Thinking about dual pmags ...

[Toobuilder]

It's funny how these threads have those who strive for perfection and those who say "good enough".

Not sure how many times this needs to be said, but there is a significant difference in optimal timing for a rich or lean mixture. An EI that uses RPM and MP has NO WAY to know which mixture setting is being used because mixture is a pilot discrete variable. Without the pilot making a discrete command to the ignition to tell it what the mixture is, the EI will be wrong for one of them. It's either going to be optimal for ROP and "good enough" for LOP, or the other way around. Without the discrete you will be too advanced and hotter than needed during high power ops or not advanced enough and slower during LOP cruise. There are no alternatives if you only have a single map. Period. Dot.

[Brantel]

Originally Posted by Bavafa

This is what I was just reviewing again and if I understand it correctly and have never tried it, that refers to change timing on an existing and running program on the fly (I.e. increasing or decreasing Max advance) and not sending a new and saved timing curve/program to the PMAG.

What I was suggesting is to have a program, lets say C, and configure that for 20-23 degree and another program, lets say D, that has the curve of 25-28 degree. You start with the C and as you climb and get to your LOP setting, change the program to D by pushing that to both PMAGs.

That?s why I posted that there are two different download methods in my post.

Nigel seemed to indicate that there can be periods of no fire while downloading and occasional serious issues where one PMag loses its mind during online changes. Not something most people want to accept for normal use.

[DanH]

Same as before, revised for clarity and a minor math correction:



Now, a bit of data fun. In #33, Ross posted a flame speed chart from an entirely different source. At first glance, it may appear to quite a lot different compared to the above chart...



However, look close at the Y axis on each chart. They are reversed; a lot of crank degrees is the same as a slow speed in feet per second. So, flip the chart from SDS so the velocity scales match, and...well, facts are facts.

[M McGraw]

Originally Posted by BillL

...Dan/Marvin, are you finding that after at cruise and LOP, and under 8000 ft, that the timing can tolerate some advance? And a little speed increase? [yes, I am challenging the rule "timing advance under 8000ft provides no benefit"]

I waited until this thread appeared to be dying down so I would not cause too much thread drift.

To answer your question, yes I find advantages to using advance well below 8000’; however, I only like WOT and the -14 will jump up to yellow line fast. My procedure is to set WOT then immediately do “the big pull” flip the switch to LOP advance curve and find peak from the lean side. The lower my altitude the leaner I need to run to keep outside the red box and below yellow line. Rather than thinking of altitude as my limit I use percent power and MP.

[DanH]

Further to Bill's question, I went back to CF Taylor for another chart, this one flame speed vs intake pressure. As before, I've de-cluttered and added familiar terms, here translating psi to inches of mercury. The box defines the manifold pressure range of interest for non-turbo engines operating below 18,000 feet, standard day, 15 to 30 inches Hg. For the entire range, flame speed to 95% alight only slows by about 6 degrees at 18,000. Bill asked about 8000 or less. Pressure there would about 11 psi, thus a climb from sea level to 8000 would only slow flame speed by about 4 crank degrees.

In general, pressure change alone (i.e. altitude) has less effect on required advance than mixture change. However, when combined, the charts spell out why we use advance with low manifold pressure and lean mixture. At the same time (and in particular for the angle valve) they show why it makes little sense for a non-racer to use the same advance map when climbing at WOT and best power mixture. Best power mixture (150 ROP, for example) requires zero additional advance (go back and look at the flame speed vs mixture chart), and the pressure loss with altitude requires only a little.

[dreed]

As a rusty pilot (really rusty) that is close to having a flying 7A soon, I've been following this thread with interest. One question I have is why the angle valves like less advance, or should I say react more poorly to excess advance?

I have a 390 in my 7A with dual P-mags

Thanks- sorry if a dumb Q I should know

[912ry]

I took the plunge last year when my mags came due for inspection. Since I'm local I picked them up at the factory. They gave me a full tour and took time out of their busy day to answer all my questions.
The install was straight forward, the mags work perfectly. Starts, hot or cold are no longer an issue. I highly recommend them!
RV-7
IO360 M1B
WW RV200 prop

[DanH]

Originally Posted by dreed

One question I have is why the angle valves like less advance, or should I say react more poorly to excess advance?

It's a good question. The short answer? As compared to the parallel valve head, the intake port, combustion chamber, and piston dome shapes all promote greater chamber turbulence. Within limits, more turbulence during combustion is a good thing, speeding burn rate. A higher burn rate requires less of a head start (more ignition advance) in order to arrive at peak pressure in the 15 to 20 ATDC ballpark.

The general configuration difference is the shape of the chamber and piston. The angle valve chamber is a moderate dome, while the parallel valve chamber is mostly flat-topped. The common piston configurations match the chamber shapes, domed and flat respectively. There are a number of interesting design aspects to the hemispherical chamber, but strictly limiting the conversation to chamber turbulence, the key factors are probably increased swirl and the addition of some squish.



To visualize swirl, imagine looking at a cylinder from the valve cover end. Note the location of the intake tube, offset from the valve. Using your x-ray vision, the port and inlet flow would thus look like this:



The goal of careful port shaping is to promote a spiral flow into the chamber. The flow doesn't run straight into anything, so it keeps a lot of velocity, going round and round as the piston travels down the bore and then up again on compression.

As the piston approaches the top of the bore, two sparks initiate combustion, one on each side of the chamber. Each forms a small kernel of flaming gas which grows, spreading across the chamber.

This is a typical piston for an angle valve. Note the valve reliefs in the center, and the raised areas at left and right. The spark plugs (and our two spreading balls of burning mixture) are above those raised areas. As the piston nears TDC, those raised portions approach the underside of the chamber dome. The effect is to squeeze those zones, forcing the hot gasses toward the center of the chamber. The combined swirl and squish accelerate the combustion process with forced turbulence...thus requiring less ignition advance.



There are other factors, but swirl and squish are typically the most important in the case of any fast burn 2-valve chamber.

Squish clearance (how close the piston gets to the head) is still quite wide in the Lycoming. How far can the principle be pushed? In high performance engines, it is sometimes set to be so close that details like rod stretch and crankshaft flex come into play. Here's one from my personal Wayback Machine, a piston for a pentroof 4 valve Honda. The flat squish areas cleared the underside of the chamber by only 0.035" when measured statically; the close clearance was created by welding filler material into the chambers, then milling it flush with the gasket surface. Very close squish like this serves a dual purpose. Obviously it increases chamber turbulence, but it also quenches gas temperature in the squished areas, killing any chance of detonation in the tiny amount of gas remaining there.

[BillL]

To add to Dans excellent and factual coverage, 4 valve engines typically use tumble rather than swirl to attain faster heat release. Also, there is a volumetric efficiency cost (reduction) due to the intake energy used to generate the swirl that will yield a lower ultimate BMEP by comparison. Squish doesn't affect Vol-eff and the benefit over the slightly increased pumping loss is worth the effort. Piston valve pockets are necessary with the piston-to-head reduction as valves open faster than the piston moves near TDC.

[Toobuilder]

I was going to respond with a quick and dirty answer, but I just knew Dan would come along and save me the effort.

Part 2 however relates right back to the title of this thread. The total timing advance as well as the curve itself is significantly different between the angle valve and the parallel valve. Throw in a few more compelling variables like compression, displacement, rich/lean, etc, and the timing schedule becomes very distinct to each configuration. Yet, the Pmag only comes with one part number that fits anything from a whezzy low compression 235 to a fire breathing high compression AV 400. Does anyone really think a simple timing shift of a few degrees is going to result in optimized timing for the entire Lycoming engine line? No way, no how.