OBDII 5mt CEL... fix

[TomsSVX]

Well I have figured out why OBDII SVXi with the 5mt conversion throw codes...

The code tossed is the P1702 tcm... This also stands for a lacking signal comming from the AT Diagnosis terminal. Now its not as simple as I hoped. This signal is a waveform that varies voltage from less than 1v to more and 4 volts. Now in order to cure this we can eithe leave the TCU connected and need to switch the inhibitor to park whenever you need to remove the key or start it... or we can work together on finding a way to either lose the need for the signal or create a new one

Tom

[TomsSVX]

pin 80 needs to be supplied the waveform... sory the pic is so big but I had to scan a print-out

http://i65.photobucket.com/albums/h225/TomsSVX/Picture.jpg

Tom

[SVXRide]

Radio Shack oscillator chip with a 9V battery?
-Bill

[TomsSVX]

Bill, I was thinking of something a little more perminant... BTW do you ever check your voicemail??:p

Tom

[Trevor]

Well I have figured out why OBDII SVXi with the 5mt conversion throw codes...

The code tossed is the P1702 tcm... This also stands for a lacking signal comming from the AT Diagnosis terminal. Now its not as simple as I hoped. This signal is a waveform that varies voltage from less than 1v to more and 4 volts. Now in order to cure this we can eithe leave the TCU connected and need to switch the inhibitor to park whenever you need to remove the key or start it... or we can work together on finding a way to either lose the need for the signal or create a new one

Tom

Tom,

The wording "whenever you need to remove the key or start it..." is rather loose. Please fully describe every condition in which the inhibitor switch is required to register park, as well the exact subsequent circuit change required at the inhibitor switch.

With this information on hand, I should be able to make viable suggestions towards solving your problem. ;)

Edit P.S. Can/does "key removed" equate with key in the off position. If not and two functions are involved, do OBII cars have an ignition switch with key in/out contacts, as do many cars which incorporate a "keys left in at door opening", warning function?

[SVXRide]

Bill, I was thinking of something a little more perminant... BTW do you ever check your voicemail??:p

Tom


Tom,
Been working in a "special" building at Lockheed Martin's facility out in Denver - no cell phone reception:rolleyes: . I'll check my voicemail tonight (yeah, got to work 28 hours this weekend:eek: )
-Bill

[JaySVX]

Interesting... I might have an idea. What are the variances in the signal coming from the tcu? Is it seperate for P, D, N, R, 3, 2, 1 or is it just P and everything else?

If it's the latter, for starting conditions only, i'm pretty sure something can be devised rather easily. If you can get me the readings, or provide me with a tcu that would give me readings on my o-scope, i'm pretty sure i can duplicate the wave-form.

I can have seperate automatic switching sequences for the two signals, tying into the clutch switch which prevents the engine from starting unless the clutch is pressed, so that it will throw park codes when clutch is in, normal driving codes when clutch is out.

Would that cause any sort of problem when you go to shift gears as far as the ecu goes?

[TomsSVX]

the issue is that the diagnostic plug for the TCU is no longer getting a signal to the ECU. This is because of its lack of presence. Now to fix this, we would need to create a waveform that varies from less than 1v to more than 4v. At a frequency that I do not know. This freq. can be measured on a stock OBDII SVX that is OBDII and still has an auto. Once it is measured we can use the velleman kit that I have for my DCCD controller and set it up to create the frequency we need

tom

[JaySVX]

Yeah, that's what i mean. Would need to be able to toy with an OBD2 system to get the wave form. When does it change, what effects the waveform reading?

I can most likely make something from radio-shack parts to emulate the wave-form, without having to go so far as to hook it to your DCCD controller, it would just be a tiny circuit board that would need to be wired in.

BTW, i got the parts for your data dump piece, be putting it together this weekend hopefully. Just need to find 1-2 more parts.

[TomsSVX]

I have a feeling it doesn't change unless there is a code present. So we need a perfectly good operating OBDII svx to be a test pig... Not to mention an osciliscope to measure the wave form:o

Tom

[JaySVX]

i've got an o-scope. just gotta dig it out.

[TomsSVX]

aw man... those aren't cheap or easy to find... great!! Now who has that OBDII that will stand to the test:D

Tom

[JaySVX]

Nah, but it's like 20 years old. Some kids got blocks to play with, i got an o-scope...

[longassname]

I don't imagine it being an analog signal.

[TomsSVX]

lost me... that was hard huh??:o

Tom

[longassname]

I just bet it's a digital signal at ttl levels and the frequencey of the crystals on both the tcu and ecu. It's a data link

lost me... that was hard huh??:o

Tom

[TomsSVX]

so whats the suggestion to rid me of this issue when I go EG33 OBDII in the wagon?? Can I simply splice the ecu data link into the TCU data link pin to get the reading??

Tom

[longassname]

Assuming it is in fact data, and i can't imagine it not being, that all depends on the requirements of the ecu software. Something as simple as holding it high or low might work..then again it might want some steady signal of something or replies to requests.

so whats the suggestion to rid me of this issue when I go EG33 OBDII in the wagon?? Can I simply splice the ecu data link into the TCU data link pin to get the reading??

Tom

[TomsSVX]

gotcha... so the signal is the data link and depending upon what that signal reads between voltage and wavelength determines.... ohhhhhhhhh... im an idiot... lack-o-sleep

Tom