Sale ends todayGet 30% off any course (excluding packages)
Ends in --- --- ---
Discussion and questions related to the course Motorsport Wiring Fundamentals
Looking for a scope trace to compare to for an engine I am trying to diagnose a starting issue with. See the attached files for a scope of the engine cranking and another of it running.
Engine is an LS7, supercharged, IGN-1A coils, big fuel system etc on Holley Terminator X Max.
Car had some poor grounding that I resolved but still has trouble starting consistently. Usually the first turn of the key it just cranks and cranks, flick the key off and back to cranking and it catches and runs like nothing is wrong. Based on my scope findings the ecu is receiving both crank and cam signals through both scenarios, starting and not starting. I'll also toss a data log screenshot (cant upload a datalog to the forum) in showing 3 tries before starting, just note the straight lines are where the laptop lost connection to the ecu during a power cycle.
Holley are convinced its a voltage problem, the battery dips to 8V for a few ms then steady at 10.6V during cranking, but the ecu just rejects seeing a crank position signal on the first few tries.
I am not sure if there is still an unresolved grounding issue, or an issue with the ECU itself. Upon further reflection the crank signal voltage changes quite a lot based on engine speed, more like a VR sensor then a hall. Especially when compared to the rock solid consistent cam sensor signal.
Holley claims that the ecu requires 5V to "see" an edge, I kinda doubt its that high for a 5V supplied hall sensor but I'm not them.
So If anyone has a 58x LS engine scope trace either from a scope or an ecu scope long to compare to mine that would be most helpful.
Here is an input capture from a MoTeC with an LS7 Cranking. Looks like the same basic pattern as your scope trace, but with a bit less noise.
Looks like their hall sensors don't quite go to 5V even when fed 5V.
I had suspected that about the voltage, interesting the cam sensor voltage is also higher then the crank. Any idea what RPM that is from?
This car has OE GM ACdelco sensors, which were replaced in an attempt to solve this issue. Odd to me the voltage changes with engine speed, didn't think that was a hall effect characteristic...
cranking with OEM starter motor, plugs in I believe, since I see the pulses speed-up/slow down.
RPM is about 218, Figured from the info in the pic I provided, one revolution in (1.1 - .825 = .275 sec), so 1 / .275 = 3.6 Rev/Sec x 60 = 218 Rev/Min.
The OEM might be using 12V for the sensors (I seem to remember there was some debate about this when I was wiring the car, I chose to use 5V since it was regulated (and I had a good result, that's why I did that capture).
I don't recall seeing any 5V feed Hall Effect sensors reading to 5V, they have all been 0V to 4V square wave signals.
A friend has posed the theory that the weak crank signal seen in the scope is the result of the sensor not having power and the small voltage seen is back fed from the pull up in the ecu.
The harness is a Holley supplied one, Holley tech told me these should be 5V sensors and wired for 5V but I wonder if there was a mistake made either in construction or by the shop that I am diagnosing this for that resulted in the crank sensor being fed by the same source as the fuel pump. I have taken Holley harnesses apart before and they feed pretty much everything except for the ecu from that.
I think the reason it catches with a quick key cycle is that the fuel pump is still priming, therefore the sensor has power.
Does anyone know what a 5V sensor does when it receives 12V? I feel this also explains why the voltage with the engine running is so much higher then the cam sensor.
My research in the past gave me that the 58x and 4x sensors (LS2/LS7) are 5V factory, and the older Gen3 LS1/LS6 24x and 1x are 12V but I have wired both with 5V and had no problems getting clean signals before. My guess is the sensor doesn't care, it'll spit out nearly its input voltage regardless of it being 12 or 5, maybe so long as its above 3.3 or whatever the sensor internally operates at.
also thank you David, haha I guess I could have done that math for myself but didn't think of it.
From what you said, I suspect it may be too much voltage drop in the wiring to AND /OR from the ecu to the battery, leading to the ECU not operating correctly, and/or too much voltage drop across the coils for them to operate reliably under cranking.
Did you measure the battery cranking voltage at the battery with a volt-meter, or is it from the display/logs? If the former, where are you taking the power for the ECU from, and what are you doing to ground/earth it? between the lower battery voltage during cranking, and the wiring voltage drops in the ECU's power and ground, it could be marginal?
It may not be the primary problem with starting, but those coils can draw a lot of current - up to ~15A each - and will need good power and ground wiring to maximise voltage across them, have you accounted for that?
As a work-around, can you increase the prime pulse time? Eventually, I think you will need to rewire the harness to provide 12v ECU power to the sensors, so it can detect cranking.
Most Hall Effect sensors can work with a wide range of voltages, so if they have 12V at the connector when it's running (there's no way a 5V supply would create an 8V+ signal), just make sure it's on when you need it.
There's definitely a drop during cranking, 8.1 ish for the first few ms then 10.6 once the engines turning. These are taken from logs, I did measure with a multimeter but the log data is faster and what the ECU is seeing seems more relevant. Meter and ecu were reasonably close together anyway.
Per Holley's instructions the ECU's big power feed has 12AWG ran for power and ground all the way to the battery.
Car has starter power direct to battery, then then rest of the vehicles electrical is jumped off the starter post.
Ground from battery goes to frame, then frame to block. I am getting them to weld a nut to ensure clean grounds to the frame and maybe we'll add a 2awg from the battery to the block, as the rest of it is 4 awg. There's things I would have done differently had they called me earlier lol. I had measured a 0.5V drop under cranking from block ground to battery negative, I want to improve that.
As for the IGN-1A coils, the ecu isn't even giving a trigger when the happens because it doesn't seem crank signal. When it does start they are perfectly happy. Also I don't think they would be drawing anywhere near 15A rms when under cranking conditions, that's more of a high dwell time high boost scenario, but I don't have physical data to back that claim.
Yeah I was going to do that today, but they were unable to get a hoist free today so maybe tomorrow I'll try that. Once we solve this I'll definitely be trying to figure out who and where made the wiring error if there is one.
My thoughts as well with the voltage output, really doesn't seem like its on the 5V regulated supply. I'm quite surprised Holley's tech support didn't catch that, as the tech claimed he passed it up the chain and they told him battery voltage. They also told him the trigger edge threshold is 5V which I also don't believe lol.
Was really hoping to deliver answers today for ya'll and myself. Hate being so close to a solution and unable to confirm it.
Got a chance to go back and finish this today. Still perplexed at the why but have it starting up on the first key cycle in roughly under 2 seconds so they are happy.
First screen shot is from before, same log as the OP but showing the measurement at the lowest point. The big dip (8.2V) is the starting going from still to turning and it levels off at 10.6V. does the same on the second try on the log but successfully starts. Not seeing a big difference there.
Both sensors are fed from the same 5V power supply pin, crimped near the ecu by Holley. Not sure why the crank sensor seems to lost power during a voltage dip vs the cam sensor being totally fine. Last screenshot is what cranking looks like after the fix, much more square and stable like the cam sensor is.
2nd screen shot is a log taken from today after I fixed a mistake I made. I thought the ecu's power cables were run to the battery terminals, but got them mixed up with a vac pump. swapped them and the ECU is much happier and starts every time. Log shows much less voltage dip through cranking at least initially.
So apparently the Terminator is very very picky about voltage. I don't understand why it seems to reject crank signal even after the voltage is back to 10.6 V, I'm not an electrical engineer but something different about the sensors and current available from the 5V supply maybe? Other ECU brands I've had shut off during first dip under cranking and come back on and start then engine after.
I kinda see why they favor grounding to the battery, I still don't like it but even with poor grounding else where it still starts the engine.
I also don't understand why the crank signal is higher then 5V and the cam is not. under cranking they are the same but engine running its much higher. I did the same tests today with my scope and even swapped probes and channels and it all came looking the same...
I'll mark this resolved in a few day in case anyone has anything to add or ask questions.
For anyone finding this in the future, Holly Terminator very voltage picky. make your log look like the "fixed screenshot" lol