MrVette90l98mt

Need help in solving a speedometer output issue with an LS3 EROD manual trans calibration (CA smog certified create engine package) kit I installed. For the last 3 months I have been dealing with the local dealership trying to get this fixed under warranty.

Facts:
-The car runs and drives with OE ZF6 trans, 500mi break-in complete
-VSS generator confirmed to be working at the dealership outputting 2.5v
-swapping out ECM with another did not fix the speedo out signal
-ECM speed output signal is outputting 11.4 (battery voltage) when it should be a square wave 2.5v signal as mentioned by GM tech hotline.
-Scanner does not see speed signal.

I'm at a loss. The original ECM was replaced under warranty but did not solve the problem. From what I have been told by the dealer is they had an extensive check list to complete prior to getting the new ECM covered under warranty, so I am assuming the new wire harness included in the kit from VSS to ECM is ok. I believe the ECM is an E67 from a 2009 Caddilac CTS-V (scanner profile mention in included instructions) Pin 39 in C1 blue connector from the ECM is the speed out
that goes to Bulkhead connector pin D, but If you look up the pin outs for 09 CTS-V ECM pin 39 is not used in a CTS-V...What?

Any ideas what could be wrong? hardware revisions not matching the instructions? ECM firmware? This is driving me up the wall! it was all plug in and play except for the speedometer which is required for CA smog.

See Post #15 for Stock '90 ZF6 VSS output specs. Signal output is also the same for '91-96 VSS, only difference is the electrical connector.


See Post #16 for how to make a T-56 VSS work for LSx/ZF S6-40 VSS input applications!

MatthewMiller

I don't know if your 1990 VSS is the same as the one in my 96, but mine does not generate a single constant voltage. Instead, it generates a variable a/c voltage depending on the speed it is spinning. But mine also outputted a proper a/c voltage even when it was bad. I finally figured out that I also needed to test for continuity across the two terminals on the VSS. It should be roughly 400ohms (at least on my 1996), but mine had infinite resistance - i.e., complete discontinuity.

I would recommend you put an ohmmeter across the terminals of your VSS and make sure it has reasonable resistance: something like a few hundred ohms. If it's several orders of magnitude higher than that (like several thousand or tens of thousands, or infinite) then it is bad, even if it appears to be outputting an a/c signal. They are cheap - I got one from Autozone for ~$35 - and it isn't hard to replace.

MrVette90l98mt

I should have added speedometer worked prior to engine swap. My car and my fsm are at the dealer, what is the official way of checking the VSS generator? I sure hope they have the competence to test this properly. I remember seeing a fluctuating signal coming out of my VSS prior to sending to dealer with my DVOM when I discovered the problem. I didn't check the VSS while the car was running but spinning the wheels by hand generated a varying peak voltage proportional to wheel speed ~.5-.7V for a couple MPH worth of wheel speed. The 2.5V they saw must have been a peak AC voltage or RMS measurement...either way, they claimed to see something that satisfied the EROD system requirements.

WVZR-1

OP - Your problem is getting the VSS signal to the cluster. The '90 VSS first stop was the ECM, then forward (4K pulse) to the devices (CCM then converted to info the cluster wanted to see). Your LS PCM can't (my thoughts) output the proper signal for the cluster and your VSS can't output the correct signal that the LS PCM wants to see.

Your VSS should output something less than 1V/AC to something 8V/AC or greater depending upon the speed. I doubt you have an issue with the VSS and I doubt that you actually had a PCM problem either except that I believe the EROD packages were maybe susceptible to defective PCM.

Is your CCM still in the car?

You need to get the VSS AC sine wave converted to what you need and where you need it.

I've never attempted to do this and I've no idea how you deal with the Cali referees etc.

MatthewMiller

The FSM way will be to connect it to a factory scanner that can process the VSS signal the same way the ECM does. It will verify the proper signal or not. Going from memory, I believe it will actually output a mph readout just like your dash.

I'm a little confused because I am not familiar with the EROD package. But...I am assuming that they mean your VSS is sending a signal that your E67 will accept. And that may be the case. However, as WVZR-1 noted, the ECM doesn't just pass the identical signal to the CCM. It uses a grounding scheme to generate a digital signal that gets sent to the CCM for the speedometer (and radio and cruise control). The output from the E67 ECM will not make sense to the original CCM. The signal isn't the same as your stock ECM created. So unless the EROD package contains a converter to make the CCM understand the new ECM's output signal, that's your problem.

As confirmation of this, is your tach working? The speedo and tach are the two gauges that don't work automatically with the LS-style ECM outputs. If the tach is also not working, that would corroborate the issue. Also, does the engine run without any problems? In particular, when you're rolling and disengage the clutch, does it settle smoothly to a steady idle or does it try to die? If it is not having that problem, then that tells us that the ECM is seeing and using the input from the VSS - if it weren't seeing the VSS signal then the car would try to die when you disengaged the clutch.

MrVette90l98mt

My immediate goal is to be able to get a speed reading off of a scanner, that way I know the ECM has all the signals it needs to be smog compliant and get it reffed and registered. I don't really care about the LCD display just yet.
Yes, the CCM is still in the car.

You're correct in the path from VSS to LCD screen [OEM set up]: VSS (AC voltage) -> ECM (converts to a square wave 4K Pulse per Mile) -> CCM -> LCD screen. I had a picture of my FSM on my phone. Prior to purchasing the kit, I read in the EROD instruction online that some manual transmission use an automatic PPM.

PN: 19257230 Auto trans cal (40T)
PN: 19257234 Manual trans (17T)

I did a lot of maths and confirmed the ZF6 uses a 17.xxx Pulse per revolution VSS which equates to 4K pulses per mile. So I bought 19257234

I believe the LS3 ECM merely converts the AC signal into a 4K ppm square wave signal same as original ECM, However the issue is the pin 39 I'm told to use as speed out is giving me straight battery voltage, not a 2.5v square wave that GM tech hot line said and I don't see a speed reading on a scanner. The original ECM was exchanged under warranty but the new ECM does the exact same thing I also have a dakota digital speed adapter if needed.

snip from EROD PDF:


Pic of My 1990 FSM I had on my phone:

MrVette90l98mt

Forgot to mention the tach works after installing a dakota digital tach box in 4 cylinder mode using the bulkhead pin C that comes from the LS3 ECM: Blue C1, pin 25

MrVette90l98mt

Car runs fine and has never died or seem to run rough during gear change. It idles at about 500-600rpm in neutral while stopped or coming to a stop, never really tried a long clutch roll but during a hard shift when it takes longer to get into a gear it never shut off. Is that a normal characteristic of an LSx, No VSS signal engine will stumble shut off? Now that 'm thinking...Does the crankshaft position verification learn procedure have anything to do with speedo out (code P0315)? OMG, if it is something as stupid as that and GM never told me or documented this in the instructions...i'm gonna lose my S***!. That's the one thing that hasn't been done yet. 500mi break in is complete, the next step is to crank position learn and set the smog I/M flags before getting smogged.

WVZR-1

You're confusing the reference done by a 17 tooth reluctor to an electronic pick up vs the mechanical driven PMG VSS.

There's many things that can be done with an ERA - can you convert the PMG to the later using one? Maybe. Call Abbott or maybe Dakota and quiz them.

MatthewMiller

That's been a normal characteristic for Corvettes at least as far back as 94, and probably considerably further back than that. The ECM takes vehicle speed into account to manage things like fuel delivery on deceleration with the clutch engaged, and to manage the engine's return to idle when the clutch is disengaged while rolling (often referred to as "dashpot" function). If the engine is behaving well when you disengage the clutch while rolling, then the ECM is almost certainly seeing and using the VSS signal. The problem is "downstream" of that.

I can't see how it could. I certainly didn't on your OE setup. The speedo circuit isn't very complicated. It doesn't need to know anything about the state of the circuit to function properly. In fact, on the stock car the speedo will read properly if the engine isn't even running but the car is rolling with the key on (or if the VSS is out of the transmission and being spun by a drill).

hcbph

Here's a thought and only a thought. Is it possible the signal being generated is not compatible with the speedometer? The reason I mention this is my 86 had a Tremec transmission installed prior to my purchase. There's a Dakota Digital converter wired into it between the transmission and speedometer to change the output signal from the transmission to be compatible with the speedometer. Might you be fighting the same issue?

Just a thought.

WVZR-1

OP - I don't believe you can evaluate the OUTPUT until you verify your giving the PCM the INPUT it would like. I think you don't attempt to substitute your thoughts (E67-CTS) for the instructions that were supplied for the package you purchased.

You'll need to use an ERA I'd think to modify what your PMG VSS is producing to the signal the PCM is wanting. I don't know that the 17 tooth was a good choice or not.

Several years ago a friend wanted to do an E-rod and I did a lot of work trying to get a package together. His of course wasn't a ZF and was much easier. He had a "funding issue" so I've no idea how it would have performed. I assume well. At one time I thought the E-rod PCM was locked and there was very limited manipulation possible, the Cali stuff I would have thought maybe even more so. Yes/no?

Like I mentioned I doubt you actually had or have issues. The tech at the dealer maybe didn't understand the older VSS so he passed on info he thought correct and the GM TECH just interpreted the information passed on to them.

Good luck and POST BACK - through the years there's been much misinformation distributed and assumed to be correct for GM VSS.

I believe that you only need to be concerned with the PCM. The speedo I believe would likely be easier and less expensive maybe using a GPS source to control it. There are a couple, maybe more that should be programmable to use the CCM and cluster.

MrVette90l98mt

if that was the case, shouldn't I see a speed reading on a scanner? I get a constant 0mph displayed on the scanner, wheels off the ground in both 1st and 2nd gear at idle speeds which should be ~8mph and ~14mph. On top of that the speed output from ECM is battery DC voltage not a 2.5v square wave.

you're right the speedo circuit is simple. All its doing is:

Input-Hardware and VSS signal output acceptable (dealer claims)
Process-Who knows? Crank verification procedure P0315 skewing / blocking output? ECM not fully functional since P0315 set? Why would pin 39 in C1 (blue ECM connector) output battery voltage and not 2.5V square wave as it should?
output-Have not even bothered to connect it since the scanner does not display speed reading.

I believe it has to be a input or processing issue...

MrVette90l98mt

After doing more research I found an error in my original calculation and did not take into consideration that the VSS on the ZF S6-40 has it's own gear reduction (15T gear on tail shaft driving 42T on the VSS shaft (info sourced from ZFDoc: http://www.zfdoc.com/techinfo.htm)

I picked up a 94-96 ZF6 VSS just to play with as they didn't have one for my year and evaluate its performance with a drill. I took video at 5 different RPMs and used video editing software to aide in calculating the average AC volt output at X rpm. Turns out it is a linear output proportional to VSS shaft speed (as all variable reluctance type of sensors are). I don't know the max voltage output as I have no way to spin it fast enough to find out. I still need to figure out a way to count number of pulses per revolution of the VSS and number of pulse at one tail shaft revolution. I will have to do this at work with an Oscilloscope. I'll bet the pulse counts are different between ZF6 and T56 / TR6060's used in OEM LS3 Manual trans cars.

here is what I found out about the ACdelco 94-96 VSS output:




Question's I need to answer or get answered are:

How many pulses per VSS shaft revolution?
How many pulses per Tail shaft revolution?
What are the differences in VSS sensors 89-93 from 94-96?

I will update this page when I have the info

Thank you all for the ideas and knowledge!

MrVette90l98mt

Problem Identified!!!

I had a chance to play with a VSS on an O-scope when I had free time at work and discovered the pulse count coming out of the sensor is approximately 1/10 of what the LS3 is expecting! I chucked the sensor in a drill, took video of the drill motor with timing marks and the O-scope image in the background. I then used video editing software to calculate the RPM of the drill and get the frequency or pulse count output. Assuming ZFdoc info is correct (15t drive gear on the tail shaft and a 42t driven gear on the VSS shaft 2.8:1 gear reduction on the VSS) the pulse count per revolution of the tail shaft is approximately 1.4 pulses per revolution of the ZF6 output shaft...way off from the 17 pulses per output shaft the ECM is expecting. I got my car back from the dealer and when I'm traveling at 60mph the scanner reads 5-6 MPH.

O-scope results:



O-scope image (94-96 sensor spun at 384 RPM):



Questions I have:

What is the voltage range expected coming out of the O.E.M. 1990 L98 ECM speedo output going into the CCM? (refer to the FSM schematic posted above in this thread)

Can I program new speedometer input and output values using HPTuner software? I will get in touch with various tuner software companies asking this question, but if somebody knows for sure please post.



Update:

Adding everything you need to know about the ZF Vehicle Speed Sensor:

Both VSS sensors for both the early and later type output very similar signals. So close the differences can be attributed to manufacturing variations.

Electrical output:
Both are AC sine waves. If I remember correctly my OEM 90 was more saw tooth like. This doesn't matter as the ECM is counting the number of pulses, not the signal shape. Here are O-scope traces of both type of VSS. These where spun using the same drill at the same speed.

Mine:



New ('91-'96) later type:


Physical differences:

Not much difference here only the connector and construction are different:

Side by side comparison Early ('89-90) is metal case with center connector:








My '90 VSS case Outer diameter



Late type outer diameter:



My VSS gear





I see no reason why the two would not interchange. You will have to change the connector. They are no worries about damage being done by hooking it up wrong as the signal is AC and the voltages are very similar. The worst that may happen could be a potential logic error (signal high connected to signal low) if that's the case swap the 2 and that should fix it if there are any problems. I've seen the early type cost as much as $150 and nearly impossible to find. The later type and easier to find, cost me $50 and it seems to fit. How much do you care about NCRS originality? lol!

MrVette90l98mt

A working solution!

It's been a while but I have a working speedo

It took a fair amount of R&D and here is my solution:

Knowing the Stock ZF VSS was far too slow of a pulse count for the LS3 ECM to understand I had to figure out a way to make 17 Pulses per tail shaft revolution. The EROD harness has a T-56 VSS connector so that was my starting point. Fundamentally the T-56 VSS works in exactly the same way as the VSS in the ZF, the difference is the way it is physically constructed. Both are Variable Reluctance sensors (passive / un-powered sensors) https://en.wikipedia.org/wiki/Variab...uctance_sensor I bought a T56 VSS and started thinking how to place this thing outside of the trans as getting into the ZF was not feasible and I want to keep the old ZF VSS for the FX3 suspension. I found the perfect place: the drive shaft trans yoke. My ICT LSx to SBC engine mounts shifted my engine 5.5mm forward that gave me a little more room to play with. I even think these parts will fit with a zero offset.

My CAD design concept:





3D prototype parts test fit:





Actual tone ring install for a test run (Tone ring MUST be made of a ferrous material {magnetic})






Final assembly for test run (still waiting on the real VSS mount)





HPtuner VSS speedo check at idle, wheels off the ground:

1st Gear idle (3.33:1 - 295/35/18)


2nd Gear


3rd Gear


Can't wait to get the final VSS mount so I can row through the gears!!!...........and then start fixing the mufflers shop's crappy welds causing a fuel trim error

Current Speedometer recalibration vs. LS3 EROD calibration:

My setup with speedometer reclaibration:


OEM LS3 EROD speedo calibration settings:




Conclusion:

The ZF6 VSS has too slow of a pulse rate for the LS3 ECM to understand. YOU CANNOT PROGRAM LS3 ECM TO ACCEPT A ZF6 VSS SIGNAL! The ECM will time out long before it reads an acceptable signal up until you're around 22mph (and that is with a crazy speedo recalibration to something like 50 wheel revs equal a mile). The reason for this is because the LS3 ECM has a minimum VSS pulse count average of 3 pulse to average to count as one pulse (4 is default). This effectively cuts the ZF VSS already slow pulse count even further! This is why my speedo was not working in the first place.

As a bonus the LS3 ECM outputs the correct signal to the CCM and LCD instrument cluster works with no other modifications!!!