Early 1.9 to 2.1 conversion - stop me if I'm being stupid!

[itchyfeet]

I hooked up a couple MAP sensors and a throttle position sensor to a data logger and recorded them while driving then dumped it into a spreadsheet and made a graph.

so throttle position right, time on x axis and what is left axis? it depends on how a MAP sensor works, is it real vacuum or just up to a threshold, it has a flat top indicating threshold

[CJH]

... and what is left axis?

Inches of mercury? The flat top just corresponds to the maximum vacuum at idle I think.

[CJH]

the article also says

There are two major factors that effect how much advance is required, engine speed and load. You increase advance with rpm and decrease advance with engine load.

so it brings me back to the thing I don't understand ... is revving the engine with no load giving the same timing as with a load

Yeah, I don't follow that either. How does loading the engine decrease the advance? It'll be out there somewhere I'm sure.

[CJH]

Yeah, I don't follow that either. How does loading the engine decrease the advance? It'll be out there somewhere I'm sure.

Increased engine load means a wider throttle opening is needed, which exposes the manifold to more ambient air pressure to fill the vacuum, so the vacuum advance signal decreases.

If that's right, then it means engine load is a secondary effect, and actually vacuum is purely about throttle position.

[itchyfeet]

I'm struggling with this, looks like you have more time for research than me

I am happy to do some timing measurements on my DJ Digijet and DG with carb ...even if I'm not convinced its the same as under load.

[CJH]

I am happy to do some timing measurements on my DJ Digijet and DG with carb ...even if I'm not convinced its the same as under load.

I think it's worth considering - I might give mine a go to see if it's worthwhile. I'm encouraged by this sentence from that article:

You don't need a distributor test stand to curve a distributor. All you need are some basic hand tools, a timing light, and a tach.

[itchyfeet]

Personally I'd be seeking advice from somebody who doesn't time their vehicle by thrashing it up a strip and seeing if they got a hight mph than last time
he probably does not care if it goes bang after a few thousand miles he just wants that quickest time

Total advance is the most critical setting. Short of a dyno the best way to find what total advance you need is some track tuning. The MPH reading at the end of the track is your best indicator of engine output. Make a run or two to get a baseline then increase your total advance and make another run. If the MPH increases advance it some more and run again. Continue advancing the ignition until the MPH starts falling off then pull it back to the point where you had the highest trap speed. It may take quite a few passes until you find the optimal setting.

[CJH]

I did some testing this evening. I warmed the engine up, then connected up the timing light and my multimeter (to measure rpm), and then used the idle speed screw to gradually increase the engine speed in increments. At each increment I measured total advance (with the vacuum pipe attached), and mechanical advance only (with the vacuum pipe disconnected). Of course the engine speed dropped each time I removed the vacuum pipe, so my data points are at different engine speeds.

What did I learn? Firstly my tippexed 10º advance mark on my pulley, which I used to set the timing to 10º a couple of weeks ago, evidently isn't quite right. When I turn my timing light to get the TDC mark lined up (the method needed to measure all the other advance values), it reads 12º. Oh well - based on the earlier post about the differences in mechanical advance between the DJ and DG distributors, maybe that's not a bad thing.

Second thing, although it's the foregoing discussion rather than the test that showed this, the vacuum advance does not have a fixed relationship with engine speed, because it's apparently determined by throttle opening. Under load a greater throttle opening is needed to achieve the same engine speed, and with a greater throttle opening you get less vacuum and hence less vacuum advance. So the following chart can only be compared to another 'zero load' chart.

Anyway, here's the chart:


The offset between the two lines is about 15º, which is within the 12º to 16º range I've seen quoted. It kicks in at this value even at idle - which corresponds to what the earlier chart of manifold and ported vacuum showed, namely maximum manifold vacuum occurs with the throttle closed. The vacuum advance doesn't seem to be diminishing even at the highest speed I measured, indicating that even at that speed the throttle probably wasn't open very much. If I have another go and try to go to higher engine speeds, it'll be when the neighbours aren't at home trying to watch the telly in peace!

My data points are a bit noisy - there's a little bit of 'flicker' in the position of the timing mark when I shone the strobe on it - possibly indicating some wear in the mechanical advance mechanism? It's been stripped and lubricated, but that didn't reduce any wear, obviously. Also, the multimeter showing rpm isn't totally stable - it moves around by maybe 50rpm at times. Possibly this is actual engine speed variation, possibly it's some smoothing algorithm in the meter responding differently at different engine speeds.

Anyway, I think this shows that it's not necessary to do a full sweep of the engine speed to measure the vacuum characteristics. And it shows I could do with more data points on the mechanical advance plot, so that I can do a proper comparison to the charts in the wiki (I'll plot a comparison using what I've got though later). It would be instructive to know at what rpm the DJ dual vac distributor jumps from retarding to advancing, and how quick the transition is.

[CJH]

And here's my measured mechanical advance superimposed on the plots in the wiki, with the wiki's version of the DG shifted to give 12º at idle to match my own distributor. I'm surprised how well the two plots line up, despite the noise in my measurements.

[itchyfeet]

getting geeky now Chris
will measure my DJ

[CJH]

getting geeky now Chris

Yep, never happier.

I'm in good company I think.

[itchyfeet]

Ignition by Paul_Barr, on Flickr

Tested both DJ digijet dizzies and the DG with carb

I think its worth being clear on the graph
Base engine DG or DJ
Distributor type DG or DJ
2e3 carb or digijet

[CJH]

Great stuff.





For the DG in the second plot, the curves look similar to my plots. But the vacuum advance/retard in the DJs look completely different. At first sight the switch from retard to advance comes at around 1500 rpm, but then the maximum vacuum advance looks similar at 15 to 20 degrees at 2000 rpm, reducing a bit as the engine speed rises.

Well done for doing three sets, and for going to much higher engine speeds than me!

[itchyfeet]

I'm begining to think those WIKI graphs are not realistic

worth relabeling the wiki as C advance only

[CJH]

I'm begining to think those WIKI graphs are not realistic

Well at the very least, the experimental method used to derive them must be different from ours. Doing it on a running engine isn't precise enough to reproduce the curves properly.

Regarding the switch from retard to advance on the DJ, that speed of 1500rpm is for the 'no load' case. I don't know how it performs the switch - obviously the advance is from ported vacuum so it comes on only when the throttle opens, but what makes it dominate the retard signal I wonder. So it's difficult to speculate whether that switch point would be different under load. My guess is it moves down the rev range, i.e. it switches earlier.

Since the maximum vacuum advance on the two types of distributor is similar, and since it appears that the retard function was introduced for emissions only, you could almost make a case for disconnecting the retard signal, and possibly taking the advance signal from the retard spigot (to get full-time manifold vacuum). Or perhaps using a DG vacuum unit on the DJ distributor on the injected engine. Or even using the DG distributor on the injected DJ, with the vacuum connected up to the retard spigot. Any of these would give you some vacuum advance even at idle and the engine might be happier. But I think vacuum advance is there to benefit mid range performance, and since the retard function has gone by the time we're into mid range, the difference is probably immaterial.