Voltage Supply Problem @ ECU (highly technical)

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
LARGE PROBLEM (88 supra yellow plug)

I have a voltage drop at the ECU on all voltage supplying pins. Pins +B, IG S/W, B1, and BATT all see around 11.6 volts with the car not running. Battery Voltage at this point is 12.8 with a fresh charge on an optima. The engine harness is new, and fuse box seems to be in good shape. At the EFI fuse, I see around 11.9 volts. (also at the main relay).

Resistance from the ECU grounds to a chassis ground is less that one ohm, so it doesn't seem to be a grounding issue at E01 or E02 on the ECU plugs. I do have a battery relocation, but am using ALT gauge wire from battery to block and starter. From the starter terminal to the OEM wiring, I am using 6 gauge(the cluster of wires that goes from the fuse box to the battery.). Voltage at that point is around 12.77.

Even the voltmeter while driving shows 12 volts even. Battery voltage at this point is around 13.7-14.0 depending on RPM. While seeing 13.7 on a multimeter at battery, my AEM is recording 12.0-11.9 volts input(and this is verified.) At the point I reach my 5500rpm ignition breakup, I record 9.95-10.0 volts at the AEM(both with a meter and the AEM logging software) The voltage rises and falls as the rpm bounces, the oddball sine waves follow each other on the datalogs. As voltages startes to come back up with rpm, it reaches around 10.6 volts and the rpm falls again.



I have come to the conclusion that this is why my car refuses to rev past 5500 rpm with AEM. This is only after I changed EVERYTHING else in the ignition system with multiple new and other parts...lol


Any ideas?
 

IJ.

Grumpy Old Man
Mar 30, 2005
38,728
0
0
62
I come from a land down under
Chris: Where are you picking up the supply to the Coils?
(as RPM/Boost goes up so does the load on the Coils if you have a bottleneck in the circuit you'll get spark blowout around Tq peak as this is where the load is highest)
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
Stock...lol I am using stock coils, ignitor, and wiring.... which could be the problem...

I am not sure the coil wiring is the problem. I just checked with the meter, and I've got 12.6 to the coils with the car off, the volt drop doesn't seem to affect them. However, I have my thoughts that the AEM may be unable to send a properly multiplexed signal to the ignitor at 10 volts...

Although..... peak torque is right around this area(I'm only at 13 psi at the moment), but to contrast that.... it has the same problem in neutral with no load.

I'm glad you saw this thread Ian..
 

IJ.

Grumpy Old Man
Mar 30, 2005
38,728
0
0
62
I come from a land down under
I'm guessing under load the 15+ year old wiring/connections are dragging it down enough that the AEM is only seeing 10v and you're right it WON'T be happy that low (MoTeC won't sync down there)

I gave the ECU it's own dedicated 5mm cable from the battery ign controlled through a relay just in case.
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
I may do that.... it's hard to find high quality relays around here though

I'll probably go order a new toyota relay and use that with 8 gauge wire to the ECU power.

Ian, you would happen to know which ECU pin has the most power draw would you? The main current source....for stock I guess(since I have a PNP ECU). Are +B and +B1 inputs or outputs? The TRSM list those two as "EFI Main Relay" so I would assume outputs. If I am correct that means BATT and IGN Switch are the two power inputs for the stock ECU. BATT seems to have constant power(I'll take it that one goes to the EFI fuse under the hood in the box) and IGN SWTCH goes to the switch.

If that is the case, I should run one 8-10 gauge wire directly from battery power to BATT, and use a relay with a heavier gauge wire to IGN SWITCH with the stock wiring triggering the relay...

Someone correct me if I am off base here.
 

IJ.

Grumpy Old Man
Mar 30, 2005
38,728
0
0
62
I come from a land down under
Chris: Sorry I don't use any of the stock wiring to carry loads I just use it all to switch relays this way I know my own wiring is more than adequate for the job.

It makes trouble shooting so much easier.
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
Yep... thats what I'm thinking about doing. I'm just wanting to switch a relay with the stock IGN SWITCH terminal of the ECU wiring. I want to use two eight gauge wires to power the ECU....lol

I need JJ in here...lol
 

Supracentral

Active Member
Mar 30, 2005
10,542
10
36
I do the same thing. Even if I install a fuel pump I turn the factory wires into triggers and run dedicated lines from the battery.
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
I'm mainly inquiring about the power sources now for the stock ECU, is it only two? BATT and IG S/W ? or are +B and +B1 also power sources?

I'm not sure if the +B's turn are power in from the EFI main relay or power out to it. I don't have a diagram handy really...

I know the BATT and IG S/W terminals(the two on the top and end of the large plug(pre 89) at the ECU.
 

jetjock

creepy-ass cracka
Jul 11, 2005
9,439
0
0
Redacted per Title 18 USC Section 798
Meh, highly technical? Cakewalk. The location of the loss can be easily found by doing voltage drop testing using the millivolt range on your meter. Same with the grounds. You can't use ohms on them. Regarding the other questions BATT is keep alive power, +B and +B1 are switched power from the main relay, IGSW is wake up power, and M-Rel is the sunken output to the Main Relay.
 

suprahero

naughty by nature
Staff member
Aug 26, 2005
14,971
0
36
54
Roll Tide
Jetjock, is that in greek because I didn't understand a damn word of it..........:biglaugh:

That's exactly what I would have told him though................if I spoke greek.
 

Nick M

Black Rifles Matter
Sep 9, 2005
8,898
40
48
U.S.
www.ebay.com
suprahero said:
Jetjock, is that in greek because I didn't understand a damn word of it..........:biglaugh:

That's exactly what I would have told him though................if I spoke greek.

It is car mechanic electrical class 101 actually. This is what he would be teaching his students after they understand the very basics. Yep, cakewalk for the electrical engineer, no doubt.
 

jetjock

creepy-ass cracka
Jul 11, 2005
9,439
0
0
Redacted per Title 18 USC Section 798
Sorry, I don't speak Greek. Only English and Russian ;)

He comes across as somewhat electrically competent so I didn't dumb it down. I'm bored this morning so here's some light reading:

Voltage drop testing is a technique used to find very small resistances, especially when the circuit is under load. It's nothing more than Ohm's Law in action. It involves using the meter on the millivolt scale and measuring points along the same conductor rather than from one conductor to another (like from +12 to ground) as is typically done when measuring voltage. Any resistance between the two measured points will cause current to flow through the meter. The more resistance the higher the voltage will be.

Anything more than 100-200 mv needs to be corrected. Conversely, the lower the voltage the better the connection is. Perfect connections would have zero volts across them but it's near impossible to do that without using special termination techniques and materials. Gold comes to mind because it doesn't oxidize. For those that don't know the contacts in your ECU are gold flashed for this reason. The connector pins and receptacles (called a "pulselock" type by Toyota) are especially made for this purpose and of very high quality.

Anyway, lets take the main ground as an example. Set your meter to DC millivolts and place some load on the circuit ie; start the engine and turn on the headlights or heater blower. Place one meter probe (doesn't matter which) on the battery negative post. On the post mind you, not the battery clamp. Place the other probe on the engine block. Ignore polarity indication. If you get more than 200 mv work your way back along the main ground. When you find the location of the loss the voltage will decrease.

You can also start at the same point and move forward along the circuit path. For example place both probes on the battery negative post. You should get zero millivolts or very close to zero. Leaving one probe on the post move the other to the negative clamp then (if possible) to the wire exiting the clamp. Follow along to the next accessible point ie; the negative terminal at the block, then the head of the bolt connecting the terminal to the block, then the block itself. As you move along the voltage will increase in proportion to the amount of resistance found in the path. Anything over 100-200 mv should be corrected by cleaning, re-crimping, replacement, etc.

Course, once you find a "bad" connection and repair it you can place the meter probes on either side of it. You can do this from the git-go too. Say you find 500 mv between the negative cable terminal and the head of the bolt on the block, indicating unacceptable resistance in that connection. Remove and clean the surfaces then check it again by measuring directly across those two points. If you can't get the voltage below 200 mv the problem might be a bad crimp or corrosion in the wire/terminal interface itself. Check by measuring between the terminal and wire entering it if it's visible or by piercing the insulation if it's not.

Your goal is to make the voltage as low as possible or at least below 200 mv when measured between two points, including the entire circuit path ie; from the battery negative post to the block. I'm very anal about electrical stuff but during my last check I found 700 mv on my main ground. The problem was the interface between the negative terminal and the bolt on the block. Keep in mind the electrical system exhibited zero symptoms with that amount of loss. It's still unacceptable though because it would have gotten worse with time until symptoms did devolop. Long before the car would refuse to start the connection would have begun heating. Cleaning reduced the drop to 50 mv. I've yet to work on a car (any car) more than a few years old that didn't have excessive drop in it's main battery/charging circuit.

The technique works on any path. In his case (losses at the ECU connector) extend one meter probe using wire. Place the long probe on the battery positive and the other on the clamp. If that connection checks out you can move the positive post probe to the clamp so as to make it easier to attach. Take the other probe (along with the meter) and check any of the + connections at the ECU. If you find more than 200 mv work backwards.

Alternately, you can check across every connection using both probes. That way you won't have to extend one. It's best to test total drop from one end of the circuit to the other in the beginning though. That way you'll know right away if there's excessive drop. Needless to say doing drop testing requires accessing every point along the circuit where the conductor is "broken" by some device. Relays and their sockets, relay contacts, switches, terminal block, crimps, on and on. That's why it's best to measure total drop first. If it's OK you can move on to another circuit.

Another technique is to simply take a long jumper and bypass the path to see if the symptom you're having disappears. The method is crude however and doesn't pinpoint the fault. It's quick and dirty though.

The key things to remember when doing voltage drop testing are:

1) Make sure meter connections are electrically solid because any resistance in them will effect measurement. Don't be probing through dirt, oil, or grime for example. Make sure the meter probes are clean too. Scotchbrite them if need be, a good thing to do on occasion anyway.

2) The circuit must be under some load. The more the better. You can't do voltage drop testing without current flowing. When dealing with main power paths like the battery cables turn on the lights, blower, etc. With circuits like the ECU power, instruments, lighting, etc, the load they draw on their own will be enough.

3) Think about what you're doing. Be sure to probe every connection, say a relay's pin and it's socket. Any such "interruption" along a wire is a potential source of resistance and thus drop. Think in terms of moving the probes as close together as possible while still remaining on either side of a connection while it's under load. It's similar to how you'd do resistance testing except drop testing is capable of finding very small resistances your ohmmeter can't. (It takes a special type of ohmmeter to measure very low resistances).

If you find the drop across a wire itself (by probing or piercing it's insulation at either end) is more than 100 mv the wire is either bad or too small for the job. With stock wiring that's unlikely (even when old) but still possible. Put another way excessive drops (more than 100 mv) are almost always a connection problem along the conductor and not the conductor itself.

Voltage drop is a very useful form of electrical testing. Google the term for more details. The Net is more than a global pornography network you know ;)
 

p5150

ASE and FAA A&P Certified
Mar 31, 2005
1,176
0
36
Central Idaho
What kind of grounding are you using? For my AEM seup I have thick bonding and grounding cables on everyting. Chassis-block, transmission-chassis, battery-block, battery-chassis.

Description of your problem sounds like a grounding issue possibly coupled with an RF emission from your spark plugs....

Also, are you using resistor-type plugs?
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
Excellent post JJ, I knew you would put me on the correct path.

Before reading this post, I added a second connection from the ALT charging post, (fused) directly to the fuse box supply. This increased the voltage everywhere, including at the aem. At idle, I now have around 14.65 volts at the AEM, and everywhere else in the car also. I did not have that before. The car seems to run much smoother, and doesn't go lean upon turning the headlamps on anymore.(appently the battery offset on my injectors isn't quite right.) But I still have the same issue I had before.

I have been through some CC electronics course, and I should have known about that, but didn't think to do it.

P5150, I've got resistor plugs in it, but I'm still having severe ignition breakup at 5500 rpm with load and 6000ish no load. The higher the gear, the more it sounds like a rev limiter instead of a misfire...

I really want to get this solved and get the car tuned.
 

p5150

ASE and FAA A&P Certified
Mar 31, 2005
1,176
0
36
Central Idaho
I have a 2j COP setup so Im not sure about settings for your car as far as dwell goes. I think that the 2jz ignitor controlls dwell automatically regardless of the settings on the AEM. I dont know what the 7m ignitor does though.

What kind of plug gap are you running? Did you try to drop the gap down a bit?
 

gixxer750

2jzget comingsoon!
Mar 30, 2005
2,333
0
0
Mississippi
running .019 gap.

I've tried every dwell setting from 11 to 24 with no change at all.

I've sent the box out to aem for diagnostics, they replaced the board with no change. I've replaced the coils, ignitor, and CPS with brand new units. I went back to brand new stock plug wires from toyota. I replaced the wiring harness, and fixed the voltage drop problem.

The car has been to two seperate professional tuners, both were baffled and one was an AEM cert tuner. I've called AEM a thousand times and nothing has come of it.

I am almost ready to sell the AEM and buy a motec, but I'm not sure if that would help anything either. I'm thinking about ditching the AEM altogether and going back to stock ECU and emanage. I just want the car to rev to stock redline and I would be happy.

I'm at a complete loss to explain this. Even now, after datalogging the rev to 5.5k rpm, the voltage drop is only to around 12.3 where it was at around 10.0 volts. So I'm guessing that wasn't the problem.