A fuel pressure test checks if the fuel system is pushing fuel at the right pressure. If pressure is wrong, the engine may lose power even if it starts, so it’s a key step in diagnosing diesel problems.
The “low side” is the part of the fuel system that feeds fuel in before it gets turned into the very high-pressure injection. Testing it helps you figure out whether the problem is fuel supply or the high-pressure injection part.
“Power Stroke” is Ford’s diesel engine line. Diesel engines need very high pressure to spray fuel correctly, and this one uses engine oil pressure to help create that high-pressure fuel for the injectors.
A solenoid is an electrically controlled switch that moves a valve. Here, it helps control oil pressure, which is then used to generate the pressure needed to inject diesel fuel.
An intensifier piston is a pressure-boosting part. It uses oil pressure to “multiply” force so the system can create the very high pressure required to inject diesel fuel properly.
A common-rail diesel system stores fuel in a shared high-pressure line (“rail”) and then sends it to the injectors when needed. The episode is comparing that style to another diesel setup that uses oil pressure to create injection pressure.
Think of the high pressure rail like a pressurized “supply line” that feeds the engine’s injectors. If the pressure in that line is wrong, the engine may run poorly or throw codes.
A PicoScope is a tool that lets you look at electrical signals as a graph. That can help diagnose sensor or wiring problems that are hard to spot with a simple meter.
Power Probe is a handheld electrical testing tool for cars. It helps you check circuits quickly without as much setup as some other tools.
Brand
Vantage Pro
Vantage Pro is another kind of car electrical testing tool. The host is basically saying there are different tools you can use instead of a plain voltmeter.
“High load, high flow” means the engine is asking for a lot of fuel and the system has to deliver it fast. Some fuel problems only appear when you’re really using the engine.
The sender assembly is the part inside the fuel tank that helps the car know how much fuel is in there. It’s also tied into how fuel is picked up from the tank, so problems there can cause fuel delivery issues.
The pickup screen is a small filter at the fuel intake in the tank. If it gets clogged or falls apart, it can restrict fuel flow and cause the engine to run poorly or fail under load.
P 1280 is a computer error code your truck sets when it thinks a sensor or system isn’t behaving right. Here, it’s related to the injection control pressure sensor reading too low.
This means the truck’s computer thinks the fuel-injection pressure sensor is reporting a value that’s too low. Since that sensor helps the engine control how much fuel gets injected, a low reading can cause weak acceleration or drivability issues.
A scan tool is like a reader that plugs into the truck to pull error codes and show live sensor numbers. It helps you see what the computer is seeing when the problem happens.
Term
Pico backproping the ECM
Backprobing is when you carefully test the wires at the connector to see the real signal going to the computer. It’s a way to measure what the ECM is actually receiving.
A breakout box is an adapter that makes it easier to test specific wires at a plug. It helps you connect test equipment without poking or damaging the connector pins.
Reman modules are used parts that have been rebuilt and tested. They can be a cheaper alternative to buying brand-new, but you want to check the warranty and testing quality.
They’re watching a voltage number while the engine is cranking. If that voltage stays steady (doesn’t sag), it suggests the car is providing the sensor with the power it needs, so the problem may be elsewhere.
A zener diode is an electronic part that helps keep a voltage from going too high or too low. They’re suggesting that if the circuit uses one, it should hold the voltage steady unless something is seriously wrong.
A key cycle is the sequence of turning the ignition off and then back on (often including the “key on, engine off” state). Many vehicle diagnostics only run after a certain number of key cycles, so a code may appear on the next cycle after the fault is detected. That’s why the host says the DTC showed up on the next key cycle.
“Key on, engine off” means the car is powered up, but the engine isn’t running. The computer can still test sensors and wiring in this state. If a sensor voltage is outside a tight range, it can set a code.
Autel is a company that makes diagnostic tools for car repair shops. They provide equipment that helps technicians find problems faster. This segment mentions Autel as the sponsor and as a maker of professional diagnostic gear.
TPMS is the system that watches your tire pressure and warns you if something’s off. After tire work, it often needs to be reset or relearned so it knows the new sensor readings.
ADOS is mentioned as an advanced safety-related system, but the transcript segment doesn’t explain what the acronym means. You’d want the earlier part of the episode (or the show notes) to know which safety feature it refers to.
Inertial rebound is when a sensor’s reading “bounces” after something hits or changes suddenly. That bounce can confuse diagnostics unless you know what to expect from the sensor.
Piezo sensors make electricity when they’re squeezed or vibrated. They can react very fast, but their signals can be confusing if you don’t know how they behave.
Pulse sensors send out a signal in repeating “blips.” When you’re diagnosing a problem, you usually care about the timing and shape of those blips, not just that a signal exists.
It’s a diagnostic term for when something seems to improve right after you replace a part, but the original problem comes back or shifts. Think of it like a temporary bounce instead of a permanent fix.
Term
neural rebound
The host is describing a “bounce back” behavior—where the readings or response don’t settle immediately and can overshoot after a change. The idea is that the system reacts in a way that makes the problem look different for a moment.
A diaphragm is a thin flexible part inside the sensor that moves when pressure changes. That movement is how the sensor turns pressure into a readable signal.
A strain gauge is a sensor that detects how much something bends or flexes. When it flexes, it changes an electrical signal so the car can measure pressure or force.
Term
capacitive
Capacitive sensors work by measuring how electrical storage changes when a part inside moves. So when pressure pushes the sensor’s internal membrane, the electrical signal changes.
Aftermarket means a replacement part that isn’t the original factory one. The host is saying aftermarket sensors can sometimes behave differently, even if they’re meant to replace the same part.
“Key off” means you turn the ignition off. Even after that, some engine computers and sensors can keep running for a short moment, so a problem can still get recorded.
OE means “original equipment,” like the same part style the car maker installed from the factory. Mechanics use OE sensors to rule out problems caused by cheaper aftermarket sensors.
A large language model is an AI that’s been trained on lots of text so it can answer questions and write responses. The host is saying people use these tools instead of looking up information the old way.
The host is saying that if you always rely on tools (like AI) to do the thinking for you, you may get worse at doing it yourself. Over time, your ability to recall and apply diagnostic steps can fade.
A diagnostic strategy is the step-by-step plan a mechanic uses to figure out what’s wrong. It’s about checking the right things in the right order, not just guessing.
LIVE
This is the Automotive Repair podcast network.
Welcome everyone to yet another episode of diagnosing the
aftermarket A to Z. I'm Matt Fonzola and ambition is a very
dangerous thing because either you achieve what you set out to
do in your life and ends prematurely or you don't and your
life is a source of constant disappointment. That and more
afterward from our sponsors, Pico Technology and Autel.
Tired of guesswork and your auto diagnostics need to pinpoint
faults quickly and accurately. PicoScope will turn your PC into
a powerful diagnostic tool. See live waveforms from sensors,
injectors and ignition systems. See the problem, solve it
validating repairs or working with ADOS and advanced safety
technologies. Autel tools are engineered to support accurate
efficient repairs and strict accordance with OEM procedures
and standards. To learn more about Autel and their diagnostic
solutions, visit Autel.com. A-U-T-E-L.com.
But something with this sensor seemed to do something that I
think the term could be wrong. I think the term is called
inertial rebound. My introduction to inertial rebound was
really with piezo sensors, particularly when I take that when
I guess I gotta be careful with piezo sensors, right? Because
we have a lot of those. I guess first look type sensors, the
pulse sensors. So piezo type pulse sensors. First look, we
have a couple of them out there made by like Joe Casey. I think
D rail might make some I could be wrong. I didn't get mine from
there. I don't think I got mine from another guy who doesn't
make them anymore. But I got a couple from Joe. Joe's electric.
They're great. They serve a purpose. But my introduction to
that phenomenon, if you will, was on the first look when they
were really the only game in town. And we would call it like
the rubber band effect and a term that I will attribute to
Aaron Kepin. It could have been Harvey Chan, but I really think
Aaron's the one that kind of termed made that term and we were
using it for the longest time. But then through studies and all
that, we find studying, I should say, we find that there's
something called a neural rebound. And on those, it's
exaggerated because of the diaphragm inside of the sensors
and the way they work, like it makes sense. But you don't think
about it, or I wouldn't normally think about it on
something like a pressure sensor of that sort, which I think the
technical term for those would be a strain gauge. I don't know
the guts of it. I don't know. I don't think they're piezo. They
might be I suspect they're capacitive. I don't know. But I
think what we're seeing is sensor rebound. The OE sensor,
the motorcraft sensor did not do that. As many times as I tried
to reproduce it, it wouldn't do it. Because it's easy. It's
right on top, right? If I can replace it, anybody can replace
it. Right on top, screw the new one. And with the motorcraft
doesn't do it. Put the aftermarket one back in, it
does it. I wish I had way more time to kind of mess around
with this and send maybe stuff to this supply, I should say
manufacturer. This Friday, people wanted to go, I get it. So
what do you do? But I thought that was kind of cool that it
was really setting on key off. As far as I could tell, I guess
I don't know if I want to necessarily say absolutely. Like
I know for sure. But I think I have data that kind of backed up
that this was setting on a key off the sensor was rebounding
from, you know, whatever idle pressure of 800 psi, I don't
think it was 1000. I mean, it could have been in certain
points, depending on when you shut it off after startup,
because usually startup can be over 1000 1500, I think, you
know, after startup, it would be over 1000 psi, maybe 1500 or
so. And then at idle, it would come back down to roughly, you
know, I think it was like 800, maybe 700. I want to say it was
800. And you shut it off, go to zero rather quickly. I don't
believe they're supposed to hold pressure. Where I think we're
seeing a inertial rebound and the PCM was awake to set the
DTC. And if you weren't looking for it and thinking about it
that way, which initially no, I wasn't, you may never see it.
And if you're not watching your scope, or even depending on the
scan tool, if it's still collecting data after key off,
some of them, once the keys off, they freeze, they're done. And
then a couple others would kind of keep going. Not long at all.
But enough to maybe capture that once in a while. It was the
scope that caught it. It was the scope. And the meter never
really saw it. Not that it was just such a fast signal, but
it's, I think it was a timing thing. And you got a few things
going on with sample rate, if you will, of the meter, and then
of course, the display itself updating, right? Because if the
meter's measuring 1000 times a second, the screen's probably
not updating 1000 times a second, right? So I put in an OE
sensor and fixed the problem. It was I thought it was kind of
cool. But like I said, I'm very easily entertained. I guess that
leaves us enough time to answer a listener question. Okay, so
it's a neat question about what do I see an effect, negative
effect of people and mainly painted talking about kids with
using artificial intelligence. Okay, so my take is that when we
say artificial intelligence in normal conversation, almost
everyone is talking about large language models. They're
talking about chat, gpt, Gemini, Claude, Grock, Meta. I feel like
I'm missing one, but those are some of the heavy hitters, right?
Yeah, it's just like using database diagnostics, right? Not
that people are necessarily always using AI for diagnostics,
but just to answer questions. The repercussion is that you
either kind of have that skill atrophy, or that area of your
brain that stored that information kind of purging it,
or you're losing how quickly you could access it, or I suppose
technically access to it period, in favor of using this
intelligent or large language model, okay, be it to write a
sentence, to write a paragraph, thought to run you
through a diagnostic strategy to have instead of reading a
bulletin from beginning to end, you just paste it in and ask
for a summary. There's, I don't want to say danger, but you
have to be aware of what's going on. So do I think this is the
fall of man, maybe the reason it's hard to get overly negative
is it doesn't have to, we can choose for it to not to and that
could be us individually, us as families, us as businesses, us
as industries, our professions, and then really honestly us as a
society, deciding that we have to think about evidence based
across the board. That's how we have to default evidence based
and then question, what are the long term effects some things
we can't know, but then why not proceed cautiously. But at some
degree, I appreciate the freedom families have parents,
particularly with their kids. I have to question sometimes if
they're aware, like how do I assure all the parents quite know
what they're up against with artificial intelligence? Because
let's be honest, they know about social media. Are they
restricting access of their kids to social media? And the
answer is no. And we know the dangers. We know them. We know
the repercussions and it's not good. So yeah, I think as a
society, we have to agree sometimes, what is best and
evidence based, right? So then if we do kind of go down certain
roads, there's a reason to do it and it's backed by data and
it's hard to overly dispute to a degree, right? Everything's the
nature of science is to question everything. That's, that's
just what it is. But on the flip side, then you better
question yourself too, right? So all that's to say, I want to
see at least discussions, maybe not necessarily regulations, but
I want to see really good discussions and somehow getting
the word out to everybody about the fire they're playing with.
And they should have with social media and they didn't. And the
information is relatively easy access to about the repercussions
of social media in children and young, young adults. So we
should be having hard conversations about that. And
thinking about what to do about it. Because I don't know if
just letting things fly and see what happened. I don't know
trusting that things will work out. Then why aren't we doing
that in our shops? You know, we do stuff to affect certain
changes. So that's so different. I think luckily, some heavy
hitters with the large language models have kind of gotten
together at least to help work together for AI detectors.
Because some of the crop of AI detectors out there are god
awful. It was nice to see that. But the reason I have a rough
time, and maybe I'm taken too long to get to this point, I
don't know. The reason I don't find it to be all doom and
gloom is we can make choices to avoid certain paths. And there's
something called neuro plasticity. And you've probably
heard me bring it up before, especially if you've listened
to episodes where I'm talking about substance use disorder
and maybe some psychological stuff. Margaret and I may have
talked a little bit about it, but I've talked quite a bit about
neuro plasticity either on this podcast or some of the other
early ones with calm talking about substance use disorder or
addiction. And neuro plasticity, it basically allows for us
that ability our brains allow us to rewrite things. So the
dependence on social media, really what we're talking about
the dependence on large language models, that can be
changed, like we have people or us or kids or whatever that
lean heavily on it. And it becomes a crutch, they lose
confidence in their ability to do something now that they have
this resource available to them that they now have to do it
themselves and should have to do things themselves at least
once in a while or take more control about what's going on
that we can rewrite our brains, we can recover from this type of
thing. It's just how far are we willing to let it go? That's my
answer. It can't be all doom and gloom. I'm not confident that
enough people in the right places and that could be at the
kitchen table or in the household are going to do and maybe
sounds judgy, the right thing, but start asking real questions.
I don't know where we kind of lost that confidence to make
those decisions or to start asking the hard questions about
the effects. It seemed to happen a lot with video games and
television and all that. And now, I don't know what's really
question. Music too. Oh my god, music. If you listen, parents
are very much against you listening to certain things. Maybe
that was it when they were falling to be barking up the
wrong trees with video games making more violent. It was
hard to find data to back that up. Watching television made you
stupid or it was hard to find stuff to make to back that up.
Although now I'm wondering, especially with me watching too
much, maybe at least I remember what I watch, I guess. I don't
know if I watch so much, just I remember it. And then, yeah, the
music with, what was it, the PMRC that it just ended up not
holding a whole lot of water and you had heroes, I guess, with
of all people, John Denver and D Schneider going in and amongst
other artists pushing them back against Congress and the data
didn't support. I don't know if there's so much Congress, maybe
not Congress, but at least a committee. The data didn't hold.
But we have data about social media. It's not good. We have
data about large language models. It's not good. So, I guess, my
first answer is my last answer. Evidence based. Those are
terms I would like to hear more and more of us refer to. Whatever
that is, if I'm presenting you an idea, and I think it's right,
what's my data? And sometimes it's an opinion based off of a
bunch of different pieces of data and experiences. But a lot of
times I see these arguments where no data, there is no data
presented to anyone. Or somebody brings about some data and
rather than counter it with data, we just try to undermine the
sources of the data or laugh the person off. And depending on
your political leanings and who the different people involved
in the conversation or what the content is or the subject matter
is dictates your behavior is acceptable for the given argument.
Meaning, I don't know, I see right leaning or more conservative
people doing the same thing more left leaning or liberal people
do. And depending on the subject matter, that's what seems to
make it okay. So, I don't know. Oh man, that's a circle, right?
They think they're further apart, but they end up going around
the other end and they're closer together. I don't know.
Anyways, that is my answer. Evidence based. And yeah, be
careful with it. Neuroplasticity can offer the path back or path
of recovery and gain those skills back. And we could be all
right. And then yeah, just start thinking in terms of evidence
based. What can I back this stuff up with? And maybe someone
else's data or you have to start gathering your own data and
understand that your data probably has biases and therefore
the nature of science is to debunk your data and go about and
find data that conflicts with your data using the same methods
you are. And then coming up with other methods to get new data
better data. And then you can use that to find better presumably
better data. That's how that works. And we can use it. We
don't have to be scientists if we can use these tools. The
scientific method works for everything. Or at least almost
everything. We can use it. That is how I'll leave you. I'll
skip Mount Rushmore today. Maybe next episode. A few people
asking different things about different Mount Rushmore seems to
be popular. If you do like it, let me know. I mean, I want to
know that this stuff interests you and what
make sure to leave a comment. If you're watching this on
YouTube, please don't forget to give the video a like. It really
does help. And if you want to not miss future episodes, click
that subscribe button. Thank you to our sponsors, Autel and
Pico Technology. And thank you to the Auto Repair Podcast
Network. And until next time, take care.
Follow Matt on your favorite listening app. He's very
interested in what you have to say. Let him know what you like
him to cover and come on the show. Matt is all for advancing
the aftermarket. Find Matt Fonslow on social media and
connect or on aftermarketradionetwork.com
About this episode
A 7.3L Power Stroke comes in with an intermittent ICP/injection control pressure sensor fault that only shows up after startup. The diagnosis starts with confirming low-side fuel pressure, then stresses testing under load and using a scope to catch fast “rebound” behavior after key-off. The host walks through ICP circuit checks (5V reference, ground, signal) and why an aftermarket sensor can behave differently than OE/Motorcraft. The episode then pivots to an AI “skill atrophy” debate, arguing for evidence-based thinking and cautious, choice-driven use of large language models.
In this episode of Diagnosing the Aftermarket A to Z, Matt Fanslow starts with a case study on a 2000 Ford Excursion with a 7.3L Power Stroke that came in with a severe lack of power. What began as a fuel delivery problem turned into a more interesting diagnostic puzzle involving a P1280 injection control pressure sensor code, an aftermarket ICP sensor, and a fault that appeared to happen not during cranking or running, but after key-off.
The case becomes a practical example of why testing under actual operating conditions matters. Static tests can pass, scan data can look normal, and a meter may never show the event. In this case, the scope captured something the rest of the tools could easily miss.
Matt then shifts into a listener question about artificial intelligence, especially large language models like ChatGPT, Gemini, Claude, Grok, and others. The concern is not simply whether AI is good or bad, but whether people, especially kids, may begin outsourcing too much thinking, writing, problem-solving, and confidence to these tools.
The episode lands on a recurring theme: tools are not the problem by themselves. The question is how they are used, what skills may atrophy when they are overused, and whether individuals, families, businesses, industries, and society are willing to think more carefully and evidence-first about the long-term effects.
Topics Covered
2000 Ford Excursion 7.3L Power Stroke Case Study
Matt discusses a 7.3L Power Stroke with a major lack-of-power concern. The initial issue was tied to the low-pressure fuel supply side. A basic fuel pressure test had passed, but it had not been tested under the actual conditions where the symptom occurred.
That led to a larger point: testing at idle, key-on, or during cranking is not the same as testing under load.
The vehicle ultimately had damaged pickup screens and debris in the fuel tank. Cleaning the tank and replacing the failed components solved the fuel supply issue, but another problem remained.
The P1280 ICP Sensor Problem
After the fuel supply repair, the truck began setting a P1280 code, related to the injection control pressure sensor circuit reading low.
The strange part was that the voltage looked normal:
Key-on engine-off voltage looked normal.
Cranking voltage did not drop.
Five-volt reference stayed stable.
Sensor ground stayed stable.
The signal looked fine while running.
The meter never really caught the failure.
The scope eventually revealed the important event: after key-off, the ICP signal voltage would sometimes dip nearly to zero. That appeared to be enough for the PCM to set the code on the next key cycle.
Scope vs. Meter vs. Scan Tool
This case is a useful reminder that each tool has limits.
A scan tool may not refresh fast enough or may stop collecting data after key-off. A meter may technically sample quickly, but the display may not show a short event clearly. The oscilloscope was the tool that made the failure visible.
The point is not that one tool is always better. The point is knowing which tool fits the question being asked.
Aftermarket Sensor Behavior
The aftermarket ICP sensor appeared to produce a rebound effect after shutdown. Matt compares the behavior to something like inertial rebound, a concept familiar from certain pressure pulse sensors and piezo-style tools.
The OE Motorcraft sensor did not reproduce the same behavior. Replacing the aftermarket sensor with the OE sensor corrected the issue.
Main Diagnostic Takeaways
Testing has to match the failure condition. A fuel pressure test that passes at idle does not prove the fuel system can keep up under load.
A sensor can create a failure outside the moment most people are watching. In this case, the important event seemed to occur after key-off.
A scope can reveal details that a scan tool or meter may miss, especially when the event is brief, intermittent, or happens outside normal observation windows.
Aftermarket parts can fail in ways that are not obvious or traditionally “failed.” The sensor was not simply dead. It behaved incorrectly under a specific condition.
Listener Question: AI, Kids, and Skill Atrophy
The second half of the episode deals with a listener question about artificial intelligence and whether it may negatively affect people, especially kids.
Matt narrows the discussion to what most people mean by AI in everyday conversation: large language models. These tools can write, summarize, explain, organize, and suggest. That makes them useful, but it also creates a risk.
The concern is skill atrophy. When people lean too heavily on AI to write, think, summarize, diagnose, or explain, they may lose confidence or sharpness in those same areas.
AI Takeaways
AI does not have to be viewed as the collapse of civilization, but it also should not be treated casually.
The better question is not “Is AI good or bad?” It is: What happens when people outsource too much of their thinking to it?
Matt argues for an evidence-based approach. Families, schools, businesses, industries, and society should be asking harder questions about long-term effects instead of simply letting the tools spread and hoping it all works out.
Neuroplasticity matters here. Even if people become overdependent on AI, the brain can adapt again. Skills can be rebuilt. Confidence can be recovered. But that depends on whether people recognize the problem and choose to do something about it.
Thanks to our Partner, Pico Technology
Are you chasing elusive automotive problems? Pico Technology empowers you to see what's really happening. Their PicoScope oscilloscopes transform your diagnostic capabilities. Visit PicoAuto.com
Thanks to our Partner, Autel
From drivability diagnostics and TPMS service to ADAS and advanced safety systems, Autel helps technicians follow OEM procedures and repair with confidence. Learn more at Autel.com
