's an automatic). You shouldn't hear any detonation. If you don't, adjust the screw in the canister for more advance by 1/4 turn increments until you do get detonation. Once you do, back it off by 1/2 turn.
This is all old-school stuff and it takes an interesting amount of Zen to do it, but it will net you a darn near perfect curve that is reliable and simple. Its much like getting a carb tuned just right by knowing how the primary jets affect the secondary mix, and accelerator pump lever adjustments help with bigger cams.
Keep in mind this all should be done while the engine is fully hot, and you might expect to have to adjust that canister back a bit in the summer until you find the true detonation threshold.
I also must insist that you are very aware of idle throttle. I wish I had a nickel for every time someone had the idle too high while setting timing. You set the idle with a vacuum gauge on a ported source. Once you have no vacuum, it's set low enough. Then you can set initial timing. Often times shadetree mechanics will do it by ear which almost always leads to retarded opening up the throttle screw first and uncovering the vacuum port. This usually stems from concerns about advancing the timing too far. You kinda can't.
I built a 454 with too much cam for the compression. I ended up actually locking out the mechanical and running it at 34 degrees initial with ported vacuum. In the summer it liked to kick back a bit during cranking and I did have a wee bit of detonation until about 3000 RPMs, but it got me through until I fixed the larger issues. With the ported vacuum on it, I figure I had as much as 54 degrees advance during cruise. This is an extreme example, but I just mention it to show you that each combo is different and if I had stuck with a normal stock ignition curve I would have had all kinds of trouble.
Also be aware that detonation is not always audible. I strongly recommend going back two steps after detonation is heard. Like above I mentioned with the adjustable can: Go in 1/4 turn increments until you get detonation, then go a 1/2 turn back.
Thanks for the info, Curtis.
I've gotta see if I can find that Dr. Jacobs book, too. I think I have it lying around somewhere.
You'll need to log in to post.
