IonSense/HardwareDesign

ION/CDI - developer info.

In fact two boards:

They are always used together and are installed in the same box.

The combination feature:


To Store/Retreive files in the wiki:


Locking:

Knowing what happens if we develop in the open we keep the lid on thisone just as AfreshTiny. AfreshTiny distribution procedure will work for now. -Jörgen

I have dubbed the boards IonBoard , CdiBoard? and DiodeBoard?. -Jörgen


IonBoard

The board for Ignition and ION sense (Standalone or in use with EFI board)

Filename: Ionboard_r008_je

Status: Unlocked

Aim: Started routing of board

ETA:

IONBoard:

Has the ARM LPC2119, CAN, RS232, one LM1815 VR interface, one hall sensor input, MAP sensor. Also has four discrete inputs that emulate an MSD DIS-4 input. The ION use one ADC channel and a MUX to choose which current mirror to sample from. -Jörgen


CDIBoard:

Has the main power supply, which feed two separate CDI power supplies (needed for rotary and twin spark applications), ION power supply and plasma power supply. It has four CDI outputs and one plasma_enable output that is shared by up to four DiodeBoards?. -Jörgen

Someone would like to sum up old notes, ltspice simulations, etc... ?

CDIBoard

Filename: Cdiboard_r005_jk

Status: Non existent, dependent on initial Ionboard_r005_jk.

Aim:

ETA:


DiodeBoard?

Possibly looking like a short thick HT wire extension. It has one HV input which is directly tied to the HV output, the cathode of the HV diode is connected to this node, the anode of the HV diode is connected to a 500v heat and abrasion resistant wire. The entire assembly is meant to sit on the plug itself. -Jörgen

I was browsing around in supply land and I cannot justify the huge cap-diode network. If you don't like self-wound transformers, a flyback type stepup (with our standard ISLV... DPAK IGBT) must be perfect. Only consumes 1 ADC input channel and 1 logic level (gatedrive) output. For prototype, safe to have both types (although the cap-diode can be soldered on a traceless carrier-board - I did it earlier for ionizator upto 20000V).

[cheap $1.5 270-μH inductor] I found [here] Is it really 5.5 A?

Filename: Diodeboard_r001_?

Status: Unlocked

Aim: Slap 30-50 HV diodes on a board use a machined pass though for the original HV signal from the coil. Make it fit in a spark plug well.

ETA:


Discussion

When you make a comment, write your friggin name after it:) Like this. -Jörgen

The experimental nature of the ION project makes me feel that we should have one board with the 'smarts' and one board with the power electronics.

proposed interface between the boards

please elaborate on this...

This will enable us to make the CPU board early and try different power board configurations on proto boards. The ION current mirrors will of course be on the CPU board; they will enable those of us who already run CDI ignition to start collecting data early. -Jörgen

I'd like to see as much of our present knowledge on one board. It's better for experiments than if it was several boards. There can be optional parts on it, of course. Maybe the first boards are somewhat more work, but this would bring usable solutions faster. Remember v3 was made much faster than the small round WBO2: just because it covers more uses, so there is more interest among developers.

Valid point, but I would like to see at least a separate IONBoard for ADC/data logging, including Ion Sense, i.e. I see Ion Sensing as data logging. This is because many racing classes do not allow the use of reactive electronics i.e. data logging must be seperated from ignition/fuel control when racing. I want two boards that I can plug together when testing and tuning, during which the data logger feeds the control board and maps the ignition/fuel tables. Then when I race, I can disconnect the two boards, the engine runs off the map and the data logger runs stand-alone. Any comments? - Patrick.

This is pretty much the plan:

Both smart boards will have switches so either could control an engine by itself (one excels at ign, the other in inj). But the swithces are organized at one section so that part can be shaved off for a smaller layout "sensor-board".

Someone could put it to VemsFrontier/Boards, I made the mistake of writing to genboard-dev first, and its late now, going to bed.


Plasma discussion:

Is there a problem with enabling Plasma on a global level? Plasma is basically a capacitor charged to 300-400v connected to the plug, nothing would happen until the impedance over the plug gets fairly low. That will only happen after a high voltage has ionised the gap, if we consider 4 cylinder plasma units we will have ignition events 180degrees apart. The resistance over the plug quickly increase as the cylinder pressure drop off during the work phase. The currents seen during a plasma discharge makes me want to keep the number of switches low. BTW, the HV diodes for each plug will effectively prevent the CDI re-strike that initiates the plasma from jumping to the wrong cylinder. -Jörgen

The above makes me want to have a high current plasma _arm_ switch in each 4 output CDI unit. This is connected to the common anode of the HV diodes. -Jörgen


IonSense/CicuitAndSimulation To continue the development of functional circuits that allow high energy sparks while also permitting effective protection and accurate measurement of Ion Currents

IonSense/Components Links to whats available,data sheets, how affordable and state of the art

CoilOnPlug Developing our own or finding what's available