classic retrofit CDI+ and advance curve

[fizeus]

I'm reading the CDI+ documentation and manuals.... I admit I'm intrigued by this product...
Just it's not clear to me how the cdi+ can control the advance curve of the distributor.

for what i know the advance is controlled by weights in the distributor that under the centrifugal force opens or close.
if the advance mechanism is locked how can the spark arrive "on time" to the spark plug if the distrubutor is 35' degree off??

[Jonny H]

Example: The signal is delivered to the CDI+ 35 degrees before TDC. If the box fires immediately, you would get 35 degrees advance. If the box 'waits' for a time equivalent to 35 degrees you would fire at TDC. So varying the amount of delay you can get any curve you want.

See here: https://youtu.be/GK-gD9_gziU

[mysocal911]

Originally Posted by Jonny H

Example: The signal is delivered to the CDI+ 35 degrees before TDC. If the box fires immediately, you would get 35 degrees advance. If the box 'waits' for a time equivalent to 35 degrees you would fire at TDC. So varying the amount of delay you can get any curve you want.

See here: https://youtu.be/GK-gD9_gziU

based on the calculated engine RPM and its change from the previous RPM

[James Brown]

the box can't predict the future (like us) But it can retard to begin with, like for starting, then it goes to advancement as rpm increases following the program you set up...magic!

[Jonny H]

^^ Yes and yes.

[mysocal911]

Originally Posted by James Brown

the box can't predict the future (like us) But it can retard to begin with, like for starting, then it goes to advancement as rpm increases following the program you set up...magic!

But properly designed firmware can modify the estimated RPM for the next spark during acceleration,
and thereby determine the optimum spark delay for the desired advance.

[Jonny H]

Originally Posted by mysocal911

But properly designed firmware can modify the estimated RPM for the next spark during acceleration,
and thereby determine the optimum spark delay for the desired advance.

True. We have honed our own prediction scheme. Without a trigger wheel, you only have 3 reference pulses per crank so acceleration comes into play.

Well known prediction scheme described here:

Alpha-Beta-Gamma Filtering

[fizeus]

Probably, something more basic is missing to me.

When I lock the advance mechanism the distributor became a fixed distributor type because weights can't move.
Now reset its position, so when cyl1 is at 5 degrees before TDC the rotor meets the connector 1 at the distributor cap and can successfully deliver spark to the plug of the first cylinder.
It will always be at this position regardless RPMs (distributor weights can't move and can not alter timing).
If the engine goes to 6000 RPM the CDI+ has to advance the spark signal 35 degrees before ... BUT 35 degrees before TDC means the rotor is quite in the middle of the sixth plug and the fist plug of the distributor (360 degrees / 6 cylinders = 30 degrees)... so the spark can be delivered to fifth cylinder (misfiring event )of the first cylinder (correct firing event) or none of them (probabilistic event)

can anyone please explain ?

[Jonny H]

Originally Posted by fizeus

BUT 35 degrees before TDC means the rotor is quite in the middle of the sixth plug and the fist plug of the distributor (360 degrees / 6 cylinders = 30 degrees)...

Not quite...

The distributor turns at half the engine speed so 35 degrees on the engine crank is only 17.5 degrees at the distributor. The rotor does not move that far - typically about 5mm for 30 degrees of engine advance. Although it 'moves' with relation to the post, there is still enough overlap whether on 0 or 35 degrees. The rotor tip is about 40 engine degrees wide.

[fizeus]

Ok ... thank you ... Jonny H.

Now it's all clear. !

Thanks again for your clarification ... I missed that the rotor turns half the engine speed !

[47silver]

how do you pin the distributor to full advance?

[Jonny H]

Originally Posted by 47silver

how do you pin the distributor to full advance?

See here: https://youtu.be/GK-gD9_gziU