Anything not specifically related to the Speeduino hardware. Eg sensors, bluetooth, displays etc
User avatar
By tobbera
#35365
Hi all

I'm looking at building myself a logic level timing light to make it easier to check timing, especially on modern engines with COP's hidden somewhere far in where they basically cant be reached with the engine running. While I'm at it I will add inputs for other inputs such as coil primary and traditional pickup on the secondary side. The brain will be based on a atmega that drives Cree high power LED as the strobe. Please let me know what you think.
Annotation 2019-06-06 005618.png
Annotation 2019-06-06 005618.png (71.28 KiB) Viewed 5476 times
Annotation 2019-06-06 010622.png
Annotation 2019-06-06 010622.png (37.09 KiB) Viewed 5476 times
Annotation 2019-06-06 005841.png
Annotation 2019-06-06 005841.png (43.29 KiB) Viewed 5476 times
Annotation 2019-06-06 005914.png
Annotation 2019-06-06 005914.png (39.75 KiB) Viewed 5476 times
Annotation 2019-06-06 005928.png
Annotation 2019-06-06 005928.png (39.14 KiB) Viewed 5476 times
By LPG2CV
#35370
Old grey did some work on this with LEDs. If I recall, the issue is some sort of blur effect as the LEDs don't stop emitting light fast enough. :)
User avatar
By PSIG
#35379
As you can't use variable coil dwell start as a trigger point (inaccurate), and you are wanting to know when the coil actually fires the spark ;) I would first go to either LL end-of-dwell, or the coil negative sensing for the flyback pulse, and trigger a very short and powerful LED pulse from that in order to prevent excessive blurring.

David
User avatar
By PSIG
#35409
As there are different ways to process triggering; have you calculated the processing delay to compare direct-triggering with processor triggering, and if it has impact? If so, I would think you would want to simplify a direct trigger to turn on the LED, and processing only to shut it off, e.g., Arduino controls an N-MOSFET or other suitable turn-off device that is always 'on', but the trigger is off. The discrete trigger turns the LED on, and Arduino processes the trigger to determine when to turn the LED off, resetting for the next pulse/flash. Perhaps something like that, which would allow accurate timing by direct triggering, and minimize blur with processed turn-off, making the worst-case being some flashes are brighter or blurrier than others, while maintaining accuracy of the initial trigger point.

If you have found that there is no accuracy difference or problem at any RPM with your triggering scheme, then ignore all that. :lol:

David
By JHolland
#35435
You will get a delay between the time that the transistor turns off and when the coil fires. The magnetic field has to collapse and the voltage on the coil build until the voltage is high enough to spark, that time varies with many factors. I can't see a logic level light being accurate.
User avatar
By tobbera
#35448
Thank you all for your input. The eventual delay between coil negative being released and the actual spark should be measurable with a scope and if need be, corrected for. I think it needs to be a substantial delay to make a difference to the human eye at idling speeds.
By LPG2CV
#35450
Just thinking out loud here.

Will that delay read on different parts of the circumference depending on the diameter of the the pulley/flywheel.

Does it in fact, already do that with a 'normal' timing light!

I uploaded logs and last tune and dyno plots look[…]

you could set 'skip revolutions' to 3, then it wou[…]

Hi, I am trying to assign Signed values to the x-a[…]

blitzbox

I've finally figured out why MAP and Lambda weren'[…]

Still can't find what you're looking for?