I've been playing with recurving my distributor, and this is for informational uses.............. (work in progress)

Car Specs:
1964 Plymouth Valiant V100 4-dr
14750 miles
Engine - 225 stock
Carb - stock
Transmission - stock 904
Rear End - 2.93
Tires 195/70R14

Testing:
a)87 Costco gas
b)vacuum disconnected and plugged
c)Consists(ed) of measuring the stock curve with dialback timing light
(for consistency all timing events referenced back to 0 with timing light
(idle at 0)), I didn't use the timing marks (i.e. I'll get consistent errors) to
determine the curve, then I set initial timing.
d)distributor locked up (removed springs) and then timing advanced in 2.5degree intervals looking for ping and noting rpm if it happens.
e)determine 'ideal curve' from b&c......
f)try to match curve....... measure
g) test run anyway to see if things get better (or does it still ping when I expect it to) goto f)
h) I did use the brake to either slow down acceleration or hold speed

Notes:
1) Stock curve will occasionally ping when set to 2.5 degrees BTDC
2) I had to get to 17.5 degrees timing (all rpms) before there was any hint of a ping.
3) Looks like up to 15 degrees Idle should be safe
4) The in by 3000rpm rule doesn't seem to apply.
5) After centrifugal is done, I'll attack vacuum
6) And if you want to know, the stock torque converter stalls at 1750...

Picture:
1) Purple Curve is stock
2) Pink is what I think the ideal curve should be (will modify if testing indicates otherwise)
3) Yellow is where I'm currently at.... and it does ping where the yellow curve is above the ideal curve....



Looks like I need some different springs..... stiffer on the primary to flatten that part out, and lighter on the secondary to let it rise faster.

BTW, the inflection point in the curve is determent by how far the weights move before hitting the secondary spring..

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

To the above I ran the vacuum advance on its loosest setting.

Just around 2800rpm there was an occasional ping. (Same area where I think centrifugal is above the ideal setting, so it could either be to much centrifugal there too)

Vacuum specs

9.5x can (19degrees crank)
5.5" start advance, 10.5" end advance.

Not really concentrating on vacuum yet, I'll get the centrifugal first.

Some more centrifugal tests to follow shortly (need to post some graphs)

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

I used a vacuum gauge to determine the best initial timing.

T'd into the PVC line for the vacuum gauge, and kept the idle speed at 600 rpm.

Here's the results

5 degrees 16.25"
7.5 deg 16.5
10 deg 16.6 (just over 16.5)
12.5 deg 16.75
15 deg 16.75
16 deg 16.7 (started dropping)

So the ideal initial timing seems to be from 12.5 to 15 BTDC

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

This is just seeing when vacuum appears on the ported distributor line, no load, just raising rpm.

Starts @ 750rpm
6" @ 860
10" @ 920
14" @ 1100
15" @ 1450
18" @ 2000

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Ok, here's some more centrifugal tests.

I cut down the total mechanical from 23 to 19.5 for these tests (safety wire twisted around end)

Yellow line is what I consider ideal...

Test 2 is with a slightly stiffer primary spring.

The curve flattened a little (hard to tell), but pinging when above the ideal line.

so I went to Test 3, which is with a primary spring that's twice as stiff.

Curve flattened quite a bit, and now I can have more initial.

Test 3.1 is with 7.5 degrees initial. No Pinging, and it's below the ideal line.

Test 3.2 is with 10 degrees initial. Slight pinging from about 2750 to 3250, which is above the line.

Note: The ideal line was slightly modified from the first test (above) due to results... (Ok it's still a guess, but a better guess... )

So that's where I'm at.............

I'm trying to get it setup so that it will follow the ideal line as closely as possible (just below); and then with 16 degrees initial, a slight ping everywhere, and at 12.5 degrees initial, no ping at all, then finally end up between 12.5-14 degrees initial timing.


Test 2



Test 3.1 & 3.2

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Ed,

Great stuff. Keep up the good work. That last blue curve looks very good, BTW.

Lou

_________________
Home of Slant6-powered fun machines since 1988

Very impressive works.

Keep up the great documentation, it will be very helpful to have this on record next time I do a recurve.

_________________
1980 Aspen 225 super six

Lighter primary spring than last time. (48lb spring)

Test 4.1 is with 10 degrees initial
Test 4.2 is with 7.5 degrees initial.



Doesn't look like I gained anything, but I actually have........... The curve did exactly what I expected, and the engine pinged exactly where I thought it would.............. Test 4.1 started to ping ~2750 and never stopped, Test 4.2 had a slight ping ~3300 and then it went away............

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Whoops, forgot to mention that no secondary spring for both tests...

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Springs:

Measured rate of stock springs (with caliper & spring gauge)

primary 29lbs/in
secondary 154 lbs/in (doesn't engage until last 5 degrees (elongated loop on one end))


I've been making the springs for further tests. I acquired some springs from http://www.mcmaster.com/ that are .25 OD, and have 20 coils with a spring rate of 17lbs/in. I've been cutting off coils to increase the rate, and bending them to make them fit. Haven't measured them - just mathematically figure rate out from remaining coils.

Test 2 was with 36 lb primary & stock 154lb secondary
Test 3 with 63 lb primary no secondary
Test 4 with 48 lb primary no secondary.

Things:

You'll notice that the observed advance curve is not linear. If you disregard the waviness which I think is partly observational error and some initial stickiness, they tend to curve up. (see test 4).

I think this is due to the pivoting system of the weights and the slight variation in the leverage due to the spring attachment point as the weights move out. (Advance mechanisms that use a cam arrangement (like HEI) can have more freedom with the shape of the curve)

The primary spring has to deal with both weights until the secondary spring is activated (They're interlocked). So the initial spring rate is Pri lbs/in, then when the Secondary spring starts, the next rate is (Pri + Sec) lbs/in. Giving the 2 step rate curve you commonly seen.

Greater spring rate flattens the curve

Greater preload on the primary spring tends to shift the advance curve to the right, advance starts later, but tends to follow the same curve, and vice-versa.

The secondary spring might have some preload if it's built into the spring, otherwise it sits there until advance hits a certain point which is also adjustable by bending the spring attachment arm.

Now why test 4 wasn't a complete waste.............

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Take a look at the following picture:



Line X.1 is the 48 lb rate curve adjusted to initial 12.5 degrees, and to the right with more preload.

It's good until 20 degrees of advance.

Now Line X.2 is the upper part of the 63 lb rate curve adjusted to start at 20 degrees.

So the theory is that I can get the advance curve to follow line X.1 until 20 degrees, then have the secondary spring kick in and follow line X.2 up. Following these lines as indicated would mean a secondary spring of (63-48=) 15 lbs that engages at 20 degrees (or 7.5 degrees without the initial).

I've made a 21lb secondary to test with which will flatten out the secondary part a bit, and I may just end up with my ideal curve......

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Looking good! I have lots of questions, but I don't want to distract from your results. Rather I hope to open the eyes of future researchers to some of the questions that you have not yet asked.

Your ideal advance curve is rarely ideal in real life. Isn't Engine load pretty important in the calculation of ideal? That suggests that it might be better to design for a heavily loaded curve for the mechanical advance, and then tune the vacuum advance to modify the curve for lighter load conditions.

The test you are doing are still VERY valid as far as understanding how to eventually modify the mechanical advance to whatever curve is necessary. I look forward to more results.

_________________
1980 Aspen 225 super six

That's my WOT ideal advance curve, I am doing the WOT/mechanical advance first (which is the recommended method;WOT,then vacuum). Less load with vacuum advance would add to the curve (different ideal) after I get the WOT curve set.

Different engines, different cams,exhaust etc.... will have different WOT curves (and vacuum advance needs). Every engine will be different. The farther you are away from what I'm running most likely the bigger the differences.

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes

Heh ED

Great work, man! I've been working on this very same thing for the last months in order to compile some useful article with SL6RAM. I Kinda felt discourgaged when I found out that WOT curves ain't that stable and even the most minute diffrence can throw your calculations away. I've done my share of govnors wrecking, now I have a 15 initial BTDC, medium fast opening curve up to 23BTDC@2750 rpm, that's when the heavy spring kicks in and approximately when vac advance starts to fade out (depending on throttle opening, load, etc) and I have dialed total mech 28BTDC@3100 rpm. Vac pod is a 8.5R stiffened 5 or 6 (?) turns CCW.

in addition, I've started to work with computer controlled curves. I'm working with a friend on getting a distributorless ignition sistem for the slant. It would be crank triggered (already got that figured out) and would have multispark capabilities, crank speed reader using another piece of magnet 10 degrees before each 120° (so we can get crank speed variation faster, with 3 readings per crankshaft turn). We're working with 3 coils (one coil per 2 cyls) . The dialing in of advance curves would be just like reading your charts (actually more like making those charts) and we'd like to get say, 5 diffrent advance curves choice from dash mounted display, and the ability to advance/retard those curves. Of course, all this trouble is in order to ease a future fuel injection conversion.

_________________
Juan Ignacio Caino

Please use e-mail button istead of PM'ing. I do log in sometimes but I'll be answering quicker thru e-mail.

Thanks


Looks like you have a little bit more advance than I got. Just from what you stated, I'd say you have a larger duration cam than I have (still stock '64)..........

BTW, my secondary spring is actually weaker than the primary.....

For a computer, you'd need a least 3 axis's.... (or a lookup table instead of a lookup array), and interpolate when you're between readings.

1. RPM &
2. Load, a map sensor for vacuum would be the equivalent of the vacuum advance can.
2a. map sensor could be expanded for boost......

Have multiple tables, and I'd throw in a couple of retard features....
start, below certain rpm, above certain rpm, shift the whole curve up or down... change tables depending switch settings, temperature, whatever..

Plus tables are alot easier to update than springs............

Actually with a computer, I think you work backwards and instead of an advance curve, you have a retard curve..... Have the trigger at 60 degrees (I forgot the key sequence to get the little degree sign) BTDC. Then you delay for spark. This would get you about a 0-55 degree advance range (5 degrees for calculations....) you could even go negative or ATDC if you want.

The commercially available ones (MSD,etc) don't deal with vacuum, I think Haltech can do anything, but it's a bit pricey.

_________________
Ed
64 Valiant 225 / 904 / 42:1 manual steering / 9" drum brakes