I adjusted the POTs (holy smokes they're touchy!) and brought the key on voltage down to 1.0x volts (they were at 1.3 and 1.4 initially). Didn't get the VSS or Turbine related codes after that.
T87 Swaps
- Thread starter kidturbo
- Start date
That little skip at the end of each square is the mosfet closing to slowing. Rather than just ON/OFF 5v signal, it's closing the switch to slow and leaves that trailing edge.
Edit: Or were you speaking of this skip?
The first screenshot above of A50 replicated signal is basically a -5v to +5v. I was shooting for a 0 to +5v square wave. And currently we have a what looks to be a +2 to +5v wave. I know the ECM will track a 0-5v signal, if we can just get that +2 down to a zero or squared off a bit, seems everyone else be happy camper for now.. Then back to the drawing board... lol
Last edited:
Super.. I kept stepping those up to see what level it took cause a High Circuit code, then backed em off.. However with that VSS Open (Permanent) never clearing on the bench, I knew it would take a couple minutes running time to tell if they were happy as set. You rolling T87 on 06 OEM sensors, we breaking some ground this weekend.
Success! I went to change the resistor associated with the TOS signal and discovered that it had broken off - with the body of the resistor right against the board, one little bend and 'snap!'. Put a new resistor in line and MUCH BETTER! The run around the block was uneventful this time 
.
. 
Yes, that's the little skip I was asking about. Did Coug just fix it? Sorry for the dumb questions.
I only soldered the terminals, so... Resistor was guy before me.
If ya ever want to start a good flaming topic, bring up solder vs crimping on the milspec harness group over on FB. The space/aircraft guys are totally crimp everything. To which I always bring up, that $10 crimp connector goes to a soldered PC__ what?? And all your hard work is dependent some computer solder job.. Can't get around it.
To answer the question, I was going with interference on the skip. Bad connection will do it.
If ya ever want to start a good flaming topic, bring up solder vs crimping on the milspec harness group over on FB. The space/aircraft guys are totally crimp everything. To which I always bring up, that $10 crimp connector goes to a soldered PC__ what?? And all your hard work is dependent some computer solder job.. Can't get around it.
To answer the question, I was going with interference on the skip. Bad connection will do it.
I solder all my larger connections because I don't have a big enough crimper
That's the kind of discussion that can start WW3!
So what's your weekend update on the VR conditioner @Cougar281 ? I have a couple more of those on my desk, in case you break that one again.. Sent it in a bulletproof case, thinking that would suffice...
We used those old school thru-hole 220 Ohm resistors as pull downs on all 3 [Load] lines so one could easily change the values. Then POT's actually pull the line level back Up to our floor value on L1 and L3. Once levels are verified on a couple trucks, those can all be swapped for smaller form factor surface mounts.
The T87 has additional 220 Ohm pull downs internal. Is how we landed on that value. But bench testing so far shows we still need those additional pull downs on the Conditioner side to keep the mosfet signals nice an square, as you've discovered. So now I guess we need to get you up to speed, literally.
We used those old school thru-hole 220 Ohm resistors as pull downs on all 3 [Load] lines so one could easily change the values. Then POT's actually pull the line level back Up to our floor value on L1 and L3. Once levels are verified on a couple trucks, those can all be swapped for smaller form factor surface mounts.
The T87 has additional 220 Ohm pull downs internal. Is how we landed on that value. But bench testing so far shows we still need those additional pull downs on the Conditioner side to keep the mosfet signals nice an square, as you've discovered. So now I guess we need to get you up to speed, literally.
Last edited:
After fixing that resistor and tweaking the POTs so the VSS/Turbine outputs to the tcm down to about 1V, it seems to function perfectly. I've gotten a P0880 a few times, so maybe I have a loose wire. Didn't have PRNDL, TFT readout, tap shift, etc, but that was of course expected. Now, if we can get rid of the pesky U codes, I might actually be able to run a fast adapt and make it shift nicer.
Glad to hear it's somewhat stable. Now we can move to fudging data or hacking the TCM into submission.
I'm gonna go hit it with HPT and see what options they have. I have that T87 MPU and what looks to be some SWD pins staring at me on this board. If we can kill some DTC's using any tool, and get a working base bin for now, probably be a our next best move for the LBZ/LMM setups.
As for LML swaps, beleive that solves the main hardware restriction. So first person to show up in my yard with an LML truck for the day, wins a free T87 upgrade. If willing to beta test the untested of course..
I'm gonna go hit it with HPT and see what options they have. I have that T87 MPU and what looks to be some SWD pins staring at me on this board. If we can kill some DTC's using any tool, and get a working base bin for now, probably be a our next best move for the LBZ/LMM setups.
As for LML swaps, beleive that solves the main hardware restriction. So first person to show up in my yard with an LML truck for the day, wins a free T87 upgrade. If willing to beta test the untested of course..
One thing I did find on the tap shift stuff is apparently pin 39 on the T87 is for the TUTD buttons, but the only place it's used is the 2016 (and maybe 2017) CTS-V - why they went hardwired to the TCM for the V and everything else was serial data, one can only guess.
As far as HPT, they seem to have DTC's pretty well mapped out, the problem is using them... I have no way to read or write using HPT, only EFILive, and oddly, EFILive has DTCs mapped out for T87 gasser TCMs, but not the Allison calibrations.
As far as HPT, they seem to have DTC's pretty well mapped out, the problem is using them... I have no way to read or write using HPT, only EFILive, and oddly, EFILive has DTCs mapped out for T87 gasser TCMs, but not the Allison calibrations.
Last edited:
Reason I'm thinking we load this one up using HPT, and then I pry the that full bin back out of it by any means possible.
Spent a couple hours this evening flashing, reading, editing, flashing, reading and comparing the T87 Allison OS using HPT and EFIlive. And soon discovered, if you combine the options available with each tool, one might be able to build a single tune with all the DTC's off, and all the tables needed to make it shift. Possibly even better than an A50...
@Cougar281 has been working on a CAX8 file to add some of our known DTC's we might need to disable to clear the light, and see how these units respond now with some missing or malformed CANbus data. But for now, flashing options with HPT then read with EFI, seems to accomplish the same without a CAX file. If you have booth tools that is.
@Cougar281 has been working on a CAX8 file to add some of our known DTC's we might need to disable to clear the light, and see how these units respond now with some missing or malformed CANbus data. But for now, flashing options with HPT then read with EFI, seems to accomplish the same without a CAX file. If you have booth tools that is.
If they had the Allison map, yes. That seems to be the hold up. Yet both software have some matching tables mapped, and each has stuff the other doesn't. So our approach is basically been flipping switches in one, and see what changes when ya read back the bins.
Sent from my SM-S901U using Tapatalk
Sent from my SM-S901U using Tapatalk
That gives us a few OS options to stay busy for a while, with T87 and T87A hardware EFIlive loadable options now in our bag of tricks.
So long as that keeps the MIL out for issues we're currently aware of, we can move to testing the CANbus data again. Hopefully find what these Allison OS versions actually require to shift normally. Torque management data being my only real concern for now.
So long as that keeps the MIL out for issues we're currently aware of, we can move to testing the CANbus data again. Hopefully find what these Allison OS versions actually require to shift normally. Torque management data being my only real concern for now.