I'm going to have several electrical questions. 1st, Is pin #8, on the 14 pin plug at the relay board, the only switched 12v+ source in the engine compartment? TIA, Kent

Yes, Can be used as a coil power wire when switching to carbs as well.

Black wire
Click to view attachment

Careful, not all model years are the same. Check the schematic for your car.

I checked all of the schematics I can find and all have the black wires at pin #8 as switched hot from the key switch. To be safe, put a meter on it and verify that it is hot when switched on at the key switch. Battery isn't necessary to check that it would be hot. Just use one meter lead on the 4 red wires, the other on pin #8. With the key to on, you should have continuity. Key to off, open. What year car?
Tom

It's my '73 Tom, which was a 1.7 originally with your fused adapter in place. My plan is to use this wire to activate a new power relay. Glad you and Mike are here to oversee this part.

Question #2 - Do I need to use a N/C relay off the starter circuit to kill the radiator fans when cranking/starting. I'm hoping the answer is no but thought I'd better ask now. TIA, Kent

Kent,
Don't know why as they don't draw that much. Is the Subie ECU controlling them? If so, that feature may already be in place inside the ECU because Subaru found it to be an issue, but I don't know for sure. If you found it to be a problem, it wouldn't be that hard to remedy. A 30 amp relay wired for the starter circuit to open the fan circuit. Basically the fan power to 30, ground to 85, fan energized to 87A (NC contact), and starter power from the key switch to 86. Whenever the starter circuit is energized, the relay opens and fans have no power. Stop starting, power to fans is back.
I think this would mostly depend on how much current the fans draw if the engine is hot and both are running to cool it. Say you drive 15 miles to the store and are inside 5 min and come back out. Would the fans still be running? This would also be determined on the battery condition and CCA. A battery with 500CCA will give 30 seconds of cranking at a big amperage without falling below a certain percentage of battery voltage. So if the starter is drawing 120 amps and fans 30 amps, that should not be a problem unless the battery is on the way out.
Tom

EDIT: Cold cranking amperes (CCA) is the amount of current a battery can provide at 0 °F (−18 °C). The rating is defined as the current a lead-acid battery at that temperature can deliver for 30 seconds and maintain at least 1.2 volts per cell (7.2 volts for a 12-volt battery). It is a more demanding test than those at higher temperatures. This is the most widely used cranking measurement for comparison purposes.

Damn, I didn't consider that the ECU would have that covered.

#3 - What is the difference between a 5AG 30A fuse vs. a AGU 30A fuse. Are they interchangeable? TIA, Kent

Kent,
The designations before the Amperage is for applications such as voltage, current flash over, and several other factors. Also designates glass enclosed or ceramic enclosed, what type of coating on the ends, etc. You would need to look each one up to determine which fuses these are meant for. This is why I stayed with ATC for my fuse blocks. It prevents someone from putting a clear glass fuse that "looks" the same in a circuit and the fuse is not rated at what the person thought. I don't know how many times I have worked on electronic equipment that some one replaced the fuse with one that looked identical until you read the inscription on the fuse that indicated it was for a 28 volt use and it was placed in a 125 volt circuit. Not good, as the fuse was no longer protecting the circuit as the proper fuse would. If you are interested in expanding your knowledge, look up "FUSEOLOGY". Very enlightening.
Tom

Edit: some research discovered. AG = auto glass
AGU = Fast-Acting Glass Tube Midget Fuses

Had to read thru some heavy stuff for my pea brain but I was able to learn this much. The 5 denotes it's physical size. In my case 13/32" x 1.5". That car fuses are slow blow to handle the large surges and are usually the 32V group. Now I need to take a nap.

The easiest location for switched 12v on the relay board is pin 1 on the 4 pin spade lug connector on the relay board. The location on the board is toward the rear on the drivers side.

Click to view attachment

That could be useful info for someone that is using the relay board. I 86'd mine during the Suby conversion, however.

#4- I need to tie in my alt light with the suby engine. I have a single wire from the Suby to connect to the GEN light but don't know where to tie it in. I looked at Jeff B's info (thx Jeff) and found which bulb it is and there are 2 wires. A R/W and a G/R wire. The R/W goes to fuse 9 and to other gage bulbs so I can rule that one out. Maybe a switched + for bulb power? That leaves me with the G/R which goes to the relay board #5 on the 14 pin plug then out the 12 pin #1 to the OIL pressure switch. I don't understand this.

Green/red would be the oil pressure light, not the alternator light.

That one has a blue wire and a red/white wire. The red/white supplies +12V, while the blue wire connects to the D+ wire coming from the alternator to the voltage regulator.

--DD

The alternator lamp work with 2 positive leads. One is switched from the key, through a fuse and to the lamp. The other is the D+ from the alternator. When the alternator spins and creates voltage (2nd positive) the dash lamp turns off. When the engine is not running, the D+ turns "less positive" and the lamp lights up.

That red/white that goes to fuse #9 on the fused side is the correct wire for one side of the alt bulb. The other side is a blue wire that goes to the alternator at D+.
Tom

Odd. I have one sch showing GR yet today I find the one you mention w/ Blue on another print out in Jeff material. Are there two and I have the wrong one?? Many TY's for the assistance, guy's.

Click to view attachment

That diagram was superseded., it was not correct. I try to keep the website materials correct, and I replaced that former diagram on the website in May 2013 with this corrected one. Keep in mind that this diagram is only for the 1974 model years, other years are different:

#5 - If you own a 914 you know the importance of a good ground. As I went through the Subaru harness I counted over 20 grounding points so Suby has a good idea, also.
My question is this. Are there rules that govern grounds? If so, can a layman understand them? Can a ground junction become ineffective if too many grounds converge at one point? Does a ground need to be sized? Should it be the same gauge as the positive circuit it completes? If adding ground points does the material type matter? e.g. brass, steel. As always, TIA, Kent

Cut & paste from my response to another forum where someone was troubleshooting an AFR gauge, but most of it still applies:

""Grounding" is a science all to itself.
Cars typically use chassis as a "Ground reference" and a "Current return path".
For sensitive electronics (O2 sensors) they need a proper "ground reference" that tells the circuitry what voltage is "Zero volts" as far as the system is concerned.

When you use car chassis as this ground reference, then it is not always at "zero volts" with respect to the rest of the car because the voltage of any particular point in the chassis is dependent on how much current is flowing through chassis and where the current is flowing.

So if you "reference" the O2 sensor to car chassis close to the sensor, then the ground can move up or down depending on if (for example) the radiator fans are on or off. This is because current flowing through the chassis causes voltage drops in this reference.

For the more sensitive sensors, you should run a dedicated "ground reference" from the sensor to the unit reading the sensor.
If you use chassis for this, then the "zero volts" reference is not always at "true zero".
If anyone cares and is confused by this, I could make a diagram..... "

"This is where "grounding" gets a little confusing.

The meter/O2 gauge needs +12V power so it needs a ground (called a 'return' in my business because it intentionally carries the return current to the power source).
This would go to the battery negative.

The part of the system performing the voltage measurements within the O2 sensor needs a ground 'reference'.
This ground does not carry any significant current so it does not experience voltage drop during operation.
That ground 'reference' should go from the Measuring unit (gauge or ECU) all the way to the sensor.
Some older sensors (VDO oil pressure for instance) did not provide a dedicated ground to wire to and depended on the car chassis to provide the path.
Newer sensors typically have dedicated reference 'grounds'.

Theoretically, you could hook the reference wire to car chassis at the sensor end and at the gauge end and in most cases it will work.
Is it accurate? Each installation could be different and give different results."




So to answer your questions: "Pickles" :-)

In all seriousness: If the "Ground" actually carries current, then the wire needs to be the same gauge as the "positive" wire or larger.

In a perfect world we would connect all "grounds" to one point, know as the "Single point ground" but we dont usually have that luxury, so the entire car chassis is the "SPG" (we call it 'SPUG' at work).

Material does not usually matter except for long term corrosion problems or highly sensitive signals that could be upset by galvanic voltages being induced.

Did I completely confuse you yet?

Not completely confused, yet. You state that the ground should be the same size "if" it is carrying the same amount of current. How is this determined? Let's say I have a circuit that requires a 12 gage wire. Does the ground carry that same force or has that amount now dropped because it has passed through the device it supplies? Does a lamp circuit that may require 5 amps require a ground the same size or can it be smaller because the lamp has consumed most of the flow?
These reference grounds that you speak of; are they best routed back to the BAT? TOM suggested (and I did this) to loop or connect all my grounds in the engine compartment via a large ga wire. And yes, my 02 sensors do have a separate ground wire. That's a reference ground, right? Thx Kent

The short answer, which is close enough for our purposes:

- Always make the ground wire at least as big as the power wire. If it requires a 12ga wire to power it, it needs a 12ga wire for the ground wire.

Note that the body of the car is about a negative fifteen-hundred gauge wire for most of our purposes...

--DD

Thank you, thank you, thank you!! I needed that!

Jumping in. Iigniting would die. 914 fixed that. Battery dead. Charger started car. Cigarette plug dead. 914 fixed that. Car caught on fire (dash wiring). 914 fixed that. Still a draw somewhere on battery. 914 installed a guillotine kill switch. Found (?) draw right headlinght stays illuminated after being turned off and dropping. Help Shalom, IGTARD

Kent,
Answer for your question about wire size for ground. The current through the complete circuit is the same at all points of the circuit. So 12 Ga. supply into equipment, 12 Ga. ground out.
Tom

Thx Tom, that's what my gut was telling me but this is one of those areas where I prefer professional knowledge over my common sense.
**NOTE** FWIW, I believe I ran into one of those "reference ground" issues, yesterday. There is a small harness that goes from my fuel pump controller to the pump. One of those wires is a ground. If left open the pump runs continuously but when connected the pump senses pressure.

OK, another stupid electrical question. I have not been able to get a detectable ground signal from my ECU to control my radiator fans. I decided it would be just as easy to install 2 thermal senders with different set points to control the Hi n Low fan features. I noticed that many people slammed these senders but upon further reading I found that many were not using relays and put too much of a load on them. That's not a problem for me as I am using relays. But this got me to thinking. If I use the single pole style, which requires the metal block for a ground, will I end up dumping a lot of stray current into my coolant? TIA, Kent

Your steel chassis should still be a better ground conductor than radiator coolant. If your engine and chassis are bonded, the path of least resistance will not be the coolant. Even if there were stray current, it would possibly be from the spark plugs that every engine has. I son't see any risk unless you have dissimilar metals contacting each other.

Where are you mounting these switches? I've seen them in the radiator before. If you are using a grounding type switch on the radiator, you should ground the radiator with a ground strap to provide a dedicated ground.

If the switch is on the engine block, the block should have a ground either on the block itself or on the transmission.

I have one chassis to tranny ground, 2 chassis to engine block grounds, 2 chassis to intake manifold grounds, 10ga ground on the fans and I've chained the engine compartment chassis ground points back to the battery itself. Is that proper or did I overlook something? TIA.

#7 - I need to connect my stock '73 back up lights and will use a relay to power the lights because the old Porsche harness was 86'd . My question is this; should I use the switch on the Suby trans to connect the circuit on the -neg or +pos side? My thought is why run 12v in that area if a ground does the trick and wouldn't be a problem if road elements were to take out a wire. Thoughts? Advise?

Use the negative side to switch the backup lights on and off.

It is usually safer to run 12V to the coil of the relay, then the negative of the coil to the switch on the trans.

One way to look at it is if the contacts of the switch were to be accidentally grounded, you won't blow a fuse; the light will just turn on. If it were 12V, then a fuse would blow.

For my stock trans w/ custom harness, I just took a wire with inline fuse-holder from the starter to the backup switch, then from the switch to the rev lights.

That would be bad with a stock harness if you left it in reverse as that terminal on the starter is hot all the time.

That was exactly my thinking as well. Right down to the backup short circuit displaying a CEL thru the back up lights. When cannibalizing that Subaru I noticed neg switching was the way.


True. And a fuse holder would be prone to corrosion in that area as well.

Will one of our electrical wizards help me out with this please. I want to employee the neutral switch on my Suby transmission. Oddly enough the switch is OPEN in neutral and CLOSED in gear. Life would be simpler if vice verse. My existing system is as shown in the first pic. BTW, the neutral switch is not going to ground at present or in this first pic.

Click to view attachment

After scratching my watch and winding my ass for the past hour this 2nd pic is my solution. Question: Did I miss something obvious and could I achieve the same result by re-wiring the existing 5 pole relay? And will this work? TIA, kent

Click to view attachment

Kent,
What are you using the neutral switch for? Starter solenoid? Safety switch so it won't start in gear ??? If that is the case, then your second schematic is correct!! You learning lectricity!
Tom

Exactly Tom. I almost took out my bench one day. Thanks for the confirmation. I'm not so sure about that learning part though. To tell the truth I learned a lot from that "2 speed 3 relay" fan set up that Bob posted. After that I began to see how you could use those switches in different ways. Now, how quickly I figure out the correct path; that is another matter.

Nice circuit.

I redrew it to better follow the flow.

Well guess what. The Suby Computer didn't like it one bit. It idled up n down between 500-1200rpm every 1.5-2 seconds. Disconnected that relay from the circuit and it idles normal again. I'll have to think about this one. It's very easy to reconnect that relay later if needed.
FWIW, it worked as intended as far as not starting in gear.

EDIT: I just had a thought. I'm pretty sure it was in reverse when the idle began to oscillate wildly. I shut it down immediately and removed the relay before confirming that it would do the same thing in forward or neutral! What if my backup lights let some voltage back thru the backup switch, which shares the same ground post and is located within 3" of the neutral switch.

I have found on my Audi conversion the ECU likes to see an exact impedance from a relay. As an example, the stock main relay is 30 Ohms. A Bosch relay is 15 Ohms. I have to run 2 relays in series to match the impedance (30 Ohms). If not the ECU will see a fault. If your neutral switch is connected to the ECU in any way, the same could be happening. It would be unusual for this switch to directly connect to the ECU but it may be the case.

Using an Ohm meter you should verify the switch is completely isolated from any other circuit. If it is isolated, the effect you describe should not happen. So you may have a wiring fault.

I'm confused on why the ECU would sense what you've done with the neutral switch. Are there connections from the neutral switch to the ECU that you didn't show in your diagrams?

If there are ECU connections, measure the voltage across the switch in neutral or in gear. You would be able to use a pre-biased transistor to interface the switch.

Spoke, no direct connections between the ECU and relays. Both the relays share the same switched ign source. The starter relay is grounded on a dedicated ss ground stud at the firewall. The neutral switch is grounded at the stock tranny ground stud along with the tranny grd and back up switch grd. I remember hearing that ECU's can suffer from ground problems and thought maybe a stray current to a nearby ground might influence the ECU. I guess not, huh? I freaked as ECU's are Voodoo to me so I wanted to disconnect the relay quickly to see if it would return to normal. It did so I will go back and explore further as Mike suggested.
Mike, your probably onto something re: a wiring fault. I re-used a "4 fuse" block that has spade fittings on the bottom. I'm going to 86 it and install a 6 fuse block with the screw terminals on the side. There are 3 circuits on the existing 4 fuse block. Backup lights/bat fused to ECU/switched fused to ECU. Sounds like I'm in the right neighborhood to begin my search. I appreciate all of you guys holding my hand while I tackle this.

Kent,
If the gear selector is in neutral, with neither relay energized, what happens? You said it hunted up and down when the gear was in reverse. next step I would do is put everything back as before when it hunted and put the gear in neutral. If it hunts, pull the power ( fuse ) for the relays and see what happens then. Oh, are any ECU sensor wires running near the relay?
Tom

That is exactly what I plan to do plus check it with the tranny in reverse as well. I hadn't thought about the ECU's sensor wires. I'll double check those.

I wanted an "in use" indicator light in the cockpit for the radiator fans. The fans are negatively switched and their ground wires pass thru the tunnel back to the ECU. I wondered if I could cut in a small lamp in that ground circuit so rather than ask I thought I would try to figure it out. I used a 12v halogen flood lamp for load, a 12v test light and 12v Batt. Pos+ to the flood lamp+ / flood lamp - to test light / test light to Batt -. And yes the test lamp did light but the flood lamp did nothing; nada! Thinking it might be location along the path I flipped them. Batt+ to test light / test light to flood lamp + / flood lamp- to Batt -. Same result. So now I wonder if this it is a demand or load thing? What is happening? Why won't the flood lamp work with the small test lamp in line? Doesn't it complete the circuit?? And what is the best way to wire in an "in use" indicator lamp? TIA, Kent

Kent,
If you have a meter, check the resistance of the test lamp. I'm thinking it may have a high resistance, therefore limiting current for the halogen lamp when you run them in series. Any idea what the current is for the halogen.
The least load for your "on" indicator light would be an LED. Very low current, so you can use small gage wires to wire it in. As a test, put an LED across the radiator fans right up by the radiator fans, in parallel - not series. See if it works ok, then pull the wires in to the cockpit to where you want the indicator lamp.
If you prefer the halogen lamp and the current is high, you can use a relay in place of the LED in the above idea, then run power for the relay from a nearby source in the cockpit. As above, test the way the relay works up by the fans first.
Tom

As mentioned, the LED indicator (actually LED + series resistor) goes in parallel with the radiators fans.

Use the series resistor to limit current in the LED. Allow 5ma, 10ma of LED current. Supply voltage is 14V; LED ON voltage is about 2V. So a resistor of (14V - 2V)/5ma = 2.4k ohms or (14V - 2V)/10ma = 1.2k ohms. Choose a current where the indicator is bright enough to be seen but not too bright.

Power in the resistors is I*I*R = 5ma x 5ma x 2400 = 0.06W; 10ma x 10ma x 1200 = 0.12W. So choose a resistor with at least 1/4W rating.

Over my head. What is "I" in the equation? Any chance you could throw a hand drawn schematic at me?

I is current in the LED and resistor. Here's a drawing of the circuit. +12V switched can just be the positive voltage at the fan or from the fuse panel.