Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

High Current MOSEFT Circuit Resistors

Status
Not open for further replies.
Actually, it forgot compression. GDI injects during the compression stroke.

Higher pressures are necessary because pressure builds in the cylinder also. If there is 3 bar pressure in the cylinder, a conventional injector can't inject.

Anyway, Jaguarjoe, am I good to go on this? Are there any more concerns to address?
 
What stroke of the Otto/four stroke system is missing? I was taught it was - 1. Intake 2. compression 3. power 4. exhaust.

The high pressure is to more completely vaporize the fuel and cause the cylinder temp to lower allowing higher compression.

Are you guy's old enough to remember the "Smokey Yunick hot vapor engine"? The man was a genius and way beyond the times in his ideas. smokey yunick hot vapor engine - Google Search

I still don't know how GM did theirs. Here are two other ways:

**broken link removed**

Preston Tucker should have hooked up with Smokey.
 
Well, here goes $40 on parts! Thanks so much for the help. If I remember later on, I'll post what's going on with the engine here.
 
Do you have access to an oscilloscope? It would be good to see that all of the pulses to the solenoids and coil are as expected.
 
Don't forget the ballast resistor for the ignition coil. Determine the coil current you would like to run at and then figure out the voltage drop aqcross the coil. Ecoil= IR= Ic(Rc). Determine the MOSFET voltage drop. Emosfet= Ic(Rdson). Add them and subtract from 12 volts to determine Eballast, then crank out the ballast resistance. R= E/I= Eballast/Ic. Calculate the ballast resistor power dissipated in watts. P=EI= Eballast(Ic) At least double the wattage you calculated to keep heat somewhat reasonable.
 
I think most ignition coils have ballast resistors built in, but the info is extremely useful to have, thanks!

I actually do have access to an oscilloscope as well as a signal generator. If I hook things up right, I should be able to see the input waveform and the four outputs, hopefully all at once.
 
They come either way. I'm most familiar with the <0.5 ohm unballasted jobs that have current limiting set by an automatic control scheme.

One thing you might look into is multiple sparks in rapid succession. Lean mixtures are harder to ignite than stoich mixtures. The coil primary inductance will determine how close together the sparks can be. E= LdI/dT
 
Multiple sparks is hard to program (PWM), so I am just going with two spark plugs off a single or, if I get some free, two coils in parallel. More current draw, sure, but I've got plenty. I'll just have to recharge the battery every night.
 
Why would you PWM the ignition coil? All it needs is to be turned on a few milliseconds before you want the spark to occur. This lets the coil charge up (dwell). Turn the coil off when you want the spark to occur. If you let the coil remain energized too long it will just build heat. Use E=LdI/dT and solve for T to see how long it takes to charge it up then keep it on just a tad longer or get fancy and set dwell based on RPM. 2 coils in parallel will halve your spark energy j= (1/2)L*I^2 because equal inductors in parallel have 1/2 of the inductance.
 
Well, knowing the second half (about inductance) is good. PWM is the easiest way to drive the coil without involving interrupts with delays. An interrupt should be less than a few microseconds, not milliseconds. PWM takes care of timing and duration. I can easily change it on the fly in 10us intervals. How would I determine the dwell time based on RPM? Also, do you know much about spark timing? I'm not sure when I should advance/delay the timing (other than from knock).
 
I still don't see the PWM concept.
The uP will measure RPM and MAP. The uP will have a table of RPM vs MAP with a corresponding spark advance crank degree setting for each combination. When the engine is at a particular RPM and load combination, the coil driver will turn off at the corresonding crank angle which fires the coil. Generally, timing advances as RPM rises, timing advances at low load, timing retards at high load. Timing is affected by octane, retarded for low octane, advanced for high. Timing is affected by AFR. High AFR's require advanced timing. The object of the whole thing is to achieve maximum cylinder pressure somewhere around 10-20 degrees ATDC. A combustion pressure xdcr would be a big help but is also big bucks. A dynamometer of some sort could help. A knock sensor is cheap but a real bear to set up. Somehow you must differentiate between valve clatter, etc and true knocking.
 
Last edited:
PWM period is based on engine period, and is constantly updates by the Hall Sensor. The PWM tells it when to fire (based on the outlined factors) and for how long (1ms should be enough, I think). Any other method of controlling it involves making the code in such a way that it actively harms other functions.
So I need to make a table (or preferably an equation) based on just testing my engine. Right now, I have a 18 degree firing angle. Not sure how good that it, but we'll see. I might build my own dyno, or else use an electric motor as a generator and hook it up to the crank. I'll increase the resistance on the generator (more load, something like that) and use an ammeter, voltmeter--though that shouldn't be necessary--and frequency meter.
For knock sensing, I'm not even going to try with a real knock sensor. Maybe if I have time... Right now I know I'll have no valve clatter, so knock should be very apparent.
 
ion sensing may be a good thing for the engine. It uses the spark plug to sense flame ionization current immediately after the spark occurs. It can be used to control spark advance and injector timing. I think it works along the same lines that "flame rods" are used to detect pilot presence in an industrial burner. This is not brand new technology but its pretty close.
 
I've seen maps before, but I wasn't sure if there was an analytical way to figure out a basic map for my application. Anyway, it's good that you think my alternator idea would work. You are correct, I do need a large drain circuit. Even at a base 3.5HP, I am producing 2.6kW, which is A LOT! Maybe I'll build some discharge coils that run off it, but those are pretty low current. Any ideas?

I've heard of ion sensing, but I'm not sure how to utilize that with my uP. It might be possible to use a smaller setup with two wires and a gap (maybe utilizing an old spark plug) to do the same. The problem is that I can't read current with my uP, I can read voltages. I think I'll talk to some more nearby automotive engineers.

EDIT: I realized that I can use a water brake (like a real dyno?) as my resistance source. The heat generated corresponds to the work done on the liquid. I am in chemistry (and physics) now, so the math is pretty easy. Another (harder) idea is to use a variable pitch propeller to turn at the same speed in the water throughout engine spark timing changes. This would allow me to read current directly and measure performance from that.
 
Last edited:
The uP doesn't need to measure current. Use a handy dandy current to voltage converter- aka resistor! If you put a 1k resistor in series with the plug and put the uP's ADC across that resistor and use E=IR, 1ma of current flow now becomes 1 volt.

A water brake is a good idea. Maybe get an old centrifugal pump?
 
Ion sensing requires a capacitive discharge ignition system and CDI's are bad at igniting lean mixtures and you'll be running pretty lean. If Ion sensing still appeals to you, Silicon Chip magazine had an article on building a multiple spark CDI unit. Multiple sparks help ignite those lean mixtures.
 
Well here's an idea. I was thinking about using a pull-up resistor actually. If I supply 3.3V to the wires, then I don't need any other circuitry. I'll see about this. I'll definitely consider it if I can figure out how to actually use the data I get from it. It should be interesting to see.

I decided a water brake would cost too much (variable pitch propeller) or would be too complicated/messy (water pump with variable valve opening). Using an alternator, or just some random generator, would be best. The obvious problem is what I would use to increase the load, but I'll work that out later. I can get a 120VAC 3600RPM motor (aka generator) for about $5. Thank god for surplus stores. My project has definitely blown its original proportions.
 
Lots of 240 volt 4500 watt water heater elements out there starting at $3 on eBay. 1500 watt 110 volt units are not as plentiful and more expensive.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top