Cadillac Owners Forum banner

Ultrasonic Reverse Parking Assist System retrofit in Gen. IV (OBDII)

12K views 35 replies 14 participants last post by  N*Caddy 
#1 ·
Yeah that is right, I am in the “research” stage for retrofitting URPAS in my ’97 STS.
After about 2-3 hours of work I finally got a complete Ultrasonic Rear Parking Assist System from a 2000 DHS (fairly cheap $50).
This is the bench test (better said Persian carpet test :)).



In case you don’t understand what is going on in the video, you are looking at the telltale assembly that normally mounts on the car headliner near the rear glass (central). You can also see 2 of the 4 ultrasonic sensors normally mounted in the rear bumper. As I am approaching to the sensors (normally the vehicle back up towards an object) the system pick up my position and illuminates the LED in the telltale assembly in the following pattern:
a) ~6 feet away first amber LED ON
b) ~3 feet away both amber LEDs ON
c) ~1 feet both amber LEDs and the red LEDs are ON
d) dangerously close all LEDs are flashing.


Here are the observations:
1) Works fine as long as the reverse signal is not grounded. In the car this input is pullet to battery voltage when shifting in reverse (by the reverse lights relay) and then pulled to GND when not in reverse trough the reverse light bulbs (the filament). So if you remove both reverse light bulbs your URPAS will work all the time. That is a caution for those of you with URPAS planning to update to LED for reverse lights.
2) It needs the RIM to operate otherwise it goes in error mode. Thank God for the diagnose mode (available by grounding the diagnose input) otherwise I had no clue why was not working. Basically the URPAS module monitors this output and expects to see it pulled to battery voltage by the RIM. Well I found a way around this issue by simulating the RIM input with a 10K resistor, so I tested with the RIM and w/o using the RIM simulate (the resistor).
3) Oddly, is not going in error mode if the VSS is missing (obviously I had none in my test), I guess is looking only for overspeed (>5MPH and ignores everything else).
4) Sound…that is the big issue here! It is not working as I would like it as in when the URPAS wants the chime it grounds the output. It actually does a frequency burst to communicate 1 chime, fast chime and I could not figure out how it requests continuous chime (I guess is higher frequency). I don’t have an oscilloscope so I was working with an LED, that is the most information I could get by observing the LED. But is enough for me to know that I am unable to use this chime output to connect a buzzer in case the RIM does not work fine on my databus. I would have to use the LED outputs in the following configuration:
a) 1 beep for the first amber LED
b) Slow beep for the second amber LEDs,
c) Continuous been for the red LED on
d) Fast beep with red LED flashing.
Unfortunately it will be an issue with the bulb test (every time shifts in reverse). Hopefully the RIM will work fine with my data line and post no code (as I am only connecting it for the chime request). Also the beeps will be there while in diagnose mode (actually that I can fix).
Let’s see how is the RIM behaving on my databus really, really hope it will be fine, otherwise it will be allot of extra work.

In this case is measure 10 times then 10 times more then cut.
 
See less See more
#4 ·
Well the 4T80 range switch is linked to the shifter - so any delay, if any, lives inside the PCM or RIM. But since it also does the bulb test, you might have to duplicate that functionality. (Or just have a minimum time lockout with the signal coming from the bulbs?)
 
#7 ·
IT TOTALLY WORKS!!!!!!!!!!!!!!!!!!!!!
The RIM integrates perfectly on the data line, there is NO CODE, and I HAVE THE CHIMES!
It is beautiful, now I am off to the body shop to get the sensor bezels painted. I will have the system ready in the next few weeks.
But the chime pattern is slightly different that described by Sub (I believe there are differences between ‘02+ Seville and ’00 Deville in this department).
I get no chime at the first amber LED
I get 1 chime when both amber LEDs are ON
I get regular chime (the same like door open, engine off key in the ignition) when all LEDs are solid ON or flashing (no difference). The RIM can read the shifter position too, it only request the chimes when in reverse even if I fooled the URPAS module to believe the car is in reverse when is not.
:banana:
 
#8 ·
Congratulations!

What year RIM did you use and from what model? 2000 from the donor DeVille?

I had a feeling that it would be fairly easy to add the ultrasonic backup alarm as it has its diagnostics are self-contained and later RIMs seem to go in without conflict. Now if we can ever figure out that TPMS system and the conflict with the PCM....
 
#9 · (Edited)
Here is the in-car test, the only difference is the VSS is not connected and obviously the components were on the driver’s floor not mounted properly. The sensors were taped along the hood and I was using my hand to simulate the vehicle backing towards an object. The engine was off but the ignition was ON and the shifter was in reverse (otherwise no chimes).



The module underneath the telltale display is the URPAS and the one with the 2 connectors (in front of the seat rail) is the RIM.



I also tested the PZM functions and all work just fine, no codes what so ever. Is the same like adding 2 HUs on your car (yeah is possible, both will control the volume and other RIM/AMP related functions).

Smike the thing with the RIM (~'00 DHS) is a very conspicuous module on the data line, is mainly a slave module doing nothing but what other modules request. The request for chimes is a standard HU function supported since ’96, the IPC can request chime, the PCM too, so the fact that another module request a standard existing function is not an issue. In the TPMS, the RFA asks a new function (display a certain message on IPC), and that would be no problem if the IPC is updated, but from your observations looks like the IPC is not the only module involved in this…I don’t remember, last we were suspecting the PCM, how about the DIM?
 
#11 · (Edited)
The harness is ready, the sensors are painted now drying overnight into the spray booth.

The bottom most harness (the one with the wire loom) is the sensor harness (running behind the bumper cover). The 4 black connectors on the bottom right side are the sensor connectors. The harness on the left is the harness running between the URPAS module (mount on the power antenna bracket) and RIM and ends with the connector to the harness for the Telltale module. The blue/black connector on the bottom left is the URPAS module connector while the 2 larger connectors (blue and gray) are the RIM connectors. The telltale harness is the one on the right side and runs behind the left side sailing panel then along the upper edge of the rear glass underneath the headliner. The white connector (top right) is the Telltale module connector.
Hopefully I have time tomorrow at least to run some of the harness in the car. The routing is identical as in DeVille the only alteration is the harness to the sensors will come out of the trunk along with the right taillight harness (trough the same gasket). In the DeVille there is a hole on the right side underneath the bumper (shared with the harness for the ELC compressor.
The electrical schematics between ’97 Seville and ‘00+ DeVille are so different yet so similar, even the circuit numbers for the various circuits is the same. In the DeVille the RIM data line is just one wire, I will actually integrate the RIM in the same loop with the CVRSS module. I opened the RIM and checked the PCB diagram and it has 2 data pins (A12 and B12) so it can be installed in a loop. Basically I will remove from the CVRSS connector one of the 2 data wires and run it to the RIM then run a small wires between the RIM and the CVRSS data pin (thus extending the network loop). And by adding the RIM to the rear electronics tray I literally run out of space, there is absolutely no more space in there as the car has every possible option installed (except OnStar). Hmm…:hmm:

Sensors are painted and ready to install.
 
#12 ·
Today I got the TellTale module along with the harness installed.
I really thought allot before cutting a hole into the liner (I had the template from the donor car). Once the headliner cut there is no turning back (is not easy to find one in good shape these days.
Thank god it came out perfect dead center (after measuring 5-6 times) and dead straight.



I also run the harness along the inner side of the headliner, even used hot glue exactly as it is done from factory, including the routing is identical with the one from factory.
Tomorrow I will install the URPAS module, the RIM and the remaining in car harness plus wire everything up.
This is the last picture of the car with no parking sensors taken just over 1h ago.



I am almost sure I won't have time to install the sensors tomorrow (that is another measure 10 times drill once deal). Hopefully will came up just like the TellTale module.
 
#23 ·
Today I installed the remaining In-Car harness, my car now has 1 PZM and 2 RIMs (RIM no 1= Radio Integration Module, and RIM no 2 = Rear Integration Module).
The second RIM has functions for fuel level, internal lights, rear defogger, ELC, Reverse relay and some other doubling the PZM. Yes that is right I can just transfer some of these to the RIM from PZM (basically I have some redundancy), but the main function of the RIM is requesting chimes when the URPAS module command so.
Everything is wired up (VSS, IGN1, CONFORT, GND, Reverse Signal and Data line), I did not installed the URPAS module yet (not connected) as I had no time for the sensors and with no sensors it goes into error mode (basically the red LED flashes while in reverse). I was tended to do at least a rear life test with the sensors taped to the bumper but unfortunately I was late for dinner (you know Easter and family and such) so I have to wait till next weekend. Well I did a quick test making sure the modules are wired correctly and I get signal to the TellTale.
Thank Cadillac for designing one spot on the far right of the electronics bay (roughly the same position where the RIM is mounted in the newer cars). I couldn't find anywhere in the FSM a reference for what this spot was reserved for and is a perfect fit for the RIM (looks like a provision for a option that never make it in production). Literally I have 0 extra space on that electronics bay now, can't add anything more (not that I have anything else to add except the OnStar stuff - but since the analog system is not working...no point adding it).
One note…the 3 LEDs on the TellTale module are a bit too bright (especially in shade/dark), and project a slight pattern of light on the headliner. It would have been nice if the dimming was taken in consideration, but since there is no data line to the URPAS no way coordinating the dimming with the dimmer switch. Too bad because the information is present on the data line. I could add this option by controlling the supply voltage based on the dimmer voltage (I would have to add the dimmer signal to the URPAS – not too complicated, but for now I better finish the install).
As an observation I can see about the same approach as with the memory seat modules. It appears to be commissioned to an independent team fairly insulated from the others (perhaps outside GM) so it had very little integration with the rest of the system (remember the lack of ignition voltage signal to the massage module leading to the 10 minute timer). Either this or the option was started a bit later into the design stage. This explains why they chose the digital input to the RIM solution to request chimes rather than adding the data line to the URPAS (more complicated firmware for the URPAS). I guess, financially it was less expensive to add an extra input to the RIM as the firmware was easy to change to support the chimes as the data line interface was already done.


This is a picture of the connector for the harness to the TellTale module (left side of the electronics bay near the ELC CB. You can see the 2 connectors for the PZM (the ‘98+PZM!!). Unfortunately I had no pictures with the location of the RIM, it was far more interesting as it has the Data wires, VSS, Chime input and few others, but by the end of the install I had no more time for pictures as I was running late.
 
#28 · (Edited)
A slight bad news...I just discovered that one of the URPAS sensors is not working. I am not sure if I mentioned but I did noticed one of the sensors was replaced (according to the MFG date in ~’02-‘03), obviously the yo-yo at the dealership did not installed the rubber seal. It turns out this replacement sensor is not working. I did notice, once in a while, I had hard time with the system to initiate, few times after boot was in error mode. Again thank God for the diagnose mode I was able to figure out (based on the LED flash pattern) that sometimes I have issues with the sensors. After spending 2 days on it I got so familiarized with the way it works that I know when to disregard some other errors that are caused by me (say I did not emulated properly the switch into reverse in the first 5 seconds after entering in diagnose mode and such). Finally I insulated the issue to that sensor so now I am in the market for a sensor (go figure the replacement sensor failed, why I am not surprised).
The thing is the FSM is not very accurate, perhaps “detailed” would be a better word. For example it mentions that when entering in diagnose mode ensure you shift into reverse in the next 5 seconds after turning ignition ON (no reason why). That translates in my word in: “ensure the reverse input has a transition from Hi to Lo in the first 5 seconds after powering the URPAS module with the diagnose input grounded.
The reason for these 5 seconds is the URPAS expects to see this transition during these 5 seconds count down, it is the only way to know the reverse input is wired correctly.
I was ignoring this since I noticed that for bench test I don’t even need to ground or pull to voltage this input. I had to look at the PCB and understand the schematic. Turns out that the input is pulled HI in the URPAS and if the wire is not connected (floating input from my point of view not knowing the URPAS internal schematic) basically it is pulled high. The firmware expects to initiate with the input pulled to GND and this is true when connected in the car as the bulbs do pull the input to GND (trough the fairly small filament resistance) since you can’t start the car while in reverse.
So only after I eliminated all these I was able to figure out what is real warning and what is bogus.

I am not sure where I am going to find a parking sensor (I see eBay has one here and there but is at least $70). To be honest the sensor is nothing but a tweeter/microphone combo and doesn’t worth this amount of money. But I paid $32 for a sticker (in 2 occasions) so I’ve done worst.


This is what I used to diagnose the sensor. Since the TellTale module and the RIM and the harness is in the car, I had to fabricate a TellTale simulate and all the wires and the RIM (lots of work just for diagnosing a stupid sensor).
 
#29 ·
Great news, I am back on track, today I scored another 4 parking sensors (complete with all the hardware) and just for good measure I took the URPAS module too, as it was the same price ($20).
Now the sensors are again in the spray booth for painting (tomorrow morning) and out of 8 sensors I got to have at least 4 working (I know one is not working but still I have 7 more).

Unfortunately I may not be able to install the sensors in the bumper this weekend either, lets see how tomorrow goes, if not next weekend.
 
#30 ·
Got my second set of sensors back from the body shop.
Remember I have 2 URPAS module to choose from, the first one from a '00 Deville and the second from a '04 Buick Rendezvous.
The P# are different but as far as I can tell other than the 5V supply chip inside there is no difference except for one note on the label “SW 3.0” v.s. the Deville one “SW 2.2”. And after a bench test you can tell the difference between the two firmware versions, the Rendezvous URPAS module is faster to react than the Deville one. If I move the object in and out from the sensors field of "vision" the Rendezvous module notice the movement much, much faster is a very obvious difference. I guess the Randevouz module goes in rather than the Deville. Also the new parking sensors are all working fine so I have now 7 working sensors to choose from.
This weekend should be the weekend when I'm finishing the retrofit.
Another thing, the 3.0 firmware also includes a "disable" input, in the Deville is not wired. I will not use it but is good to have the option.
 
#31 ·
Today was THE day when I finalized the install.
Measure 10 times and drill once (the top edge of the tape defines the middle line of the URPAS sensor holes):



The first hole ready, I was so worried about chipping the paint, but nope it all worked fine.



This is the final product, notice the harness goes into the trunk sharing the same rubber grommet as the right side outer taillight harness (not finalized in the picture). It was not easy to squeeze the 6 extra wires trough the same tight grommet but with patience (about 1 h) it worked.



And here it is WORKING in a ’97 Cadillac Seville STS, never meant to be yet it is, and it was not that difficult.



The first flash is when I shifted in reverse (is the bulb test sequence) then you see as I am approaching the vehicle parked behind the LEDs turning on in the sequence 1 amber LED, 2 amber LEDs, all LEDs on, all LEDs on flashing. The corresponding chime is heard in the background 1 chime at first amber LED on then continuous chime when all 3 LEDs are on.

Also in the process I accidentally drove a ‘03 bi-turbo V12 CL 65 AMG (>600HP!!!!). That is really a sweet ride and I notice the URPAS sensors are identical with the ones in my Cadillac. :D
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top