[junglejustice]

So, the 164 gearbox has a little gear-driven plastic ring & pinion inside of the transaxle that drives a sender of sorts...

What I would like to know is the type of signal that this generates - is it a voltage or a resistance? A digital frequency, or an analog signal? AC or DC current - what?

The 6-speed 156 GTA gearbox that I am installing in my 1995 LS with the 3.45 litre 24 valve motor, has no such sender. Those cars pick up wheel-speed from the ABS sensors and somehow transmits it to the speedometer on the 156.

What we plan to do to get my 164 speedo working with the 156 GTA gearbox, is to tap in to the 164 ABS wheel-speed sensor(s) and then come up with an electronic converter that will send an accurate signal/voltage to the 164 dash to read speed. I need ideas please!

Say, if the 164 gearbox sender actually generates an AC current with voltage that increases as it spins faster and that is how the speedometer is driven, then we need to somehow take that ABS signal and converted it to a variable voltage...

Dito for the ABS signal that I plan to tap - what is it? Analog?

 

[AlexGS]

Hello JJ,

The last time I repaired a speedometer sensor (leaky caps), I took some photos while I had the oscilloscope on the output. Unfortunately those photos are on my PC at home, and I'm not home for another couple of weeks. I can tell you, though, that the output of the speedometer sensor is a square wave of approx. 8V amplitude (I think it was), about 50% duty cycle, with the frequency picking up to several hundred hertz at 'high' speeds.

I don't know how the AC voltage of a wheel speed sensor is going to compare. Usually there would be a pulse-shaping circuit to make it into a digital signal. Perhaps you need to construct this from the circuit out of a speedometer sensor.

I would have thought that the easiest solution would be mechanical, i.e. to fit a speedometer sensor to your replacement gearbox. The mounting for the sensor drivegear is pretty crude anyway - it's only mounted in the casing, with no support at the base of the drivegear.

Cheers,
-Alex

 

[junglejustice]

Thanks Alex!

Yeah, I would think that drilling and tapping in to the tranny would be the easier part. It's replicating the plastic ring-gear on the differential that would be harder to replicate.

I'll go for the electronic conversion for now! Thanks for the numbers.

 

[AlexGS]

Maybe, to save interfering with the ABS sensors, perhaps you could do something dodgy with the intermediate driveshaft along the back of the engine. I think that some old cruise-control kits relied on a sender attached to the driveshaft of a RWD car. Or, I wonder if you could incorporate the guts of the speedometer sensor (which is a magnetic-reluctor type) with some kind of rotor on the driveshaft - having the correct number of blades to give the desired number of pulses. Though, I suppose it would lose accuracy going around corners - but then, that was going to be a problem with the ABS-sensor solution as well.

Sorry that I haven't really given you any certain numbers at all - when I get home in just under two weeks, I can be more helpful. Two weeks probably isn't long compared with the rest of your project - six months now?

I know. What about a way to calculate the speed from the engine RPM - with stored samples so that the PIC microcontroller can 'tell' when you make a gearchange, and apply the correct ratio for the new gear. The pulse that drives the tachometer will be easy to measure, plus a signal could be taken from an ABS sensor to provide some indication that the car was starting to move...

...though, I guess having done that, it will be easier to have the PIC microcontroller work with the ABS signal instead. I think you would have to program a loop that counts five incoming pulses (each time waiting for the input to go 'high' then 'low') and then outputs one pulse - something like that ratio. Hopefully the input will be of a high enough impedance (10k resistor in series?) not to affect the ABS ECU.

-Alex

 

[Alfisto Steve]

Just brainstorming here.

Does the 6-speed differential case have a access hole for a sensor?

How possible would it be to adapt Milano pick up sensor that reads off ring gear teeth and uses amplifier to send signal to speedo.

Or even a Spider flywheel RPM sensor or a 164 crank pulley RPM/timing sensor?

Another idea would be 156 sensor mounted near ABS sensor to read ABS serrations off axle.

 

[junglejustice]

Hmmm.... All great suggestions - no short-cuts though!

I like the thought of a steel ring-gear on the half-shaft with a 164 mag pick-up mounted close to it to measure rotations - would be a bear to get calibrated though and we learned from the custom crank-pulley triggers with programmable engine management systems that the trigger wheels typically need to be pretty big - the teeth need to be large!

I'll be back at the shop later this afternoon to pick away at it some more and then I will look around some more for potential access holes! I'll also remove the sender from the old auto tranny and bring that home - my brother has an oscilloscope and all kinds of electronic toys at home - maybe we can figure something out!

 

[andrewg]

The diff casing is different from the 156 to 164 transmissions. There is no hole in the 156 case. The 156 diff does not have the provision to mount the plastic speedo drive. The thought of using a 156 unit would be that since the ABS generates a different signal, what are the changes that the 164 speedo will correctly understand that signal. There will be a lot of trial and error to get this issue resolved.

 

[junglejustice]

HEY!!!!!!

*** - we're supposed to be working right now!

Call me!

 

[andrewg]

If you would get up before lunch time then there would be time to work on your car!!!

 

[junglejustice]

So, I would like to revive this thread and move this part of the conversion forward. Any further input would be much appreciated.

Myself and at least 3-4 other guys all need to come up with a solution, to run the 164 speedometers with these 156 boxes. I need some direction please, since the car is also throwing a 1252 (no speedometer input - I think it is?) error code as well...

As a reference, this thread kinda goes hand in hand with the following thread; http://www.alfabb.com/bb/forums/164-168-1991-1995/53427-1252-error-code-24-valve-motronic.html

Any way, I guess that the first question I have is - does the ABS signal(s) from the wheel-sensors make it all of the way to the ECU and if so, what colours are the wires at the ECU (or where else can I pick it up to test with?) I'll start by measuring the signal while driving to see if it is variable and what it is. (Maybe put an o-scope on it and run it on the rack or something?)

Next I am going to remove a sender from an extra tranny that I have sitting here, spin it on a drill and see what range and type of signal it puts out. Any better suggestions?

 

[jontg428]

I have a better idea.......

Try this out man, should do the trick!

 

[Alfisto Steve]

As I said in other post first off instrument cluster sends signal to speedo sender in tranny and then that signal which is created from that sender spinning is sent back to speedo and then on to various ECU's for engine, A/T tranny, ABS and Cruise control.

Now how do you send a like signal sine wave from 164 ABS system to simulate tranny sender signal in a 164 is the $164 question. I believe we are talking apples and oranges here.

 

[AR4me]

Here is another thought: Why not get a 156 speedometer/cluster, take the electronics for the speedo out and put it into the 164 speedo/cluster? This assumes, of course, that the 156 and 164 ABS wheel speed sensors generate a similar signal, i.e. that the 164 ABS wheel speed sensor can drive the 156 speedo.
Jes

 

[Alfissimo Int.]

Here is another thought: Why not get a 156 speedometer/cluster, take the electronics for the speedo out and put it into the 164 speedo/cluster? This assumes, of course, that the 156 and 164 ABS wheel speed sensors generate a similar signal, i.e. that the 164 ABS wheel speed sensor can drive the 156 speedo.
Jes

This is a pickle for sure. The problem I see just from a general view is this, the ABS sensor runs to the ABS control unit then out to the speedometer. So I am not sure if that signal which is most likely the same type signal the 164 puts out is changed to something else when it runs from the sensor to the ABS control unit out to the speedo.
If it stays the same then you can possible get the ABS sensor and wire it in where the speed sensor for the 164 is obviously!

problem I see wit the 156 speedo is the clusters are coded. May possibly be similar and can swap it for the speedo on the 164 but I would not be sure it's accuracy. Although you could take it to a speedo shop to calibrate.

Hope that helps any or not?!>

 

[evilgidget]

156 speedo wouldn't work. From memory, the 156 signal goes to the ECU, then gets converted to a signal for the speedo. 164 seems to do it the other way.

I like Steve's idea of drilling the case to accept a sensor to read diff teeth, then use the amplifier. It'd be a doddle. Go have a chat to your local TV repairman or radio enthusiast - should be able to knock up something that'll work in a few minutes

 

[Alfissimo Int.]

156 speedo wouldn't work. From memory, the 156 signal goes to the ECU, then gets converted to a signal for the speedo. 164 seems to do it the other way.

I like Steve's idea of drilling the case to accept a sensor to read diff teeth, then use the amplifier. It'd be a doddle. Go have a chat to your local TV repairman or radio enthusiast - should be able to knock up something that'll work in a few minutes

Yes, that seems like the solution to me as well.
Good luck!
Jason

 

[MrT]

In my humble opinion, the lazy man's solution would be to learn what the 164 speedo sender signal looks like (probably low voltage inductive pickup pulse train, as the sender looks like an inductive pickup similar to the crank sensor) and then build a simple electronics box to translate the wheel sensor pickup signal into a form that the speedo/ECU can use. The ABS box is in the console with the ECU, isn't it? It's possibly even as simple as a digital divider to multiply by the tooth ratio between 159 ABS pickup and 164 speedo pickup. Or so it seems to me. Is Steve certain that the ECU sends some kind of drive signal to the speedo sender? If so, then this might be more problematic. But it seems like it should be a tractable problem.

Michael

 

[Alfisto Steve]

You using my name? If so correction. I said sender at tranny gets 12v voltage from instrument cluster and sends about a 3.5 volt signal in a sine wave back to speedo and signal then goes to ABS, Cruise and motronic ECU then on the A/T tranny ECU if automatic equipped. I listed page in 1994 wiring manual stating that above.

I didn't say anything about drilling hole in differential that I remember. Much more to it than that to use a 164 sender because sender driven by geared pinion shaft that would need a drive ring gear on differential gear to drive it.

Now an idea that just popped into my head would be to investigate using Milano sender that "feels" movement of ring gear by sensing gear movement. Look at using Milano sender and drilling a hole to install it since no gears needed and then add the amplifier that is under rear seat of Milano.

NO THAT IDEA IS STILL TO HARD AND OLD TECH.

NOW another idea you maybe could look at is using something like engine rpm/timing sensor or CAS sensor or even a seperate ABS sensor mounted on a seperate bracket next to ABS sensor to read serrations on axle so you can send 12v signal and get feedback signal. Now whether sine wave the same I have no idea. You need a scope to see.

 

[MrT]

correction as requested

Howdy, Steve. What I tried to say:

Is Steve certain that the ECU sends some kind of drive signal to the speedo sender?

was that if you were certain that the speedo sender requires some drive power (like 12V) from a source (which I _thought_ you'd said was sourced from the ECU but what you really said was),

instrument cluster sends signal to speedo sender in tranny

then making a simple pulse translator box might not be the full solution. I also said I didn't know the signal requirements, and that the first step to doing something would be to find out what the signal actually needs to be. Just making comments and trying not to obscure stuff.

Michael

 

[junglejustice]

So, a brand new direction here and yet - an old and familiar road for some I guess... Those of you who have gone down the path of installing a Milano transaxle in a GTV6 have undoubtedly been faced with a similar dilemma - getting the GTV6 speedo to work!

One of the solutions out there is to take a pulse converter from Dakota Digital, hooking it up to the signal produced by the Milano transaxle (or any gearbox pulse generator for that matter) and simply calibrating it to drive the speedo in the GTV6. Well, this is not that much different I guess.

They pull signals off of flywheel mounted pickups (like on some of the Spiders) to run tachometers, they pull signal from ABS toothed-trigger wheels - anything that will rotate and create a magnetic field! They said that tapping in to the ABS signal from one of the wheel-sensors would work fine, but they don't like the potential liability of doing something that may impact the brakes and therefore prefer to sell you a separate "motorcycle" pickup (60 bucks) to mount "on" (close to) one of the ABS wheel gears to produce it's own signal!

You feed that signal in to their SGI-5 converter (80 bucks), calibrate it and should be good to go. Dunno - we'll see - I ordered the stuff...

I also pulled the little generator off of the old tranny tonight, plugged it in to the 2-pin signal connector, spun the little square input cable by hand and the European manual 164 Q speedometer bobbed up and down, so we should be golden! I'm just going to feed it signal right there and let the car/dash/ECU/instrument cluster/main harness do the rest...