This not a lot of detailed info about these old 1G's out there, so I'm having to do a lot of guessing...
My problem is the Check Engine light does not illuminate. The bulb and all circuits external to the ECU board are fine. The board is also good between the 12VDC into the board from the bulb and ground. I suspect the trouble is the transistor, resistor or any associated capacitor, but the board has been worked on before and the markings on these parts are no longer readable.
Without a schematic, I wonder if someone more familiar with this type of circuit could guess at a possible replacement for the transistor and capacitor.
I know this is not the "proper" way to do it, but the alternative is buying a new ECU - several hundred $$. So I'd like to try a few things that might have a chance of working before doing that.
Here's the only schematic I can find that shows the transistor and resistor on the board, but no info as to what the parts are. Also, can't find what is connected to the base, and can't get in to check the base signal. I can swap out the parts from the component side of the board OK, but can't trace the leads without damaging something else on the board -- due to the condition of the board.
Probably sounds goofy to you guys, but until I can get that light to work and the car to pass smog, I can't justify spend much cash on restoring the car.
If you can get to the trace connected to the base of the transistor and solder a wire to it, you can connect a meter and measure the voltage there. It needs to go to about .6 volts above the emitter to turn on the light. You could also connect a 2.2 K resistor to that wire. When the other end of the resistor is connected to +12, the bulb should light. If it does, the failure is in the base circuit and you'll have to pull the board to diagnose further. If it doesn't light, then replacing the transistor or resistor or repairing traces on the board is indicated, and you can read the rest of this long post for more information on how to figure out the part values.
The transistor is an NPN. Get an NPN of the same physical size, with the highest current rating (see the datasheet) you can find. Even better, determine the current rating of the check engine lite bulb, and get a transistor rated for at least that current. The resistor between the collector of the transistor and the bulb limits the current available to the bulb. The value of that resistor can be determined by measuring it with an ohmmeter, unless it is burned up/damaged/cracked etc.
If the resistor value can't be measured, you can figure it out mathematically or by experiment. Here's an example using math:
Lets say the bulb is rated at 12 volts and 800 mA current. That means that when the bulb is hot, it has a resistance of 12/.8 or 15 ohms. Now an auto can provide say 14.5 volts when the engine is running, so we want to drop that down to no more than 12 volts. And to lengthen the life of the bulb, dropping down to 11.5 volts would be better. So, if the target voltage is 11.5, we want to add a resistor that will drop 3 volts. The circuit would look like this:
+12 ---[Bulb]---[Resistor]---[Transistor]---Gnd
That's what you have in the schematic you provided.
When the transistor is switched on, it will drop a small amount of voltage, but for this example we'll ignore that.
So our circuit is the bulb, which is 15 ohms when hot, in series with an unknown resistor that we want to drop 3 volts. That means there will be 9 volts across the 15 ohm resistance of the bulb, and 3 volts across the unknown resistance when 12 volts is applied to the circuit. We can solve for Ru (the unknown resistance) using ohm's law: Voltage (E) = Current (I) * Resistance (R) We know the voltage across the bulb's 15 ohm resistance is
9 volts, so 9 = 15 * I. That means I = 9/15 or .6 amps
That same .6 amps flows through Ru, dropping 3 volts, so
3 = Ru * .6 and therefore Ru = 3/.6 or 5 ohms. That 5 ohm resistor will heat when .6 amps flows through it. The power dissipated will be 1.8 watts, and it makes sense to use a higher wattage resistor. So in this hypothetical example, a 5 ohm 5 watt resistor would be a good choice.
Here's how to do it experimentally: Connect the bulb to a variable voltage supply. Gradually increase the voltage until the bulb is bright enough. Note that "bright enough" means less than full brightness. Measure the voltage and the current. Call that the target current. The unknown resistor needs to drop the difference in voltage between bright enough and 13.8 volts, at the target current. At this point you can determine the value of the resistor you need to add mathematically as follows, or continue experimentally as in the next paragraph. Vdrop = 13.8 - measured voltage; I = target current, so
13.8 = R * I, therefore R = 13.8/I. R is rated in ohms. The power dissipated (rated in watts) by that resistor must be at least Vdrop * I You want a standard wattage resistor that is at least 50% higher wattage than computed, and double is better.
Continuing with the experimental method: Now add a 1 ohm 1 watt** resistor in series with the bulb, and raise the voltage until you get the target current again. Repeat the process of adding 1 ohm resistors until you reach the target current at 13.8 volts.
** = the wattage presumes that the target current is
1 ampere or less. If higher than one ampere, use 2 watt resistors, if higher than 2 amperes you would need 5 watt resistors - but it is highly doubtful that dashboard bulbs would draw current that high.
and all circuits external to the ECU board are fine. The board is also good between the 12VDC into the board from the bulb and ground. I suspect the trouble is the transistor, resistor or any associated capacitor, but the board has been worked on before and the markings on these parts are no longer readable.
type of circuit could guess at a possible replacement for the transistor and capacitor.
buying a new ECU - several hundred $$. So I'd like to try a few things that might have a chance of working before doing that.
resistor on the board, but no info as to what the parts are. Also, can't find what is connected to the base, and can't get in to check the base signal. I can swap out the parts from the component side of the board OK, but can't trace the leads without damaging something else on the board -- due to the condition of the board.
work and the car to pass smog, I can't justify spend much cash on restoring the car.
Yeah, that's basically the same circuit I was putting together for a different application, but it would not indicate a real fault.
and all circuits external to the ECU board are fine. The board is also good between the 12VDC into the board from the bulb and ground. I suspect the trouble is the transistor, resistor or any associated capacitor, but the board has been worked on before and the markings on these parts are no longer readable.
type of circuit could guess at a possible replacement for the transistor and capacitor.
buying a new ECU - several hundred $$. So I'd like to try a few things that might have a chance of working before doing that.
resistor on the board, but no info as to what the parts are. Also, can't find what is connected to the base, and can't get in to check the base signal. I can swap out the parts from the component side of the board OK, but can't trace the leads without damaging something else on the board -- due to the condition of the board.
work and the car to pass smog, I can't justify spend much cash on restoring the car.
Thanks, struck out with that -- or finding someone to repair this one.
and all circuits external to the ECU board are fine. The board is also good between the 12VDC into the board from the bulb and ground. I suspect the trouble is the transistor, resistor or any associated capacitor, but the board has been worked on before and the markings on these parts are no longer readable.
type of circuit could guess at a possible replacement for the transistor and capacitor.
buying a new ECU - several hundred $$. So I'd like to try a few things that might have a chance of working before doing that.
resistor on the board, but no info as to what the parts are. Also, can't find what is connected to the base, and can't get in to check the base signal. I can swap out the parts from the component side of the board OK, but can't trace the leads without damaging something else on the board -- due to the condition of the board.
work and the car to pass smog, I can't justify spend much cash on restoring the car.
Thanks a lot Ed. It's been a long time since I dug into components. I will post any progress I make with this.
The +12 from the LED connects to this component unfamiliar to me. Anyone recognize it? It is labeled CM104. Can't get a clear view for a photo. Rough sketch....
formatting link
Most connections on this part of the board are underneath parts I can't get to. No markings on it that I can see. There are 7 of these on the board, all packed close together under a metal piece which is not practical to remove.
Ok. That means the bulb draws about 71 mA and provides about 200 ohms resistance when hot. If the schematic is correct, then you can use a 22 ohm 1 watt resistor between the transistor's collector and the bulb to replace the existing resistor, if it's bad.
Thanks Ed. My problem is finding the transistor. The circuit connects from pin 64 of the connector to CM104. I'm lost from there. There's no way to know where it leads from there, and no schematic showing where the base signal comes from.
I'll keep looking for a schematic. In the meantime, I'll take the suggestion to his the junk yards.
Looking very closely at the photos, I would guess that the "CM" designation is for "Connector Mating" ! There seems to be several CMxxx designations printed on the board. If there are any numbers printed on the actual device it may help to know.
It also makes my guess at the components value invalid !
Hmm... I traced the pin-outs to the connector and labeled them. There's only one input from each component, so that seems to eliminate both the transistor and resistor that are in series with the light. Check this out...
I identified the pinouts on CM104 if that's any help....
formatting link
Not Mitsubishi, but here's a typical block diagram from Toyota...
Groan. In fact, double groan. Tracing the thing out on the ECU, if that is possible, still might not yield the fix, because the transistor/resistor might be good. If you could trace it to the transistor/resistor, you could use just about any NPN transistor to replace it and add the 22 ohm resistor, but that still might not fix the problem. And, who's to say something else won't become defective in the disassembly/troubleshooting/repair/reassembly activity. Grumble.
connects from pin 64 of the connector to CM104. I'm lost from there. There's no way to know where it leads from there, and no schematic showing where the base signal comes from.
suggestion to his the junk yards.
Ditto to that, Ed. I'm not one to give up but once I get a working ECU, I'll tear into this one and keep looking. This particular version was known to have used faulty capacitors. This is the 3rd time in 10 years I have traced the problem to the board. Unfortunately for me, there are no more caps to replace. Lots of corrosion though where I can't get to it without taking a chance on ruining something.
My comment on smog checks.... this car is WAY underneath the pollution emissions and runs like a top. Two things prevented it passing -- the check engine light and the mechanic couldn't find the timing mark. LOL!! I couldn't find it either. Had to keep tapping the engine and marking the pulley feeling every stinking 1/2 inch of the edge. Tried smearing it with oil and every other trick. Looking for it with a bright timing light wouldn't even work. Finally found it and marked it with paint, and yes the timing was exactly right where it should have been.
As much as I would like to keep it, once it's up and smogged, it's going to El Cheapo paint shop for a shiny bright red coat, adding the gaudiest wheel covers I can find, a license cover that says "BABE MAGNET" and For Sale sign and parking it right in front of the high school.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.