Driving 97 kick tach after J18A engine swap

[maroka]

I guess I should have been more clear, sorry. I have a MK2 Golf which I intend to engine swap with a VW ABF 16V engine, so that's why I started digging around trying to see how people may have sorted the original tach to work with the Speeduino/Mega and COP. I should have thought about that a lot earlier, but I am happy to admit that my electronics knowledge has expanded since opening this thread (Mostly thanks to you and all the info you share). I have ordered one of these Chinese scopes, but it says it's still in transit and I should expect it within a few days. When I get it, I will make sure to check out what kind of duty cycle, frequency and voltage level the ECU broadcasts on the tach line. I read on a forum thread about VQ swaps that a guy was trying to tap into the "RPM Output" wire of the ECU Going to the Power Steering control unit and posted a waveform that had 0-4V levels and the frequency corresponded with the amount of power strokes for the V6, so my guess is that the cause here will be similar. I'm Still kicking myself for not thinking about that earlier (well I did build that circuit with the arduino, but the "tone" library might not have sent the correct types of signals on the output pin).

I will try just the transistor with a current limiting resistor instead of the relay. If I understand correctly, the tach will be seeing a 12V signal, until the transistor is switched on by the "tach" wire on the J18 ECU, then it will pull the 12V to ground through the resistor and the tach will see 0V.

[fixkick]

I guess I should have been more clear, sorry. I have a MK2 Golf which I intend to engine swap with a VW ABF 16V engine, so that's why I started digging around trying to see how people may have sorted the original tach to work with the Speeduino/Mega and COP. I should have thought about that a lot earlier, but I am happy to admit that my electronics knowledge has expanded since opening this thread (Mostly thanks to you and all the info you share). I have ordered one of these Chinese scopes, but it says it's still in transit and I should expect it within a few days. When I get it, I will make sure to check out what kind of duty cycle, frequency and voltage level the ECU broadcasts on the tach line. I read on a forum thread about VQ swaps that a guy was trying to tap into the "RPM Output" wire of the ECU Going to the Power Steering control unit and posted a waveform that had 0-4V levels and the frequency corresponded with the amount of power strokes for the V6, so my guess is that the cause here will be similar. I'm Still kicking myself for not thinking about that earlier (well I did build that circuit with the arduino, but the "tone" library might not have sent the correct types of signals on the output pin).

I will try just the transistor with a current limiting resistor instead of the relay. If I understand correctly, the tach will be seeing a 12V signal, until the transistor is switched on by the "tach" wire on the J18 ECU, then it will pull the 12V to ground through the resistor and the tach will see 0V.

this level of hacking is complicated by the fact car makers mostly do not cover this topic at all , swaps.  nor signals told, some on purpose (odometer laws)
notwithstanding silly PCM pages with key on voltages, only.

The inputs on most ECU/PCM are limited to 15v max.  most  do this 2 ways,  with series resistor (protecting) and CAP filter for noise. and some have zenor diode clamps.  cars are nasty noisy things happening, like AC clutch coil noise, and protection from that and overheating,top the list.
sure scope is best in all cases to see what really is there,  and can be vast things possible.  (even data streams)
But Id not ever send a negative signal in to the PCM  nor go over  +15vdc (alternator max) or risk damage,  and yes the coil neg lug is nasty and is in fact filtered.

BACK EMF is here and off the scale;G16 1991

the 91 this protection and there is no need for the 100v spike to make it work

first note the 2k resistor in the Suppressor  module that reduces this PIKE first. EMF spike. The D110 kills the negative component of the signal and the R106
forms a voltage divider and 2vdc to the fail logic deeper inside (port expander IC chip.? )
The key point here is that the input spick and ringing are not to be set to the logic inside just the 12 to 0volt pulse. and becomes in this one case 2v pulse.
Most store bought Tachometers have complex input selections, not just 2 or 4cycle and cylinder counts, but levels too.
Im not sure what you are doing with VW parts at all but this is how some tacho lines look like , 1991 8v here, G16a. just to show how Suzuki did this clean up.


ON most COP based engines, you have 2 or 4 coils for 4 cyclinder engine, PCM driven.
how the PCM creates a tachometer signal is any ones guess, and I'm sure not all cars are the same.

Most would be a 12v or 5 volt square wave, pulses.  how many per crank turn is no standard either,  the cluster designer talks to the engine EFI designer and they decide that and keep it   a secret but the scope in hand tells the truth, what they did. pulses per Crank turns. CPCT?

On my 2008 jeep it is all serial data, on the CAN BUS. (a network for cars) some signals are encrypted too. yup,  good luck hacking that LoL.
In effect my cluster head is just 1 more network appliance (device) with 3 pins, 12v, ground, and CAN.
I'm sure no GOLF has that. 1983 to 1992 made.  but  Germans do as they want.
NOT saying I know a thing but the filters used yes, and what not to do, best is avoid any negative signal and not over +15vdc, Alternator max.
The links you posted are from folks just  trying so hard to emulate the huge noisy coil signal,  and I would never to that,  id use NO COIL at all .
that 400v spike on my car could blow up vast things easy, for sure other cars parts hacked in or clusters or OTC over the counter tachometers.

try a clean 12vdc pulse. is the most safe way to do this.

good luck with your project.

[maroka]

Hey there!

So I tried hooking up 12V to the tach and a mosfet (since that's what I had on hand), as a switch, pulling the tach wire to ground through a resistor when turned on by the ECU signal wire. The needle does jump a tiny bit only when cranking, as soon as the engine starts, the needle stands still.

My Chinese scope has arrived too, but I don't really know how to use it yet. I wrote a simple arduino program using the "tone" command. It should output a 5V signal at a set frequency and 50% duty cycle. I set it to 250Hz for the test and when I hooked up the scope and pressed the "auto adjust" button I got some funny results
here:
https://photos.app.goo.gl/rRjqmkx7gT6qkiPp9
Why does it display a negative pulse and nothing near a square wave? I also tried manually setting the pin to output 5V, then 0V with a manual delay and I got pretty much the same result. Either this scope is crap, or I can't set it up right.(I'm leaning more towards me being dumb).

What could be causing that?

[fixkick]

Hey there!

So I tried hooking up 12V to the tach and a mosfet (since that's what I had on hand), as a switch, pulling the tach wire to ground through a resistor when turned on by the ECU signal wire. The needle does jump a tiny bit only when cranking, as soon as the engine starts, the needle stands still.

My Chinese scope has arrived too, but I don't really know how to use it yet. I wrote a simple arduino program using the "tone" command. It should output a 5V signal at a set frequency and 50% duty cycle. I set it to 250Hz for the test and when I hooked up the scope and pressed the "auto adjust" button I got some funny results (so this is arudion out put)/? if yes then the driver in the Arduino is set wrong,
i know that arduino and computers like this very well,  and there are many output modes in the Arduino possible
what works best is not known yet , impossible to know  at all; you need to do the below tests first then with those facts learned we design the ardiuno sketch to do the right thing, both signal type and polarity  and all that.
We even do not know the output of the current cars tacho signal at all
nor what the meter wants. 2 unknowns.

the real problem is there are 2 devices here,  old PCM to old tach  and not a different engine,? to different meter>? IDK>
I am not clear at all what this means
"MK2 Golf which I intend to engine swap with a VW ABF 16V engine,"  what does this have do do with anything suzuki? IDK?
are you saying there is no J18 engine, no J18 PCM , no J18 car here?
What car is the target car.? what is the donor card, all parts moved tell.
what PCMs are used and what tach meters used in target car
none if this is clear at all. and very very confusing.




here:
https://photos.app.goo.gl/rRjqmkx7gT6qkiPp9
Why does it display a negative pulse and nothing near a square wave? I also tried manually setting the pin to output 5V, then 0V with a manual delay and I got pretty much the same result. Either this scope is crap, or I can't set it up right.(I'm leaning more towards me being dumb).

What could be causing that?

to convert from one simple out put to another device,  simple one must know 2 things.
The ECU sends the tachometer output
and the actual tachometer input wants some other unknown signal.
the problem is both are alien to each other,
first we scope the car ecu first. tacho pin,

those narrow spikes are buss capacitance changing voltage states,  means wrong settings on i/o port  of the Arduino
but we need to answer the above green questions first.

the path is simple (pure electronics only)
the tacho meter must get the same signal it got off the old cars PCM. in the exact same way , voltage , polarity and wave form type.
the conversion can even be simple in some cases but we do not know what the tachometer wants yet.

more later.

[fixkick]

Ardunio digital I/o ports, not analog
this  is first, for sure, programming (sketching?)
the I/O mode set is key to success, if wrong this then your code is worthless (the input and outputs both must be right)
we did this at Intel (I worked there for 6 years as tech) with 8255 chip, way ahead in time then. (cutting edge)
(flow chart, linear: input wrong pcm ,>> process signals >>> output what new TACHO meter wants to see.)
this includes polarity and wave form shapes and even duty cycle  and if we must count up or down to make the tachometer work.

this post is what to do after learning the original PCM outputs(tacho out).

https://idyl.io/arduino/how-to/manipulat...ino-ports/

the output pin  can be set to totem pole outputs or open drain ,  the latter needs pull-up resistor work. Totem Pole outputs are super clean square waves by design.
in all cases for you the totem pole out is best.
and input pin set to input only and with pull up mode set,
the arduino comes in 2 forms, 3.3 volt outputs or 5v.  but there work on 12vdc input tachs, see?
so a buffer is used, to do that . (we will do that later, after we know what the signals mUST to work into the unknown tacho meter of any kind.
best is the 5vdc Arduino this is because 5v outputs can more easily drive,  the gate on many powerful Mosfet transitors.
no idea what Arduino you have.

good write up is here, many others are weak. See how GPIO pins work .(super good in fact , even teaches defaults on pins) wow.

see this on making Arduino  to drive powerful  singles or higher voltages than 5v at 40mA.

this is getting way ahead of my self and we need to use the scope first on the original ECU > Meter, (tacho) to see what is THERE.
guessing will not work and that is where the scope comes in.

the best way to Sketch is start simple  write a loop that toggles and output pin and scope it , then you know it works and sketch on from there.

end Arduino things.

[maroka]

Sorry, I am using an Arduino Nano. I just tried it on the d13 PIN (the one with the integrated LED) and I get a beautiful square wave output! The scope works! Tomorrow I'll get it hooked up to the tach output on the ECU and I'll see what we're going to get there!

https://photos.app.goo.gl/QT8PNQSviEZt5MmP6

[fixkick]

i can not at all understand this vw , swaps , I am totally confused on all that,
donor car and parts moved to what target car,, all part moved.
swaps are not easy for sure using wrong tachometer on target car.

[maroka]

i can not at all understand this vw , swaps , I am totally confused on all that,
donor car and parts moved to what target car,,  all part moved.
swaps are not easy for sure using wrong tachometer on target car.

Sorry about that, I mentioned the VW just because I have one and I want to swap in a different engine in it and run a standalone ECU. Because of that, I had started digging around the internet on how my VW's tach can be driven by the standalone. Then I remembered that the Vitara's tach could very much be driven the same way. 

I guess it wasn't really necessary mentioning the VW, since it doesn't have any relevance to the matter at hand what so ever, but I just thought that I could throw it in just to be more clear about my thought process. I ended up just causing confusion. I'm sorry about that.

To recap, the Vitara has its own original tachometer still in place, I need to drive it using a signal from the J18A's ECU. The donor car is a Baleno 1.8GTX, the only things from the Baleno are the Engine and ECU, everything else is stock 1997 Vitara.

[maroka]

First off, happy holidays!

Finally! I was able to hook up the scope to the tach output on the ECU. It seems to be sending out Vbat pulses at every ignition event (~30hz at idle, 2 pulses = one revolution, so divide by 2, multiply by 60 to get the revolutions per minute ~~ 900RPM at idle). It seems though it's not anywhere near 50% duty cycle, in fact the scope says it's mere 5% at idle and it increases as the RPM go up. Though if I recall correctly you said that the duty cycle does not really matter to the tach.

Video: https://photos.app.goo.gl/AzgY9mYyDX2EuC9J8
The signal gets fuzzy at the end because of the crappy ground, I just put an alligator clip on the metal on the clutch pedal and moved it on accident.

This confuses me though. It would seem that the Tach does not really like the 12vDC.

Do you have any thoughts on why that could be?

I apologize if I sound dumb or my statements don't make sense. For me it is a great chance to learn about electronics and I am making progress, albeit kind of slow.


Now I'll curve the topic a little bit.

My friend has a MK1 Golf with a 16V engine out of a MK3. The MK3 has a single coil with a distributor. In order to get the tach working we were told to hook up the wire to the ground pin on the 3 pin coil connector. That did not work, so another guy told us that there is a little cover that we could remove and that uncovered 2 solder points. Though I don't remember which one we soldered the signal wire to, the fact is that the tach starter working off of that wire. Turns out, the MK3 has the coil driver built in, so the ground on the connector is "filtered" (I guess) and that wasn't sufficient to drive the tach. The solder point under the cover I guess is the actual coil ground before it gets filtered by the driver circuit.

My point being:
Is it possible that the Vitara's tach is configured in a similar fashion?

I came across this: https://www.pirate4x4.com/threads/faq-2-...x.1003571/
You have probably seen it already. I am not too familiar with filter circuits, so I can't say what role the resistor plays in that particular circuit.

Once again I want to thank you for all the knowledge shared so far!

Enjoy your holidays!

[fixkick]

First off, happy holidays!

Finally! I was able to hook up the scope to the tach output on the ECU. It seems to be sending out Vbat pulses at every ignition event (~30hz at idle, 2 pulses = one revolution, so divide by 2, multiply by 60 to get the revolutions per minute ~~ 900RPM at idle). It seems though it's not anywhere near 50% duty cycle, in fact the scope says it's mere 5% at idle and it increases as the RPM go up. Though if I recall correctly you said that  the duty cycle does not really matter to the tach.

Video: https://photos.app.goo.gl/AzgY9mYyDX2EuC9J8
The signal gets fuzzy at the end because of the crappy ground, I just put an alligator clip on the metal on the clutch pedal and moved it on accident.

This confuses me though. It would seem that the Tach does not really like the 12vDC.

Do you have any thoughts on why that could be?

I apologize if I sound dumb or my statements don't make sense. For me it is a great chance to learn about electronics and I am making progress, albeit kind of slow.


Now I'll curve the topic a little bit.

My friend has a MK1 Golf with a 16V engine out of a MK3. The MK3 has a single coil with a distributor. In order to get the tach working we were told to hook up the wire to the ground pin on the 3 pin coil connector. That did not work, so another guy told us that there is a little cover that we could remove and that uncovered 2 solder points. Though I don't remember which one we soldered the signal wire to, the fact is that the tach starter working off of that wire. Turns out, the MK3 has the coil driver built in, so the ground on the connector is "filtered" (I guess) and that wasn't sufficient to drive the tach. The solder point under the cover I guess is the actual coil ground before it gets filtered by the driver circuit.

My point being:
Is it possible that the Vitara's tach is configured in a similar fashion?

I came across this: https://www.pirate4x4.com/threads/faq-2-...x.1003571/
You have probably seen it already. I am not too familiar with filter circuits, so I can't say what role the resistor plays in that particular circuit.

Once again I want to thank you for all the knowledge shared so far!

Enjoy your holidays!

For sure easy (IF)
if current signal is known from newer ECU or older if be the case.
and what the stock TACHo meter wants to see, we all know there is no standards here. ever.  the oem does what they please end to end.
but  3 to 12vdc signal is common, and on ours a filter so the 100s of volt back EMF pulse does not wreck the tachometer.
on COP  engine , they do not use the COP huge pulse they just drive a 5 vdc signal out of ECU called tachometer. out.  no nasty HV pulses there.
and in the right polarity too. (and duty cycle,  our old tachometer is analog and works on the total energy of the 12vdc pulses. (integrator)
the more pulses per second there are the more RPM is presented (a DC voltmeter it is) , crude as crude can be. (but cheap)

we do not know the , 1989-96 G16 sure do. but not the others. well not me, but I can scope them all an photo my screens and learn any car made.


nobody has a scope so all they do is guess. and guessing can work. wild.

my guess is the 6k ohm reduction is the input resistor voltage divider.


that makes the signal stronger to a wrong tach and  now works.  

he changed the input circuit , in the tacho.

the PCM is must  simple transistor, some cars do not use totem pole driver only and open collector transistor and must have  resistor to 5vdc or is dead.
they did that because nothing is more cheap than that method.  so adding pull up to 5vdc can cure a dead tach too.
or the tacho already  has that resistor.
in fact the one he changed (no schematics) may be the pull up to 24k , made the signal stronger. and worked,  of the 2 ways it could.

good luck finding what works.