loudhvx, look inside

Good read for sure and pardon my joining into this discussion so late!

Just waded through it all and sounds interesting but I have some questions bouncing around in my head after digesting this;

Isn't there a fairly large lag between vacuum increases and what output you'll get from the TPS? Wouldn't that give you a meaningless TPS reference on the graph until RPM and TPS are stable in correlation to one another? Also if you are going to use an O2, how do you correlate that output to mixture and wouldn't there need to be one for each cylinders header pipe?

As far as a inductive sensor and magnet goes I'm a little confused, are you talking something like a linear transducer? I wouldn't think a hall effect would be that reliable with all of the vibration and heat potential compared to some sort of potentiometer signal. We rep a large sensor company and I'd be hard pressed to come up with a better solution to determining the CV slides position without going overboard in expense.

Wasn't this a problem back when automakers tried fuel injectors in a throttle body set up with a TPS, O2 & vacuum potentiometer tied to an ECM that also adjusted timing? The result was that they couldn't guarantee proper fuel mixture until these things became stable to one another. Seems like a carb would always have a flat, lean or rich spot or spots in various rpm's at least temporarily if not consistently at certain spots due to jet/needle transition.

I guess it would be a superior empirical data collector for tuning compared to the dead reckoning of plug reading so I'm sure it is worth the time and effort but it seems to be getting very close to the next step which would be to convert it completely over to a squirter.

Again, very good read even if you are pushing my understanding of your data collection method, I'll be watching this one to see how it works.

TPS in an injected engine of course gives you better data than in a carbed engine, since there is significantly more lag in the carbed engine before you get the fuel moving. While tuning it is important to do nice steady pulls while accelerating and having nice zones where you hold rpms constant. This will give you nice data to work with, since as you said it takes a while for everything to stabilize. Running around blipping the throttle this way and that will get you nowhere.

The madnet sensors I was talking about are Honeywell units, not sure how they work, but they sense the magnetic field intensity and you can use them for linear position. The PN for these is HMC1501. The array you would need for sensing position with a magnet would be quite costly I think.

These days the automakers can avail of faster processors, better sensors, etc to provide better fuel injection. By looking at the differentials of the sensor inputs the programs can anticipate what to do in advance.

loudhvx wrote: Actually, I think your voltage divider idea will work, but the other components have to be changed a little. The zener will clamp to 9v, the diode will pull down to 1.7v, then the divider will divide by 3... so 3v high, and .6v low. As long as the micro's input is high enough impedance, it should work. No pullup needed on the micro.

The 560 ohm R1 may have to be 1 watt or even higher, if the ignition is turned on while the crank is not turning. If the minimum high voltage is low enough, you can make R2 higher (3.9K) to reduce the low .6 even lower.


(Also, I had to edit the spec on the first diagram. It should be 1 watt, not 1 amp on the zener spec.)

Is that divider right? Seems to me like if the voltage across the zener is 9V then the Vout will be something like 6.7. Since I've already got components coming, I think if R3 is 4.3k and R2 is 10k, I'll get something like 2.7V high and 0.5V low. Is that right?

Sorry. I don't know how I f'd that up. :ohmy:


Yes the 10K should be above the other resistor, and the other resistor would be better to be 5k. 4.7k is more easily available, though so may be better. 4.3k may be a little low. Radio Shack should have the 4.7k in stock if the 4.3 is too low.

Just wanted to make sure my math was checking out. You've been so helpful wiht this, thanks.

Going to revisit this topic. Got my micro controller, its a FEZ, it's programmed in a .NET environment, it's pretty cool, way more processing power than an arduino. Just gotta learn to use it now.

My question today is, how do I turn a negative into a positive? On my acewell speedo dealie, there is a neutral light and a high beam indicator. I want to use the high beam one for my oil pressure switch, since I'm frankly more concerned about my oil pressure than whether my high beams are on or not. But the switch just pulls to ground, and the high beam thing calls for a positive 12V input. Is there a simple way of rigging this up in hardware? Keep in mind I've got components coming for the circuits you already designed in this thread, and there's doubles, so if I can use that crap it'd be great.

I'm afraid you'll have to get an PNP BJT. 2N3906.

I sort of overbuilt the specs since I don't know what's inside the Acewell.



I made one that has a green LED for good, and several red LEDs for bad.

home.comcast.net/~loudgpz/GPZweb/GPZoilPressureWarning.html

I have to get a transistor to activate my relay anyways so thats no big deal, I'll get two.

Problem here is that radio shack was bought by circuit city and they suck balls now, nowhere near as many components. I'll find one though.

I revised the resistor specs a little, and changed how much the circuit can handle. The 2N3906 doesn't handle as much as I thought. The 1K resistors will still work, but using 1.5K reduces the heat in the transistor a tiny bit. The pull down resistor can be anything.

I usually use whatever I have, and I typically keep more robust parts. The 2N3906 will be fine for an LED, but for a bright light bulb, I'm not sure.

I say just try it. If it smokes, the transistor only costs 20 cents. It won't damage the Acewell.

If it needs more power, we can easily step up the transistor to a Darlington power trans. Then you could power an 1157 to blind the rider.

Sounds good. I assume the acewell runs LED's, can't imagine they packaged a bulb into such a small space.

Seems like I have all I need now to wire up everything when my components get here, gonna be a lot of fun wiring everything up, and programming my controller. I can't thank you enough for all your help, I could have figured it out probably, but it would ahve taken a while.

Well, don't thank me until it's actually working. It's usually a long uphill battle before stuff like this gets functional.

Oh I know that haha, just happy to have a little guidance. I should be able to get everything functioning with a little work.