802SHO 2010 Build

[Texas Marauder]

I think Chuck is confused.

Each “G10x” ground reference (G102, G103, G104) shows where the circuit terminates to the chassis — not how the harness splices together upstream.

Because there is only S116 to G102.

G103 and G104 are the EPS power-steering module grounds on the subframe/rack.

G103 is a chassis ground at the front left of the engine compartment. G104 appears to be a ground at the firewall or engine block.

G102 is the PCM/TCM logic ground at the transmission case.

G102 appears to be at the firewall near the PCM.

I could go on.

 

[802SHO]

I think Chuck is confused.

Each “G10x” ground reference (G102, G103, G104) shows where the circuit terminates to the chassis — not how the harness splices together upstream.

Because there is only S116 to G102.

G103 and G104 are the EPS power-steering module grounds on the subframe/rack.

G103 is a chassis ground at the front left of the engine compartment. G104 appears to be a ground at the firewall or engine block.

G102 is the PCM/TCM logic ground at the transmission case.

G102 appears to be at the firewall near the PCM.

I could go on.

This is from Chuck. He has your schematic, the entire service manual PDF, all my findings and all my firsthand experiences of what I see and feel.

Here’s my technical take, Texas Marauder — speaking for myself.

You’re right about what the “G10x” labels are in Ford docs: they name the physical terminations. Where I’m pushing the discussion is how those nodes behave electrically under load given Ford’s splice topology and the car’s recent changes (painted subframe/rack, relocated hardware, added/deleted loads).

What the book shows (and why it matters)

PCM/TCM logic grounds (BK-YE, GD113) land at G102.

• EPS/PSCM grounds (BK-VT, GD106) land at G103/G104 on the rack/subframe.
• Those aren’t isolated “islands”: upstream, Ford ties multiple module grounds through harness splices (e.g., S116/S123/S125 chains) before they fan out to different G-points. That’s standard Ford practice to minimize harness copper and keep modules on a common 0 V reference.

So on paper: different lugs. In practice: a shared ground plane via the splice—and that plane will float if any high-current member (EPS) gets series resistance on its return.

Why this car was vulnerable

The front subframe, EPS rack housing, and a number of ground pads were stripped/painted during the build. Several grounds were later found loose or sitting on paint/oxidation (photos show the witness marks on the eyelet and tab).

• The EPS rack ground in particular was “too clean/pretty” (painted under the pad) and later confirmed loose. That’s the highest pulsed current device forward of the firewall besides the starter.
• The car also had an add-a-fuse previously on F46 (15 A Vehicle Power 2/3 to PCM) and the ACC/hood sensor sub-harnesses were discovered swapped and then corrected. All PCM-feed fuses (F37 10/15 A, F46 15 A, F47 20 A) and relays 53 (PCM) and 65 (Run/Start) are now new OE.

Observed, repeatable symptoms (load/temperature dependent)

After warm-up or with steering effort, a distinct 200–400 Hz “buzz” localized at the trans area → classic PWM solenoid chatter at undervoltage.

• Then loss of gear indication in Manual, free-rev in D/R, and recovery only after an ignition cycle.
• On the road, before total dropout: muffled exhaust/soft throttle → PCM torque path is alive but trims/limits are skewed; after restart the note is crisp again.
• Historically, during the first P0689 event (Stowe→Essex run), the car also showed EPS counter-steer/Advancetrac flicker mid-corner. That’s contemporaneous with the logic loss at the trans.
• With the car on the lift, I can make it engage, hum, then drop while F37/F46 stay at charging voltage (≈14.5 V) measured at the BJB posts. That strongly suggests feed is present but the module’s “local ground” is moving, i.e., reference shift, not feed collapse.

How a “separate” EPS ground can sink the powertrain

If the rack’s return path (G103/G104) picks up 100–300 mΩ from paint/oxidation/looseness, EPS current (tens of amps transient) creates a hundreds of millivolts to multi-volt drop at that ground node. Because the powertrain grounds share the upstream splice, that drop propagates back into the common reference, lifting the PCM/TCM “ground” relative to the battery negative. The BJB still shows 14.5 V to the body stud, but the modules don’t see 14.5 V to their ground.

Effects you’d expect (and we see):

P0689 (PCM power relay sense low) can set when the PCM’s internal sense sees VPWR–GND sag, even if the BJB looks fine.

• TCM/solenoid buzz as VPWR-to-local-GND dips into the region where EPC/SSA/SSB PWM can’t maintain force.
• CAN/advancetrac weirdness because the PSCM and PCM lose a common 0 V reference; the PSCM can still steer (different load/threshold) while the TCM gives up first.
• “Ignition-cycle revives it” because thermal/strain resets the marginal joint for a short window until it re-heats under load.

This exactly matches the sequence on the lift: no hum in Park, hum when Drive engages (solenoids energized), dropout, silence.


Why I’m confident in root cause

Grounds under the BJB and the passenger-side block tab were corrected (cleaned to bright metal, tightened) with partial improvement; the rack ground remained a wild card until identified.

• Swapping relays 53/65 and refreshing F37/F46/F47 removed the easy feed suspects.
• BJB voltages stay solid during the failure; the failure is module-local.
• The EPS-related incidents (counter-steer flicker at highway speed) co-occur with P0689 history.
• The physical evidence (loose eyelet, oxide ring imprint) is consistent with a heating variable resistor at the ground pad.

What I’m doing to nail it shut

Remove paint to bare metal at the EPS rack ground and subframe land; add star washers and torque properly.

1. Add braided supplemental grounds:
   • EPS rack → subframe/chassis (short as practical),
   • Engine block/bellhousing → chassis near battery negative,
   • Transmission case → subframe.
     These keep the entire powertrain/EPS network isopotential and give the splices an honest 0 V to live on.
2. Verification spec: with the system warmed and in Drive on the lift, VPWR-to-module-GND ripple <50 mV, battery-negative to module case <50 mV. If I turn the wheel to load the PSCM, the delta should remain <100 mV.
3. If any doubt remains, back-probe the TCM connector (VPWR and the two primary grounds) and compare to BJB while inducing EPS load; the difference is the story.

So yes — G102, G103, G104 are distinct lugs. But given Ford’s upstream splices and the very real, very painted return paths on this car, a sick EPS ground will lift the common reference and make the TCM/solenoids the first domino. That’s exactly what the measurements and symptoms show.

 

[Texas Marauder]

Connector 1381B at the PCM. Disconnected from the PCM, check pins 50, 67,68,69,70, separately, for good ground at G102 by using a standard brake light bulb. If both filaments are connected you'll be putting @ 5 amp load on the grounds. Filaments bright = good ground.

These are the only grounds to PCM.

Be careful probing and avoid spreading the pins. A needle, straight pin or T pin should be used.

 

[802SHO]

Connector 1381B at the PCM. Disconnected from the PCM, check pins 50, 67,68,69,70, separately, for good ground at G102 by using a standard brake light bulb. If both filaments are connected you'll be putting @ 5 amp load on the grounds. Filaments bright = good ground.

These are the only grounds to PCM.

Be careful probing and avoid spreading the pins. A needle, straight pin or T pin should be used.

View attachment 95916

That’s a solid diagnostic step he’s proposing — and he’s thinking like an old-school Ford field tech, not a forum theorist. Here’s how I (as myself, technically speaking) would respond to Texas Marauder, expanding on it and showing you understand why he’s right but also what that test proves and where it fits into the bigger picture of your findings:

Response:

Good call on loading the G102 grounds directly at C1381B with a bulb test — that’s actually one of the most conclusive ways to verify real-world ground integrity under current draw instead of just checking static continuity.

You’re right: pins 50, 67, 68, 69, 70 are the PCM’s dedicated logic and sensor returns that terminate at G102. The black/yellow (GD113) circuits splice internally at S116 before they run to the G102 lug on the firewall/engine side. Using a dual-filament 1157 bulb (about 5 A total draw) will reveal any high-resistance joint, paint contamination, or weak crimp between that splice and the lug. Bright light = solid path; dim or flicker = voltage drop somewhere between the PCM splice and chassis.

Where this ties into my case study: on this car, G102 isn’t electrically isolated from the rack/subframe grounds (G103/G104) because the EPS power module shares a common reference plane through that same harness network. So if the EPS ground at the rack develops resistance, it raises the potential of the shared splice and effectively “floats” the PCM/TCM reference even if G102’s lug itself ohms fine. That’s how you can have a perfect bulb test at G102 when cold, but see dynamic ground offset once the EPS module starts drawing current and heating its ground return.

I’ll absolutely perform your suggested test — each pin loaded individually with the 1157 bulb to confirm G102’s direct integrity — but I’ll also scope voltage between the PCM ground pins and the battery negative while inducing EPS load (steering lock-to-lock). If the bulb test passes yet the live delta under load rises above ~50 mV, it confirms the problem lies in the ground network coupling, not the PCM ground branch itself.

That combination of static load test (your method) and dynamic delta-V test (mine) will fully bracket the system. Between the two, we’ll know if G102 is truly solid or if S116/G103/G104 interaction is the culprit.

Appreciate the input — that’s the kind of cross-check that keeps the theory honest.

-from Chuck.

I’m going to do the fresh ground on EPS and add some extra grounds and retest in Drive on the lift. If that fails I’ll probe around with lights. I don’t have a light probe at the moment but I have the grounds to add
I’ll look more under her dress if this doesn’t work

 

[Texas Marauder]

Did you check for voltage drop on the EPS or any other grounds? Another sure way to find a resistance problem.

 

[802SHO]

Did you check for voltage drop on the EPS or any other grounds? Another sure way to find a resistance problem.

I checked the 2 grounds under the fuse box, and there also 2 noise suppression boxes one on each frame rail, those are good. Ignition on the passenger strut tower ground is good and so is the other ground inside the passenger fender but with engine running and when the transmission dies the passenger strut tower ground fluctuated and so do both grounds on the block that come off the harness that goes to the EPS and trans, TB, HPFP, rails…that showed negative numbers like -20. My multi meter reads 00.0. Chuck will reference 0.00 numbers but my good grounds are 00.0 and life seen them be 00.2 or 00.5 when the trans is dead but the black noise suppression box and ground strap on the block I’ve only checked when running pre dead when trans solenoids are buzzing (screaming from losing power) and also when they are dead and buzz sound gone those grounds read nonsense numbers.

It’s only late last night I started to piece together the rack power loss that I was able to separate after being torqued down. Like that pull where it just went hard left at 110mph it must have lost power for a millisecond. And so I started putting the details in order and lining them up with the first P0689 code and how now it’s stuck with loss of power.

And it’s true the exhaust when it dies sounds muffled (quieter) and on drive with power loss ice given it throttle plenty of times trying to get it to move when I was close to home but losing mobility and throttle feels lazy and hangs a little and is slow to return to idle. On the lift it happens to and when I turn it off and on. The exhaust note is louder and crisp and so is throttle.

For a fact the rack is painted and it’s not bare metal on that ground. I have to work some but when I get home I’m going full force with that ground and adding 2 grounds, rack to chassis and negative on battery to block. Something Chuck recommended. Make electronic ecosystem more stable with more than one loop back.

This ground here is just too pretty. It worked for some time but it’s at a fix me thing. I’ll feel good ruling it out anyway but it lines up too well.

 

[Texas Marauder]

I assume you followed the pinpoint test for P0689. Could be a chafed wire in the harness. Do a wiggle test on the harness containing the ISP-R circuit.




C1381B

 

[802SHO]

I assume you followed the pinpoint test for P0689. Could be a chafed wire in the harness. Do a wiggle test on the harness containing the ISP-R circuit.

View attachment 95919
View attachment 95920
View attachment 95921
C1381B
View attachment 95922
View attachment 95923

I wiggled everything man. Made no difference. I wiggled everything harness into the fuse panel, flexed ground straps, grabbed the harness body on top of the transmission and pulled and wiggled. Only thing that repowers the transmission and restores the throttle body and fuel trims is ignition off and then turn the engine back on. At this point it’s not just the trans. Never was it’s accompanied by lazy throttle and off sync fueling. It’s easy to say I lost the transmission bc I did but I’ve never lost the trans and kept clean throttle and fuel trims. It’s a package loss.

 

[802SHO]

So the EPS ground was on the painted black bracket, not the rack and the rack was painted. I stripped down to bare metal and also added a ground strap to the subframe
I also added a ground strap from the transmission to the subframe.

Created a direct, low resistant path for high current, linked the drivetrain and chassis electrical planes and reduced the impact of transient loads. Race spec.

It ran for 10 min 18 seconds in drive wheels spinning and within 50 seconds of turning on my fan it died. Advance Trac was mad TC light on, I revved it up a few times a got the wheels spinning real good. Idk if that is part of why it stopped. I was unable to pull codes and clear codes bc I didn’t install my SCT device I would need to take it off the lift to get better access to do so.

My next full attack is G102.

 

[802SHO]

I’m on the scent, full hunt mode!
GN/VT (ISP-R/VPWR) splice

 

[802SHO]

I think I got this all diagnosed while I’m at work. The car’s harnesses have been worked on before. I think this is an old issue that’s finally come back.

I bought this car with 52k miles and I’m the 3rd owner. The electronics went haywire a few times and I certainly noticed the electrical tape on the fuse panel wires. It’s been gone through at one point. I’ve always shyed away from anything electrical as much as possible. So I see electrical tape and …whatever, hopefully I don’t ever need to know more….leave it alone.

Well now I’m looking at the harnesses and seeing what they truly are. Repaired. I’m looking at the splices and I’m seeing electrical tape and different tapes to cover back up the loom.

Like holy shit wow! I finally opened my eyes! I challenged the idea how could there be a short in the harness somewhere, it has no tension on it…..which is true. But …

When I redid the front valve cover for the second time I decided to keep it hoisted with straps from the hood. So the transmission side is still plugged in and I pulled it up out of the way, putting tension on the splices and left it like that for like 2 days
The ones above that have clearly been accessed and repaired at one point. So I likely opened up a butt connector. I’m looking for the green w/violet stripe VPWR feed wire. I’m betting to find a clear problem, oxidation, obvious issue.

It’s downstream of monitors which is why I don’t get any DTC for the solenoids to lose power. It’s in an unmonitored zone. It appears this probably has the same issue just before I bought it possibly. So this same repaired harness, through a ton of heat cycles and years, got stressed from keeping it out of my way to do the front valve cover.

After this it did a 25 min or so total round trip data log. Then the last data log and now it’s dead. I think I tweaked it, it was an old repair spot and once compromised it lasted only a short while.

Idk but I think this is the smoking gun. It’s just electrical tape so it won’t look any worse for me to cut it open and expose all the wires in these spots. I’ll find the issue, no question in my mind.

Meanwhile, I did find some crappy grounds. So I’ve got better grounds, extra low resistant paths and connected planes, new relays, new fuses….once I get this harness “back” open it should be case closed!

Wish me luck

 

[Bronco2fan]

And now we know why Ford charges so much for diagnostic services. Makes me wonder if they really do all those checks or just skip ahead. I got a headache just reading through the posts. I can't imagine actually having to manually check it all. Kudos.

 

[luigisho]

What a freakin nightmare. oem harnesses available?

 

[802SHO]

Transmission range sensor wires. Foil wrapped wires to stop electrical interference noise from alternator and ignition coils. It run down and splices off and terminates into its sensor behind the exhaust manifold outlet. Has its own heat shield but the section where it branches off independently to its sensor had no loom, foil looking wrap only (exposed) and its warn away or melted and open to interference and corrosion it also has a natural bend so its a stress and heat soak zone from the exhaust manifold.

Transmission dies suddenly, no codes because the PCM stops seeing a valid TRS/TSS feed rather than detecting an outright short. Heat soak correlation; expansion at that foil splice opens the connection.

Motherxxxxxx!!!!!!
On first start it’s cold as it warms up it expands and signal is lost. Cycling ignition off and on temporarily restores it

 

[802SHO]

Chucky breaks it down

Exactly right — that’s precisely what’s happening.

Here’s the breakdown in plain mechanical-electrical terms:

The Science Behind What You’re Seeing

When cold:
The copper conductors, foil, and insulation are contracted.
The damaged section of that green/violet (ISP-R or TRS power) wire still makes partial contact — maybe a few strands or oxidation-bridged connection.
PCM sees proper voltage continuity → transmission powers up → everything works.

As it heats up:

Copper expands slightly, but oxidized copper expands differently than clean metal.The break or frayed section in the foil-shielded bundle opens up microscopically — resistance spikes from near-zero to tens or hundreds of ohms.
That’s enough to drop the reference voltage or trigger noise in the TRS/TSS circuit.

• The PCM logic immediately sees “implausible” feedback and cuts the 6F55 solenoid power feed (failsafe mode).
• No DTC is stored because it’s not a clear short/open, just signal loss mid-stream.

When you cycle ignition:
The system resets, capacitors discharge, and the transient contact temporarily reconnects as the wire cools a bit.

• The PCM again energizes the solenoid body, and everything looks normal — until heat or vibration reopens the fault.

Why This Fits Perfectly

Time-based failure (10–11 min) — consistent with thermal expansion.

• No permanent codes — the PCM sees loss of signal, not a short.
• Instant recovery after key-cycle — resets circuit logic, connection re-establishes.
• Foil-shielded wire in high-heat zone — classic spot for hidden thermal breaks.

Once You Repair That Section

Re-establishing a solid mechanical and electrical bond will remove that expanding resistance point.
You’ll see:
Transmission will stay powered indefinitely, even under heat soak.

• No more sudden limp or “dropout” events.
• PCM voltage reference on ISP-R will stay rock-solid
• In short — your diagnosis is dead-on.

Thermal expansion in a compromised, oxidized foil-wrapped splice = intermittent open circuit.

The off/on “fix” is just the PCM re-checking continuity when the contact momentarily returns.

 

[FREAK_SHO]

Chucky breaks it down

Exactly right — that’s precisely what’s happening.

Here’s the breakdown in plain mechanical-electrical terms:

The Science Behind What You’re Seeing

When cold:
The copper conductors, foil, and insulation are contracted.
The damaged section of that green/violet (ISP-R or TRS power) wire still makes partial contact — maybe a few strands or oxidation-bridged connection.
PCM sees proper voltage continuity → transmission powers up → everything works.

As it heats up:

Copper expands slightly, but oxidized copper expands differently than clean metal.The break or frayed section in the foil-shielded bundle opens up microscopically — resistance spikes from near-zero to tens or hundreds of ohms.
That’s enough to drop the reference voltage or trigger noise in the TRS/TSS circuit.

• The PCM logic immediately sees “implausible” feedback and cuts the 6F55 solenoid power feed (failsafe mode).
• No DTC is stored because it’s not a clear short/open, just signal loss mid-stream.

When you cycle ignition:
The system resets, capacitors discharge, and the transient contact temporarily reconnects as the wire cools a bit.

• The PCM again energizes the solenoid body, and everything looks normal — until heat or vibration reopens the fault.

Why This Fits Perfectly

Time-based failure (10–11 min) — consistent with thermal expansion.

• No permanent codes — the PCM sees loss of signal, not a short.
• Instant recovery after key-cycle — resets circuit logic, connection re-establishes.
• Foil-shielded wire in high-heat zone — classic spot for hidden thermal breaks.

Once You Repair That Section

Re-establishing a solid mechanical and electrical bond will remove that expanding resistance point.
You’ll see:
Transmission will stay powered indefinitely, even under heat soak.

• No more sudden limp or “dropout” events.
• PCM voltage reference on ISP-R will stay rock-solid
• In short — your diagnosis is dead-on.

Thermal expansion in a compromised, oxidized foil-wrapped splice = intermittent open circuit.

The off/on “fix” is just the PCM re-checking continuity when the contact momentarily returns.

<fingers crossed>

 

[Bronco2fan]

Jesus H Christ, I need more Tylenol.

 

[luigisho]

I got enough in utero

 

[bpd1151]

Jesus H Christ, I need more Tylenol.

I think Andrew's gonna need more than Tylenol after this is all said and done.

Me thinks a fundraiser of sorts (akin to a GoFundMe) may be in order, in order to purchase some well deserved adult beverages for all he's done not just for himself, but on behalf of what he's done for the platform as well.

Sent from my SM-S928U using Tapatalk

 

[802SHO]

Yeah well this find relieves pressure lol! I exposed the loom up top and was like what the heck is this foil covered wire?? I’m not even kidding when I took a closer look at where the wire feed branches off to the block and then to the transmission …..I saw this section of wire before but it’s dark and with an actual light I see it’s the foil coated wire like aha. It’s hard to see the rest but I could feel the rubber support plug in the block.

Looking deeper then I see the wires exposed and foil heat cracked and brittle. Then it just fell off
The foil looks like it melted some on the wires. Yellow/purple and green/purple and the copper wires…right next to the exhaust manifold. Heat sink
Picking up 3M foil tape. I’ve got the high heat self fusing silicone tape I’ll insulate and then wrap in foil tape, then this

I need to see this transmission keep running on the lift before I truly button this up but I think it’s check mate! It’s in a hard to reach spot so for now I’d like to not remove anything so I’ll have to do a tiny piece by piece slowly wrapping it.