Carbs and low-load spark advance
 

The title should have been "I don't understand how the lack of vacuum advance is working for the twin-plugged, carbureted guys".......

Many of the carbureted (PMO/Weber) modified builds here have been twin-plugged. Some use a 12 pin distributor, some use Electromotive HPV/HPX, basic dial-based Xdi, and other DIS units. I may be wrong here, but in most every case, these ignition systems have straight centrifugal or dial-based spark advance curves, with no vacuum (load based) spark advance to my knowledge. This couldn't possibly result in optimal cruise/part throttle running. I believe this is causing much of the difficulty that carb tuners here are experiencing when trying to tune the cruise/part throttle/low load area.

My understanding is this (and may be flawed): The centrifugal (or dial-set) advance is optimized for wide-open-throttle performance. Under WOT conditions, the charge density is high, relatively easy to ignite, and has good flame propagation. However, during cruise/part throttle/low load, the fuel/air charge density is fairly low, difficult to ignite, and has slower propagation. For these reasons, most ignition systems, whether distributor based or ECU based, incorporate a load based spark advance feature, advancing the spark to as much as 45 degrees BTDC under low load/high manifold vacuum conditions in some cases. For OEM distributor based motors, its the familiar vacuum servo seen on the side of the distributor, and the ECU motors use MAP sensing.

So the above leaves the twin-plugged carb'd guys, for the most part, lacking the required spark advance under low load/high manifold vacuum conditions. I've seen guys on this forum struggle to understand the hesitation/stumble with lean AFRs they see during cruise/low load conditions when running straight centrifugal advance distributors or dial advance DIS with carbs. The lean cruise/part throttle AFR readings are likely due to incomplete combustion of the difficult-to-ignite low density charge resulting in unspent oxygen in the exhaust stream. It appears that if they ran an ignition that incorporated additional load based spark advance, the cruise/part throttle burn would be far more efficient. Take a look at any of the more successful stand-alone ECU based timing maps for evidence of this. Here on this forum, take a look at this timing map thread for examples of how much low-load spark advance many tuners are running:

http://forums.pelicanparts.com/porsche-911-technical-forum/975423-official-efi-ignition-map-sharing-thread.html

So I'm in this boat. Newish (to me) 3.4, high compression, twin plugged, PMO 50mm carbs, headers, DC 60ish cams. The motor came to me with a pair of old Electromotive HPV DIS (dial adjust timing) units to support the twin plug setup. No possibility of adding MAP based spark advance there. I've researched a bit, and without getting into a full blown ECU solution (AEM, MOTEC, MegaSquirt,etc), the Electromotive XDi 200 seems to be a good compromise. The big feature that separates it from 12 pin distributors, HPV/HPX, basic XDi, and other DIS solutions, is that it has a MAP input for vacuum advance/boost retard. So, with a unit like the Xdi 200, initial spark timing is set (10 BTDC), and a curve from there is specified (otal of 26 BTDC @ 3000 RPM for twin plug). Then, the additional spark advance based on load is specified (an additional 15 degrees at max vacuum/minimum load).

I really have no practical experience mapping spark advance. I've only done a bit of research to try to understand why carbs are so tricky in the cruise/part throttle/low load area. This research pointed to a lack of low-load spark advance as the likely problem.

Thoughts?

Thanks,
Tom

Hi Tom, what’s the compression you are running?

Based on your info, peak torque will be mid to high 5k. Timing should be “all in” above that event. The amount of timing will depend on the mixture that you plan on running, both wot high load and part throttle conditions.

Hi Aaron. Thanks for your reply. Compression is right around 10.5:1. I dont really have a question with respect to how much advance, or where it should be all in.
My post is really about how not running an additional load-based spark advance feature or mechanism, can't possibly be optimal on a motor that is street driven, or spends any significant amount of time at part throttle. Many are running dual plug distributors or DIS that lacks additional load-based spark advance.
So, I should have concluded my post above by asking for anyone that ran without additional load-based spark advance, and then implemented it, and if it really makes as much of a difference in tuning as I suspect it would. Should I really dump my old HPV/HPX for something that will give me load-sensitive spark advance. Remember, I'm carb'd, not EFI'd.

Tom,

This is a great question. I’ve just accepted the guidance of no vacuum spark advance. My distributor has the diaphragm but it is not connected.

Why aren’t we running vacuum advance?

Could it be that the large overlap cams prevent large vacuum signals? Lobe center only 102 degrees

My vacuum system only sees 9-10” vacuum at idle (hot).

Maybe it won’t actuate properly with such a small difference between idle and WOT.

I'm not surprised at the low vacuum with narrow lobe center cams. I'm expecting that with this motor. I'll be logging the MAP data before attempting any spark advance maps based on MAP. I'll start with just a base programmed curve (10 @ idle, 26 @ 3000 for example) . I'd add the MAP based after logging.
Still curious if there are any carb motor tuners with input regarding adding MAP based spark advance.

Problem is if you have a long duration "carb" cam you don't have an acceptable vacuum signal or at best a pizz poor signal. Why most builders ditch the vacuum advance with carbs and why many aftermarket FI/ITB guys end up running in AlphaN/TPS only. The MAP doesn't work well with a poor vacuum, also the reason most car makers including Porsche use a plenum based intake.

I have a pretty hot engine, 3.0, twin plug, webers, 120/104 cams, mechanical advance and my engine runs as smooth as silk.

I plumbed all the vacuum ports in my Zeniths to a common manifold. And I only get 9-10”. If the system is only ported to one cylinder the vacuum gage bounces like crazy.

So this makes sense. Running single plug and 34 degrees advance from 3000 up. Runs well.

The vacuum pot on the distributors used on the 2.4 MFI engines (and 2.7) was not a vacuum advance but vacuum retard. Some Later engines did have vacuum advance. When switching to carbs the vacuum retard is generally not used.

john

Carbs and low-load spark advance
Didn't Traffic have a song by that name?
.
.
.
.
.

humm...oh wait. That I think that was Low Spark of High Heel Boys....
Sorry.

As Tom suggested in post #1, "....during cruise/part throttle/low load, the fuel/air charge density is fairly low, difficult to ignite, and has slower propagation. For these reasons, most ignition systems, whether distributor based or ECU based, incorporate a load based spark advance feature, advancing the spark to as much as 45 degrees BTDC under low load/high manifold vacuum conditions in some cases."

So yeah. At part throttle, low-load situations, when A/F mixtures are lean, ignition timing can be advanced in this zone to take advantage of the slower burn mixture and increase fuel mileage a bit with more complete combustion. I think PAG's intention was to provide the best power with a simplified system instead of a more complicated part-throttle system with all the associated do-dads.

You will notice they did pay attention to more mundane driving conditions on their less energetic engines for more efficient part-throttle cruising by adding the usual vacuum mechanisms to operate in that range.

So, take that BB out of the vacuum hose or reinstall the vacuum hose, then see if all related components still work (vacuum diaphragm, arm and movable breaker plate, then confirm operation with a timing light and see if it makes a difference (dialed down butt dyno) in your daily drives.

Sherwood

VFR750 - What cams are you running? I expect a low vacuum signal value at idle with cams that have alot of duration and/or narrow lobe centers. What does the vacuum signal look like at 2,000-3,500 light throttle cruise?

Good question. I’m not sure.