Digital CDI for large engines

Post by **AndreyB** » Sat Jul 01, 2023 6:16 am

With cars we usually use https://en.wikipedia.org/wiki/Inductive_discharge_ignition while smaller and larger engines like to use https://en.wikipedia.org/wiki/Capacitor_discharge_ignition

How do those *digital* capacitor discharge ignition systems control spark energy and even spark duration? For instance how do you produce either 150mJ or 300nJ spark?

kb1gtt · Post by **kb1gtt** » Sat Jul 01, 2023 10:21 am

On small engines I think spark duration and energy are mostly fixed. Components are selected and the trigger signal is also fixed. However I understand that you can get a certain level of control. With CDI you basically charge a cap, then you quickly discharge the cap through a step up transformer. If you charge the cap up to something like 8V, you'll get less energy when you dispatch that energy through the transformer and gap. If you charge the cap up to 12v, you get more energy.

I think the coil on plug setups found in cars are typically CDI. CDI has less inertia in the system which allows for faster and probably hotter spark potential. Spark energy in a coil will likely be lower heat for a longer duration.

I would think parameters for CDI duration and energy would primarily involve gap, transformer step up ratio, cap uF value and capacitor charge voltage. With some searching I'd bet an equation could be found or made which correlates those parameters.

Post by **AndreyB** » Sat Jul 01, 2023 4:40 pm

https://ww1.microchip.com/downloads/en/Appnotes/00001980C.pdf

shows software-controlled DC-DC

: image.png (138.48 KiB) Viewed 52715 times

Post by **AndreyB** » Sat Jul 01, 2023 4:52 pm

https://www.omnicalculator.com/physics/capacitor-energy

E = ½ × C × V²

: image.png (12.68 KiB) Viewed 52712 times

kb1gtt · Post by **kb1gtt** » Tue Jul 04, 2023 9:58 am

The cap and transformer are going to prevent some level of energy from converting into spark energy. Keep in mind the stored energy is a max available energy.

Hmmm. That pic circuit look very similar to the direct injection driver boost circuit.

Post by **AndreyB** » Sat Jul 15, 2023 12:35 am

https://www.homemade-circuits.com/dc-cdi-circuit-for-motorcycles/

Post by **AndreyB** » Fri Aug 25, 2023 7:59 pm

Altronic 69694G Flange Ignition Coil 850 grams of magic

: image.png (10.76 KiB) Viewed 51854 times

Post by **AndreyB** » Sun Aug 27, 2023 11:33 pm

Well, so far I've failed twice. First fail is lame, second fail is epic!

Both scenarios same ECW-FG80685JA 6.8 uF 800 VDC capacitor. Both scenarios using 100Ohm 1W unknown voltage resistor(s)

First fail $8 diode-based High Voltage DC-DC Boost Converter. Small board https://rusefi.com/forum/viewtopic.php?t=286 probably BTS2140 might be fake.

Here I got spark! Problem: only one spark every 15 seconds. This (weak?) converted takes 15 seconds to reach 380v on the capacitor. I assume weak boost converter is the problem here?

: image.png (244.37 KiB) Viewed 51795 times

: chrome_6R3zbCDxRJ.png (319.21 KiB) Viewed 51795 times

Post by **AndreyB** » Sun Aug 27, 2023 11:39 pm

Second fail is so much cooler!

Here I've used $12 High-Voltage Power Supply Module https://www.amazon.com/dp/B01MYLWHZW

As arrived module output was 80v, things were peaceful. I've started cranking pot to increase voltage...

First it has burned one resistor, then it has burned two resistors in series! Note the heat shrink has failed to keep the magic smoke inside. I did not even have a chance to attempt sparking.

: WhatsApp Image 2023-08-27 at 17.04.35.jpg (206.87 KiB) Viewed 51792 times

: WhatsApp Image 2023-08-27 at 17.04.04.jpg (212.78 KiB) Viewed 51792 times

: WhatsApp Image 2023-08-27 at 17.19.50.jpg (253.7 KiB) Viewed 51792 times

kb1gtt · Post by **kb1gtt** » Mon Aug 28, 2023 12:57 am

I recall a QUCS or ltsoice simulation somewhere. Perhaps this would be worth attempting to simulate it.

kb1gtt · Post by **kb1gtt** » Mon Aug 28, 2023 8:55 am

This might have some useful information.

https://www.homemade-circuits.com/?s=cdi

Post by **AndreyB** » Mon Aug 28, 2023 6:16 pm

I know what's going on, I am burning/shorting IGBTs!

I've replaced unknown/old/fake BTS2140 with a legit ISL9V3040 and everything seems to start working at 220VDC. Once I've cranked voltage up I got a single unscheduled spark on the "spark plug" and then yet another resistor went in flames.

Power supply claims 390V so I should not be exceeding the 400V limit of ISL9V3040. There is also "SCIS Energy = 300 mJ at TJ = 25°C" am I exceeding that?

kb1gtt · Post by **kb1gtt** » Tue Aug 29, 2023 9:16 am

Scope traces? It's hard to guess when we have little data. If you scope it out, then make sure you using 10X scale or make sure you have a probe capable of 400+V. Often a 10X probe can handle around 400V. So beware and keep an eye on the voltage capability if you try to use a scope.

I might guess you have some kind of kick back, which is exceeding the voltage. I sounds like you are hitting the Over Voltage Protection (OVP) decently hard, which probably heats up the driver chip. I suspect it then eventually causes a thermal shutdown, or breaks the OVP circuit. Once OVP is broken, then I expect you get additional damage, as the kickback voltage moves upstream breaks other components.