How to: Use DHP with a 2004

[plastic_indian]

Owners of 2004 model year cars will get an error when trying to open a BIN pulled from the ECM with DHP; DHP does not support model years beyond 2003.

Reviewing BCM and PCM part numbers confirmed it was theoretically possible to substitute modules salvaged from an earlier model year, so it was off to the boneyard to find a suitable candidate. A like-optioned Victory Red two door 2003 offered up the requisite modules. The modules were swapped into the 2004, the antitheft learn completed, and the engine started/ran normally. After running for a short time, the SES illuminated and the PCM stored P1810. A test drive confirmed transmission operation to be abnormal. When P1810 is stored, the PCM will illuminate the SES, crank up transmission line pressure, assume the selector is in D4 and freeze the transaxle adaptive functions.

P1810 is effectively a fault stored when transmission fluid pressure switch outputs are implausible or do not match expected values. The transmission fluid pressure switches and the transmission fluid temperature sensor are integrated in one assembly attached to the valvebody. Reviewing the relevant circuit diagrams for both 2003 and 2004 model years revealed the cause of the fault. The 2003 4T45E transaxle utilized six fluid pressure switches (P/R/N/D4, Lo, DR, Rev, D/2/1 and TCC release); for the 2004 model year GM eliminated the P/R/N/D4, Lo and D/2/1 switches. The earlier TFP switch assemblies were plagued with problems; perhaps the revision was the response to this.

This fault can be artificially eliminated by providing to the PCM the signals expected from the missing TFP switches. As the TFP switches are providing confirmation of the transaxle manual valve (the spool valve mechanically positioned by the gear selector lever) position, the transmission range switch (TRS -- aka neutral safety switch, start safety switch, et al) can provide the same information. The TRS provides selector position information to the PCM on four circuits: signals A, B, C and P. The PCM programming contains a table wherein each combination of values corresponds to a specific position. The problem lies in harnessing the output of the TRS, parsing it to determine manual valve position, and then generating a useable signal for the PCM.

Before moving forward, a brief word about how the PCM monitors these signals -- both TRS and TFP -- is in order. The PCM provides a 12 volt reference to each of the switches. The switches complete or open, as appropriate, the circuit path to ground. The PCM hardware includes a pull up resistor on each reference voltage output. When a target switch is closed, potential voltage on that circuit goes low (near zero); when the switch opens, potential voltage goes high (near source [12v reference]).

Service information gives the TRS and TFP signal logic tables for 2003 and 2004 model years. After comparing the tables and consulting the appropriate wiring diagrams, it is revealed the 2003 PCM expects but never receives signals on pin 22 of connector C1 and receives inaccurate signals on pin 17 of connector C2. C1/22 is, in fact, not even present in the 2004 wiring harness. The harness connector can be disassembled to fit a pin in cavity 22. The connector components are manufactured by Delphi; the GM part number for the pin is 12084912. Pins can be salvaged from connectors in the wrecking yard, purchased in bulk online or acquired from a GM dealer. In most instances, however, the GM dealer will have, but NOT *inventory,* connector pins and similar bits; when one of these components is used by a tech during a repair, the part is one of myriad costs covered by the 'supply charge' shown on the invoice. Recently, a certain 2004 owner bought lunch for one of the fine countermen at a large local northside Chevrolet dealer and was permitted to poke around in the small parts room...

Using either direct splicing or fabrication of a breakout harness, parallel electrical connections to TRS A/B/C/P and PCM pin C2/17 can be made. As C1/22 was added to the PCM harness, it can be routed and connected in whatever manner desired. In lieu of drilling additional holes in the firewall, the main harness passthrough grommet is relatively easily removed after removing the underhood fuse panel (as known in GM speak as the underhood bussed electrical center/UHBEC) and its mount bracket. Very little room is present to build a breakout harness at the TRS -- direct splicing is much easier.

Now that electrical connections to the inputs and the outputs are established and a table can be drawn of TRS A/B/C/P signals versus signals expected on the two PCM pins C2/17 and C1/22, the only problem left to work out is how to interface the two.

Enter the Arduino Uno (www.arduino.cc). Code was written to allow the Arduino to monitor analog signals from TRS A/B/C/P and then selectively control two outputs. As the Uno is a 3.3v/5v device and the car is a 12v+ system, N channel MOSFETs (Fairchild FQP30N06L units sourced from SparkFun) were incorporated in a printed circuit board (etched at home using the toner transfer method) to allow the Arduino to interface with the car.

Code, Fritzing files (www.fritzing.org), Excel files of the logic tables and Excel file of the header chart for the finished PCB are on GitHub: https://github.com/FuelishPursuits/4t45e_interface

In the Fritzing file, the breadboard view is a faithful representation (connections wise) of the successful proof of concept; components are standard Fritzing library components in lieu of the true parts. The MOSFETs are the above mentioned Fairchild Semiconductor pieces. The voltage regulator is an NTE966. The resistors are whatever 1/4 watt bits were kicking around in the parts drawer; their impact on the circuits' function can be tweaked on the fly by altering the UThreshold and LThreshold values in the Arduino code. The headers are generic .10" spacing Dupont style bits. The PCB view is the drawing from which the actual toner transfer was exported. The schematic view is an unrefined spaghetti bowl nightmare bourne of Fritzing's effective yet crummy autorouting. Them's the breaks.

The whole mess is enclosed in an ABS plastic project box from Adafruit. Said enclosure is mounted on the left side of the engine bay betwixt the UHBEC and the battery; the bracket for the UHBEC has, however, been trimmed back after the OE airbox was eliminated. The wiring is terminated with GM/Delphi weatherpack connectors (sourced from the NAPA Belden book) to facilitate convenient assembly/disassembly.

[O1GAGT]

Interesting, setup.
DHP does work on 2004 and 2005 Grand Am's... you likely just needed to flash a supported .bin file. If you want I can see if I have a supported .bin I can send you.

As a side note, I swapped a 2004 transmission into a 2000 GAGT. I forget what I did, but it was along the lines of switching a pin in the harness and then adding a pin and grounding it out. The car ran and shifted fine in drive however I lost the ability to manually shift from 1,2 and 3.

[plastic_indian]

GM had a couple reflashes for the '04 m/y (perhaps 05 as well...) to address garage shifts and 'enhancements' for a DTC that was problematic (can't recall which DTC). Perhaps those recals broke compatibility. I did not think to try and rollback to an original calibration. If you have a known working BIN for the 04 that would be great.

Cavity N of the A/T connector is empty on the 04 -- that is probably the one added. If getting normal automatic operation is that simple, I evidently misread the wiring diagrams. No matter - learning the Arduino was fun, will undoubtedly be useful in the future, and the transaxle is -- so far anyway -- operating 100% to specification.

[O1GAGT]

I will have to check and see if I can locate one on my old computer.
I can't remember, but did I give you a CD or jump drive with the tuner? If so, I believe I had a archive of .bin files.

[plastic_indian]

Flash drive. LA1 BINs are all 2003 or prior.

[The_Maniac]

For 2004, you need OSIDs 12589761 or 12589762 to use a DHP. I have an '04 and no issues with my DHP connecting and adjusting it.

[O1GAGT]

Go to https://tis2web.service.gm.com/tis2web/ and enter your vin to identify your calibration OSID. It should have a few variations.

I have attached the ones that The Maniac identified above.
If you need any others, let me know and I will see what I can find.

[AaronGTR]

Nice. I was pretty sure you could still tune an '04 with DHP with the right bin file, but I've never dealt with those compatibility issues myself. Cool project though. Just goes to show, you can make anything work with enough knowledge and perseverance.

I still haven't been able to get my DHP working again. I wish I could, as there were a lot of things I still wanted to try with my tune, but I haven't really had any time to work on it, and I've got some body damage to fix on the car first.