[Jason Pittenger]

To an extent, all engines will have an increase in idle rpm with an advance in timing, but that won't be an indicator of correct tune.

I have been reading up on ignition timing quite a bit. I'm no expert, so here's my understanding. Please correct me if I'm wrong, because I want to do it right!

From what I understand, the best timing at any given RPM (assuming no knock) is the one that gives the most torque. I saw a video that gave a rough graph of torque vs timing. When well retarded from optimal, the graph is pretty steep and each degree adds a lot of torque. As you approach optimal, the graph levels off and then starts to drop when you pass optimal.

I'm not on a dyno yet, so "highest torque" at idle is highest RPM for a given idle setting. Same tuning process as setting idle mixture. One could also say, highest vacuum at a given idle RPM.

It will be next to impossible to tune partial and full throttle openings below 2-2.5k without lugging the engine, so I'm planning to revert to stock timing once the throttle cracked, up until 2-2.5k. Hopefully it won't make the throttle response weird.

 

[Spockster]

Your last paragraph is a good plan, since less advance is needed at low rpm and then should follow an upward curve increasing with rpm.

I keep forgetting you have adjustability over the entire rpm range. Unlike mechanical systems that are somewhat a compromise between best and worst.

 

[Jason Pittenger]

I'm still learning more about timing, and hoping to share what I've found both from research and experience.

According to this page:
“If the spark occurs too soon, the engine will fight against the pressure in the cylinder creating extra heat and possibly cause a pinging problem that may do engine damage. If the spark occurs too late, the maximum cylinder pressure will occur too late, thus not converting the fuel energy into power, but wasting it as heat that the cooling system must get rid of.”

It sounds like my advance to 8 degrees at idle with no throttle will reduce the engine heating at idle. Additionally adding some advance at cruise and partial throttle acceleration should also reduce heating. I have also considered retarding the timing during a cold start to try to warm up the engine faster. Apparently some OEMs do it with cars, though their main goal is heating up the catalytic converter. I don't have a temperature input to the controller, but I could look at the vacuum / RPMs associated with cold start on choke. I don't have the cold start data log in front of me right now, but I would expect high vacuum and an RPM 2000-2500, as that's where my bike idles with choke.

I keep forgetting you have adjustability over the entire rpm range. Unlike mechanical systems that are somewhat a compromise between best and worst.

This controller is almost infinitely adjustable! On the left column, I can select 10 different vacuum pressures (105 kPa = no vacuum = 0 inHg, 0 kPa = full vacuum = 30 inHg), and 15 different RPMs for a total of 150 different points of adjustments in timing! It will interpolate between them. Not only that, but I can choose what the 15 different RPM points are and what the 10 different vacuum points are. Right now, my RPM points are every 600, but I could set each column to whatever I wanted. I have two individual tables to work with as well, so I set up a 2nd table with only three columns (5 @ 1100, 25 @ 3500 and 25 @ 8500) for stock timing. I'll be able to switch over on the fly (even while riding) at the flick of a switch, to be able to compare to stock timing whenever I want.

 

[Spockster]

I've had the wrong springs on mechanical advance and the engine would just start bucking when you tried to step on the gas. The advance curve was too steep.

On the dirt cars we were stuck with either a locked down advance or playing with springs and initial settings. Using a locked advance meant using a special procedure for hot starts, or getting a push truck.

It's really trick to have full control over the spark curve. Will be interesting to see dyno results.

 

[Jason Pittenger]

I uploaded the wrong log files in post #19.
Here are three log files that I made on Saturday.
Proper file extension is .xlsx or .csv depending on the name.
The .csv file has the raw data and no graph.

I graphed the RPM, the carb vacuum, the ideal ignition timing (stock 1100 RPM and lower = 5 degrees, 3500 RPM and higher = 25 degrees), and what I think is the actual reported ignition timing.

The timing data, I calculated as:
Timing = 36 - (DATA / 2.56)

This formula doesn't really make much sense, but the graph using that formula lines up pretty close to the ideal timing numbers.

On the other hand, this same column of data I had previously calculated as dwell time (before my goofy calculation above).
That would make more sense, as this section of data had an idle advance of 8 degrees, and a decel advance of -5.

I'm still decoding this data....sigh

 

[Jason Pittenger]

I found the reported advance!
Data bytes 105 and 106 are the advance for each cylinder.
Blue is hand calculated advance (based on the reported RPM), and Yellow/Gray are the reported advance values from the controller.
It only reports back integers, but it otherwise perfectly follows what I expect to see.

I assume 105 is cylinder 1 and 106 is cylinder 2, but it really doesn't matter since I'm running the same timing on both.

I updated the logging program file in post #16 to version 1.3, so that it now logs the actual advance used.

I have a bet going regarding how ignition timing affects engine heat at idle. I say more advance, coupled with less throttle to get the same RPM, will result in lower engine temperatures. My coworker say advancing the timing will add heat. Since I'm able, I'm going to run 1100 RPM with both at stock (5 deg) timing and compare cylinder and exhaust temperatures. Then I'm going to run one at 7 and one at 4 and compare cylinder and exhaust temperatures, then swap them and compare cylinder and exhaust temperatures again. Loser has to buy the winner lunch!

I think that more advance with less throttle so that the power output is the same, results in less heat in the engine and exhaust. I think he's thinking that more advance with the same throttle means more power, so more heat.

I mean, what else am I going to do while I'm waiting for spring?!?! I'm at least going to wait until the temperature is > 20F!!! We're supposed to hit -25F this Sunday.

 

[Jason Pittenger]

I collected more data today.
This was just a smooth run up in RPM, with attempts to hold at my test RPMs.
3000,3500,4100,4800,5500,6200,6600,7000,7500,8000 and 8500.
These RPMs correspond with 40,50,60,70,80,90 MPH, then 6600,7000,7500,8000,8500 fill out the rest.

The red is the RPM and the gray is the vacuum inside the carbs.
The X axis is time in seconds.
Higher is less vacuum, with 0 being a perfect vacuum and 100 being no vacuum.
Ignore between 1100 and 3000 RPM. I was experimenting with retarding the timing in that range.

I noticed that from 3000 RPM to 4000 RPM, there is a lot of vacuum in the carbs. IE it doesn't take much throttle opening to hold those RPMs.
At 3000, 3500 and 4100, I measured 66.9, 69.7 and 68.2 respectively.
Above 4100, the amount of vacuum decreases like a straight line until ~5300, where the average remains fairly constant until redline.

Know what happens above 4100? The ignition is no longer advancing, so I see room for improvement at part-throttle operation!

Stock timing stops advancing above 3500 RPM. Mine is supposed to, but it was set to follow stock, except at vacuum readings below 67, and I had some dips to 65.
I plan to bump the timing at each of my test spots and see if I can hold each RPM with more vacuum (lower number). That would indicate more efficiency. I'll hold WOT timing at stock until I can carefully test it on a dyno.

Now, the reason for the weird timing below 3000 RPM. I was hoping to find the operating point for when the bike was running on choke and retard the timing. This would bring down the RPMs while on choke, and IN THEORY, add more heat to the engine to speed up engine warm up.
In practice, adding choke puts it in the 2000 RPM and 85 kPa range. This is the same range that you pass through as you blip the throttle to come off idle. The end result was TERRIBLE off-idle response. You can see in the graph, the engine struggles to rise above 2000 RPM as I gradually added throttle, then shoots up to 3000 RPM where I had to remove throttle to hold 3000 RPM. It would be un-ridable!

 

[Spockster]

Interesting data. Can't wait to see a dyno run. There's one close to me, they run all night sometimes.

On the last graph, what is Advance 1 and Advance 2? Two profiles in the new ign box or is one of those stock?

Seems like a hard pull in neutral. 50 seconds at the upper rpm? Am I reading it right?

 

[Jason Pittenger]

The graph is RPM (blue) vs carb vacuum (gray). I did a slow ramp up of RPM (ie just enough throttle to increase the RPM) to get a baseline for vacuum readings with stock timing. So it's not 50 seconds at high RPM.

Advance 1 and Advance 2 are the individual cylinders. The controller allows you to set one cylinder different from the other. Not an entire timing map per cylinder, just a possibility of setting one cylinder 1 or more degrees ahead or behind the other. The data coming back doesn't show less than whole degrees, so if the actual timing is right at a half way point, one might round down while the other rounds up. Otherwise, they match each other.

I did a few more of those today, bumping the timing up each time. I found I was able to give it 2-4 degrees advance with a noticeable increase in vacuum. I think at higher RPMs, there is a resonant effect, causing the sparatic readings.

I also logged a couple rides, including a WOT pull in 3rd from 2k to redline, and a closed throttle deceleration back down to 2k. I don't have the data with me at the moment, but it looks like there is a slight restriction at the top end at WOT. It starts to pull a vacuum even at WOT. This suggests an ear shave might help top end HP.

I tuned the idle timing at 7, with timing of 8 at 900 and 1 at 1300. It's a trick I learned to stabilize the idle. As idle drops, the timing advances pushing idle back up. As idle rises, the timing retards, pulling idle back down. It's really cool, on warm start it pops up to about 1500 and drops right back down to 1100 on the nose. Same with coming off throttle. RPMs drop like a rock and stop at 1100 like it's a brick wall! Cracking the throttle puts it at a different vacuum level, so it bypasses the retarded timing at 1300. This also prevents the need to adjust the idle as the bike warms up.

Speaking of walls, the rev limiter is a nice feature! I missed a shift and hit neutral instead. This time however, the engine didn't wind out to 13k but held right at 8700.

Lastly, at high vacuum (deceleration), I pulled the timing to -5 degrees for a late burn. It's a nice predictable double pop coming off the throttle, with more pops as you slow down. To each their own.

 

[Jason Pittenger]

Here's part of my ride yesterday. I'm blipping the throttle at idle, then accelerate to a 60 mhp cruise, with a brief 4th gear acceleration before slowing down for a red light.

And here's a WOT run in 3rd gear. Notice the droop in the red as RPM goes above 7k?

This is my current timing map. Throttle is the Y axis. It idles around 68 kPa. Decel is below 50 kPa. WOT is ~99 kPa at my elevation. At 97+ kPa, I'm running stock timing

 

[Jason Pittenger]

I added a switch to toggle between stock timing and my custom timing.
I did make a change to "stock" timing where it runs 7 degrees at 1100-1300.
True stock is 5 at 1100 and 7 at 1300. I did this so I don't lose 2 degrees of timing at idle if I switch back to stock.
Here's the idle recovery with "stock" timing (left) and my timing (right).
Notice with the "stock" timing, the RPM slowly settles in after reaching 1500.
With my timing, it drops straight to 1200 before settling in.
The green is the actual timing, and I'm using it to regulate the idle.

I also added a shift light that starts blinking at 8000 (blink rate increasing with RPM) and it's full on at 8500.
It's zip-tied to my clutch cable.
It's more of a novelty at least during day time. It's not quite bright enough.

Here's a video demo of the idle recovery, rev limit, and shift light.
Switch up is "stock" timing, down is my timing.

You can also hear the decel pops as I back off the throttle.
I rode through a city street with brick buildings on both sides.....and WOW it's loud!
Another reason I'm glad I added the "stock" timing switch. I can turn off the pops whenever I want.

Lastly, my tach is slightly inaccurate.
1100 RPM shows up just under the "1" mark on my tach, and 8500 shows up almost at the "9" mark.
Notice too, when the rev limiter kicks in, the tach goes crazy.

The final step is do some typical riding and compare MPGs with new and old timing. I'm running 1-2 degrees over stock at cruise, so it should have an effect.

 

[Spockster]

Hey, I know how to make it go off like a howitzer if you're interested. It's only single shot though.

Waiting to see ride results, will be interesting.

 

[Jason Pittenger]

As I'm transitioning from tuning to just enjoying, here's some of my initial observations.

I rode to-from a friends house last night. Bike sat on the street for a couple of hours, so it was plenty cooled off. Temperature was low to mid 40s.
I started it with choke, and gently rode off. After going about a mile, I took the choke off and it idled right at 1100.
Earlier in the day, I rode for a bit when the temperature was warmer. Even at the end of the ride, it still held 1100.
I'm loving the locked in idle! I never have to touch the idle speed, or deal with an idle that too high/too low.
It drops into idle almost instantly, and it recovers just as quickly when a slight release of the clutch drags it down.
The rock steady idle is just a joy!

On my right mirror stalk, I have a switch to toggle over to the stock timing (eliminates decel pops). I flipped this switch last night as I was in a neighborhood.
I love the ability to turn the decel pops down at the flip of a switch.
I do lose the improved idle quality with this, but I guess I can't have it all!

I'm still working on some cold start adjustments. I set up another switch on the left mirror stalk that retards the ignition.
I tried 30 degrees, and and the bike dies on choke. I tried 5 degrees, and the RPM only drops from 2500 to about 2300.
Somewhere between 5 and 30 is the optimal amount to get the bike to idle lower on choke.
The goal is a quieter cold start through a lower idle speed on choke.

I just filled up this morning, and I got 28.45 MPG! But it's not a meaningful measure.
I spent a good amount of this tank just running the bike in my driveway, WOT pulls, and draining the carbs.
Not at all what I'd consider typical riding.
The experiment of MPG starts now. This tank, I'm going to ride with my new timing. Next tank, I'll ride with stock timing.
Of course, I don't know how much temperature will play a role in MPG, so I might have to repeat the experiment in the summer.
For now, I'm filling up with premium, just to be safe and avoid any knock.

 

[Jason Pittenger]

I set up the switchable ignition retard to 10 degrees. Now a cold start is about 1300-1400 RPM on full choke and the timing retard. I always hated rolling through the neighborhood with throttle closed and it would hold speed. I also hated how loud it was when first warming up. I'm LOVING how it behaves when cold now!

The retarded timing reduces power, so it's a good reminder to take it easy while it's cold. It also puts more heat into the exhaust (and in theory, the engine) warming it up faster.

I'm 75 miles into the tank now. I hope to be onto the next tank by the weekend.

 

[Spockster]

Turning the choke off early helps warmup quicker, since it will be running leaner.

 

[Jason Pittenger]

Today was officially the coldest ride I've ever gone on (30F)!
But it gave me a chance to try out a really cold start with my new timing adjustments.
She started up with no hesitation, full choke and just a small blip of the throttle.
She settled into a ~1300 RPM idle with no drama.

 

[Jason Pittenger]

MPG results are in. 42.57 with my timing and 41.97 with stock timing. That's a small enough difference to be due to lean angle while filling the tank, or idling at a light just a bit longer, or even ambient temperature. At least it didn't decrease!

I'm still waiting on an anti-pulse valve shipping from UK to do more tuning. It's sitting in Chicago for 5 days now! With that installed, I'm hoping to see more difference between part throttle cruise and WOT at the MAP sensor. Right now, part throttle looks like 90% and WOT 100%. Not much to tune off of.

I also have to figure out how to advance past 28 degrees. Setting the controller to 29 or higher doesn't have any effect.

For now, I have my low cold idle, my rock steady warm idle, my rev limiter, my shift LED and my decel pops to enjoy! If I can get an MPG bump too, this mod will pay for itself!

 

[mmart]

Dude the science learned alone was worth the effort. Everything else is gravy.
Good stuff

 

[Jason Pittenger]

Well, the anti-pulse valve FINALLY arrived! No complaints with the company, it was held up in customs for what felt like forever! For whatever reason, I can't find anyone in the US that sells this. It's basically a one-way restriction, designed for VWs to smooth out the vacuum pulse going to the distributor.

https://us.gsparkplug.com/1x-vacuum..._GB&utm_term=b[/URL]"]The Green Spark Plug Co

Here's the results. I tried 4 different configurations, and rolled the throttle slowly from off idle up to redline. I then graphed the vacuum reading vs RPM. Note, the actual vacuum in the engine didn't change, but the reading at the sensor did. Some of the spikes up or down, below 3000 is me lightly bumping the throttle before a full run. I'm mainly concerned with 2000+ and especially the 3000-5500 cruising range.

The 4 different configurations are:
1. Connected to just one carb, no valve (purple)
2. Connected to just one carb, with valve (green)
3. Dual carb, no valve. This is how I have been riding (red).
4. Dual carb, valve between T fitting and sensor (blue).

If you look at low RPMs, the vacuum levels are MUCH smoother with the valve after the T fitting. At high RPMs, the signal is much cleaner, single or dual carb, and it's lower with both carbs. I'll have to retune with the new signals, but it give me more signal to work with. The decel portion is all below 45, and there's no noticeable difference between configurations. For reference, WOT is about 97-98, pretty much flat across all RPMs. At idle, the vacuum readings (for the 4 listed configurations) are 68, 71, 67, 66. Not much difference there, though the one carb with valve was oddly high.

I also took some graphs of cracking the throttle, but I haven't looked at them yet. There wasn't any difference in feel with the different configurations.

Also, here is an audio/video comparison of running with full choke and ignition retard, vs no choke and no ignition retard. The RPM is about the same, but the exhaust pulses are stronger with choke + retard. It's also more even. At regular idle, you can hear some unevenness, which I think adds character. I cut the video before any of the unevenness sets in.

 

[Jason Pittenger]

I figured out how to advance timing past 29 degrees.
There's a setting in the software to allow going past the timing lobe on the crankshaft.

I'm running 93 Octane during these tests. I should be able to go back to 87 Octane when I'm done, since all my advancements are at part throttle operation. I'm a little scared to try it though.

I ran some more tests with my new air setup (anti-pulse valve).
I slowly ran the RPM up from idle to just over 8000 RPM, and measured engine vacuum.
Then I bumped the timing up 1 degree and ran it again.
First run was 5 (1100) to 25 (3500) degree timing, second 6-26, then 7-27, 8-28, and 9-29.
At each setting, I ran the test 3 times to get an average.
The goal was to find what timing gave the lowest reading (highest vacuum).
I found that with more timing, the engine is able to maintain a given RPM with less throttle.

And this is the resulting graph.

It's kinda a mess like that, so I broke it down into pieces.

This shows 5-25 (blue) vs 8-28 (black), from 1500-5000
I saw about a 2 kPa improvement across this range (0.6 inHg).

This shows 8-28 (black) vs 9-29 (red)
At 4500-5000 RPM, 29 degrees showed an improvement as compared to 28 degrees.
Otherwise, 8-28 was the same or better than 9-29.
No advantage to 9-29 below 4500 RPM.

From 5000-8500, there was a big improvement going from 25 (blue) to 29 (red) degrees.
Look especially at the 5000-6200 range! About 10 kPa (3 inHg)!

Comparing 28 (black) to 29 (red), you can see that improvements were starting to diminish.
However, it's possible 30 degrees will show further improvements.

So now a full scale comparison of 5-25 vs 9-29 timing.

I'm excited because of the big improvements in the 5000-6000 range.
This equates to 70-80 mph.
Even the 4500-5000 showed some improvements (~63-70 mph).

This is my resulting table.
WOT is the bottom line, and that's the stock timing.
Anything above that is interpolated to the values on my graphs, based on engine vacuum.
I HAVE NOT run this table yet.
I need to log a ride on stock timing first, to double check IAP values to see how they compare to my slow rev tests.

I had no indications of engine knock at any point.
I checked the spark plugs and they look like the top picture on this site.

Spark Plug Reading Information

Spark Plug Reading Information

www.seadoosource.com

The manual suggests that timing between 4-6 at idle, and 24-26 at 3500+ RPM is acceptable.
I extended the horizontal lines over to the side of the graph, found in the service manual section 15.
I'm guessing that that's a safe timing then, even at WOT.
Otherwise the manual wouldn't put tolerances on the timing chart.