Login Register

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Driving 97 kick tach after J18A engine swap
#11
All that information is overwhelming to say the least! But it's really helpful. It seems that if I want to make something that works I would need a scope. I looked up some boat anchors Big Grin as you call them on the local websites where people sell their stuff, but they seem to be asking over $200 for some old Russian scopes that have a screen the same size as a calculator's.

Then I found some guy selling these Chinese scopes for a very decent price:

https://www.aliexpress.com/item/32878694...web201603_

They are rated for 200Khz but he says in his listing that for sine wave they seem to be around 100Khz max. It doesn't hurt to try, so I might get one of them.

You gave me way too many tips, almost feels like cheating! Big Grin

So first order of business is, make the needle move. I seem to have lost the arduino sketch that I wrote to generate the frequencies, so I'll try to write up a new one, and hopefully test it with a scope to make sure it works!

I am kind of familiar with interrupts, since I tried to build an arduino rev limiter a while back which used the time between 2 interrupts to calculate the RPM at any given time. I did it on my MK2 golf which had an RPM output pin on the coil itself which was about 7v so I used a resistor divider and it worked fine.

The J18 ECU has an RPM output pin, I guess it doesn't hurt testing it with a scope to see the frequency at a given engine speed. I might use that as an input to the circuit.

So, make the needle move, then make the arduino count the rpm correct, then make the arduino drive the tach correctly. I better get to work! Thank you so much for the help so far, but I'm afraid I might need some more. I'll keep you updated!
Reply
#12
swaps are not easy . many things there are that Suzuki never documented, at all.
good it has an RPM output pin, that is good news...
so that needs scoping first.
there are many ways to drive this pin (ECU side)
open drain or collector transistor. or? and with or with out internal pull up resistor.
a full totem-pole driver .(best)
it might have a series resistor on that pin too, but can be discovered easy with any scope even that china one (Pico brand like) 50v inputs peak not too bad beats 5v max.
The output can be 5v or or 12vdc and can be current limited at 1mA 10 or 100MA.

a2/a10 tacho out, here? 97. j18 sport.
https://fixkick.com/ECU/18L/18LiterECU1.pdf

What to do is use MOSFET or 2 in on the pin, to buffer that single to make 12vdc sqaure wave to the TACHO 97 stock
if the pin is open drain say then it looks dead, until you add say 510ohm resistor to 12vdc (ign. power ,key on power)
then connect that to tacho. direct, see if tacho works if not scope the pin under load, see you get 12vdc to 0v swings. if not we fix that first ask.

we do not know now the source impedance of the ECU tacho out pin nor input impedance of the tacho 97 but we will learn that easy as we go.
doing simple tests. using 1 scope and a box of test resistors for loads. (and the real connections too)

best i can tell is the tachometer must not have lower that 10,000 ohms impedance. (DC resistance same)
that makes for no greater loss of 12vdc square wave never blew 10vdc square wave.

this make this easy, from that view..
even 1mA drive can do that. 2k source impedance.

a scope tells the signals polarity
after seeing that tells use whether to use 1 or 2 buffers to make it work if direct fails. see? 1 is inversion ,2 is non inverting buffer home made. 0v is dwell time charge COIL so we make sure that polarity lands on the tacho pin, I would guess the tacho pin is 4x rate already but IDK, that, the scope does tall that,
voltage swing see.
see polarity
see pulse rate, all 3 at once.
http://www.fixkick.com
Reply
#13
did you make an IMMO delete, my guess a black box that sends a serial stream (RS232) code to unlock the immo in the PCM?
I know Rhinoman did this once.
http://www.fixkick.com
Reply
#14
tests
see what the PCM tacho pin does naked to scope probe tip. dead nor not.
if dead we add a pull up resistor 1k ohm is good to 12vdc source. (ecu power or ignition on, key on DC power)
if now you get nice 12vdc square wave we look at polarity.
check the if dwell pulse is at the 12 volt end (inverted wrong) or at 0vdc end correctly. as it must be.
if at 12vdc end this dwell pulse we must add 1 more transistor outside ECU on that pin. (1 transistor and 2 resistors ,$1 cost)
this if needed buffer will driver the 97 old tacho great.

last is PRT, pulse repetition time. of the negative pulses seen. we must make this match the 97 tacho to be seen as 1 coil spark engine not 4.
I do not know what your ECU wil do here. for PRT. Ive idea at all what this ECU sends on this pin but is fun to learn. The old tacho needs what a call 4x pulses.
4 spark pulses that charge a capacitor in the Tacho head, very crude but works.
http://www.fixkick.com
Reply
#15
if the PRT is wrong, we then do a 1:4 scale'r or as needed , seen on the scope.
see?
the scope is the key , for all interfacing, like this.
http://www.fixkick.com
Reply
#16
You seem to have guessed the pin right as the color also matches. But the pin numbers don't match up, not only in that schematic. They don't match my ECU in any schematic I've ever looked at for these engines. I used a write up on some forum I found, I still have the file, so I'll post it. It seems that the ECUs for the Balenos had their own pinouts which do not seem to be documented. It's not spot on, but it matched 90% of the pins on my ECU.

About the immo delete, yes I used an emulator. You also have to mod the contents of an EEPROM on the board for it to work. My guess is that the emulator just prevents the contents of the EEPROM from being changed. I still get an error 83 and a CEL which is kind of annoying. I've read that on the plastic box Denso ECUs, you can just remove this EEPROM and the car starts and runs fine. Some people report having to disconnect the Immo box, some say you can leave it plugged in. In my case, I just didn't end up running the data wire from the J18s ECU to the Mitsubishi immo module, so you could say I left it unplugged. And many people also suggest unplugging the module. I might try adding this connection later on just to see if the car still works and whether the CEL disappears. It seems that the wire is shared between the ECU, Immo box and OBD connector under the dash, I guess it's some sort of a bus.

I am currently out of town but will order the scope immediately after coming home. I am really curious as to why this ECUs rpm output signal didn't directly drive the tach.


Attached Files
.docx   baleno wiring.docx (Size: 13.82 KB / Downloads: 0)
Reply
#17
Update!

Won't be able to test very soon since the J18 spun rod or main bearing, don't really know yet. 

A little background. I fitted an LPG Injection system to it, since on the old engine it had just a basic mixer system like a gas stove. The new system has an ECU, seperate injectors for LPG and it needs tuning of course. So we went out on a drive so I could tune the LPG software. At one point I flipped it over to Petrol and after a little bit it started pinging like mad, as if you dropped a box of nuts inside the engine. Switched it over to LPG and it stopped (because of the much higher octane), that's when I noticed that the Reducer's temperature is around 50deg, should be 85 as the engine. That's what tipped me off and we decided to stop, lift the hood, and sure enough one of the rad hoses had popped out and leaked pretty much all of the coolant out. Throughout the whole thing, the temp needle on the dash didn't move from the center. We drained the oil and it had tiny flakes of metal inside it. We waited till the next morning to pour some water into the engine, put some new oil in it and started just fine. After it warmed up, the oil press light started blinking on idle, so we knew it's one or more bearings. We're pulling the engine out to fix it and should be able to test the ECU tack pulses within about a week I hope! 

One odd thing I noticed, whenever the engine is warmed up, and we try to start it, it just doesn't want to turn. It starts turning and as soon as it gets to the first power stroke it hangs for a little bit, then the next, like the battery is flat, but it's fine. So I'm thinking the ECU is advancing the ignition way too much to the point where the peak pressure occurs before TDC so it's trying to spin the engine backwards. I have no clue why it's doing that though...
Reply
#18
the stock temp gauge is a 1 wire sensor (thermistor) now the horror of that, (location location location wrong)
horror 1: the water level drops.
H2: then the sensor touches AIR and guess what it CAN NOT measure air temps at all and for sure not touching coolant at all ,just steam now (horror 3)

never trust that gauge. and the hot engine ping is first clue as is steam.
best is get real gauge, on that car. for sure old cars. I think the lower rad hose is best there with inline hose coupling for sensor mount. (summit racing.com)
that is the cool side but at least water is there full time and to the near end.
a case can be made (race cars do this) 2 gauges, one for top of rad tank and 1 for bottom then you have truth.

spark question when it will not crank,
1: disconnect coil packs ,cranks now/ spins fast now, then is spark timing way off,
check spark timing, check that rear of CAMshaft CMP cam angle sensor is not loose, make sure chain did not slip, wrecking all timing. with low oil pressure the cam chain tensioners are off line, and now chain is very loose and if old and warn can jump. off time.
2: jumper cable start it. (2 batteries now) My engine uses 100amps to crank yours be like 200max, the cheap battery does 500. so...
3: turn engine by hand see if it feels right.

my G16 do 50psi cold engine cranked (spark killed) oil pressure, make sure cams look ok. if this engine has that nasty VVT? variable valve timing thing on end of cam , it can fail.
engine is interference, pistions do hit valves, timed wrong, for sure. (new in crate engine does)

id be all over oil pressures. make sure its even ok no spark at all. just cranking.
may be cam tensioner blew up opening up the oil gallery and just that bad. (even cross fingers for simple cheap thing bad)

they have locks for starting zero RPM tensioner piston is locked inside and when oil pressure comes up , it unlocks, and is now dynamic tension. (so cranking will not sound HORRID chain slap)
see these pistons inside the tension'er. may be a total wreck leaking oil like mad.
engine out is best way to cam front cover off it.
the silly front cover is RTV sealed. and not easy to get a good seal up after it is pulled. (take extra care here)

(RTV= silicon caulk, color coded,for engines) mine is orange, color. (oxy sensor safe etc)

keep an open mind do not blame main or rod bearings first. find the leak inside, all places possible. Rods/mains last.

hope to you , something simple wrong, cheers to you and good luck in your quest.
http://www.fixkick.com
Reply
#19
Thank you for being so active on here! It feels so nice knowing that there is someone with a lot more knowledge and is also willing to share it!

Now on to the fun part.

Yesterday I pulled off the wiring harness to help my dad, but I had to drop off the missus at her parents' house, so I had to leave. My dad pulled the engine today and took off the oil pan. He said there were tons of debris stuck to the bottom of the pan and most likely they are inside the engine now, so a good rinse is needed. Bearings are shot, he said little to no scoring on the cylinder walls, crank is fine too. The bearings seem to be the same part number as the J20 and H25, so I hope they won't be hard to get. I will tell him to check the stuff you mentioned too. Just because there was one problem, doesn't mean there couldn't be more!

My guess as to what happened. My dad took a part of the bottom of the G16 pan and the top part of the J18 pan and welded both parts together, since the oil pick up needs to be at the back as it is in the G16, but the J18 had a front pick up. He also extended the oil pickup pipe. My guess is that while cutting and welding particles were flying off and sticking to the crank and other metal parts. My dad told me that he cleaned the pan itself, but not the engine(the whole engine was under a rag). I guess we had it coming, and only have ourselves to blame.

I don't have a timing light to check the timing, but we marked the stock position and moved the CMP about 5deg clockwise (when standing in front of the engine) to retard the timing. The issue persisted, but the car was a little down on power. Oh and I also didn't wire up the test switch terminal on the ECU to the diagnostic plug, since I didn't know what it was for, but I'm beginning to realize my mistake, so I'll get it hooked up asap.

Once again thanks a bunch! I'm new to this whole car electronics thing, but I am really keen to learn. I have some experience dealing with Bosch EDC15/16 diesel ECUs and Siemens MS Petrol ECUs(both BMW, since that's what I've had access to) so if by any chance, some day, you need help with these systems, I'd be happy to share all knowledge that I have! I feel kind of guilty just asking for help and not giving any in return.

Regards!
Reply
#20
next time you work on rods and cranks see those bearing shells and the very tiny oil slot there?, dirt inside an engine lands there and the bearing will starve and no longer have pressure there and now is metal to metal hitting. ( A CLEAN engine can go 300,000 miles here. )
the oil there on those bearing is a pressurized oil cushion , and most not fail. (or metal to metal hits and bam,. wrecked)

clean is king, inside any engine,
good luck with your project, I did not know the pan was custom, now do.
'
http://www.fixkick.com
Reply


Forum Jump:


Users browsing this thread: 4 Guest(s)