ECU's, interfaces and baud rates

[pentium]

Re: ECU's, interfaces and baud rates

The software is called RhinoView. This is version 0.8. Slightly buggy but it works.

In the versions since, the developer has removed support for serial interfaces and firmware locked to his own FTDI based solution which he sells for a ridiculous amount. Likewise with version 1.0 and onwards you have to buy the software as it is no longer free.

This was just my way of telling him to eat s***.

[Longbow]

Re: ECU's, interfaces and baud rates

Just curious what parts of the car's control system you are working on?

[pentium]

Re: ECU's, interfaces and baud rates

The ECU. That's it.
The computer in the Tracker/Sidekick is of a stupid simple design, as in Suzuki gave you the bare minimum needed to make a TBI engine work. It's also non-programmable and doesn't even have a bit of memory to be used for economy "training" so all the data port can be used for is outputting fault codes if you don't want to do the jumper trick listed int he Hayes manual or you want to monitor all of the car's sensors realtime

[UHU]

Re: ECU's, interfaces and baud rates

Love the Panasonic Toughbook - a Toughbook in use always implies that its owner means business!

TBI - Throttle Body Injection?

Have you solved the emission problems? Was it HC (Hydrocarbons) too high or CO (Carbon Monoxide) too high?

Have you cleaned all the earth points?

Engine to chassis, Battery to chassis, there are normally a couple of wiring loom earths all over cars and under dashboards (those often earth the ECU) - you would need a wiring schematic for the car they are all listed there.

Disconnect earth, wire brush till shiny and copper grease them.

Then once all the earths are at the same potential - engine management problems have been known to go away.

An example:-

Was buying an obscure part at a franchised BMW main dealership (fortunately it took them an eon to find the part) and watched a mechanic in the car park "Diagnose and Fix" a BMW M5 without even moving it from its customer parking spot (no customer present).

1.Start car with difficulty - great clouds of black smoke from exhaust - revving just made more smoke - leave car running - return to car with a single battery jumper cable.

2.Connect one end of jumper cable to hefty bare metal part of engine.

3.Connect other end of jumper cable to chassis / body side of engine earths - do each one in turn.

4.Black smoke abruptly stops - rev up still no smoke - move jumper cable around smoke comes and goes (he was identifying exactly which cable and which end of the earth cable needed cleaning - leaving more work for a later wallet busting repair bill)

5.Engine off, nonchalantly disconnect and toss jumper cable on ground, disconnect (unbolt) earth strap, wire brush/clean one end only, reconnect engine earth strap.

6.Restart with absolutely NO BLACK SMOKE! more violent revving (I wonder how the catalytic converter survived all that excess/neat petrol it was subjected to?)

7.Three more mechanics appear - all four pile into the BMW M5 and go for a "test-drive" the M5 comes back smelling of burning rubber - the mechanics look totally crazed as they pile out of the car - one of them has the BMW service test tablet PC with him (probably so they could switch off the built in traction control)

Central London BMW main dealer Labour rates are £130 excluding 20%VAT (Sales Tax) - I wonder how much they charged the M5 customer for cleaning one end of one earth point?

Have you tried engine oil flush? Try Wynns in a steel can. Add to your old engine oil (you could drain to min. mark on dipstick then add flush) circulate for 20 minutes then change oil and filter. Frees piston rings which can have a dramatic effect on emissions.

Let me know how its all going.

[pentium]

Re: ECU's, interfaces and baud rates

The main grounds were cleaned when I had the transmission serviced. Again when I was working on the alternator charging harness.
My hydrocarbons have always been really, really high along with NOX.




I've replaced or tested just about every electronic component in the emissions system, including the plugs.
I have not actually started troubleshooting the emissions issue yet as I'm in the process of moving the completed interface into an extruded aluminum enclosure so it doesn't pose a risk of shorting out and blowing my ECU up again. It also needs a more durable USB connection. It seems the vibration of driving causes the B connector on the FTDI dongle to act intermittent and when it drops the software instantly locks.

[Longbow]

Re: ECU's, interfaces and baud rates

I have an older Swedish car that will throw engine emission codes if I don't install the correct type of O2 sensor. Usually the OEM sensor prices are huge, so an aftermarket sensor gets installed. The trick is to identify the maker of the sensor and the actual maker's part number. NTK is a big player. And to save time, check the O2 signal to make sure it is swinging the required voltage to make the ECU happy. A lot of these problems are due to worn out cat and/or 02 sensor. Figures.

[redwire]

Re: ECU's, interfaces and baud rates

pentium, I asked an engineer who designs ECU's and uses emissions bench regularly.
If you have no trouble codes, he figures it's catalytic converter.
A 1992 vehicle has no ODB2 (second O2-sensor) for diagnostic coverage of the cat.
HC's are unburned fuel, so from crappy combustion- misfires or oil.
Mileage is up there 311,000km so some burning oil could also make it high or damage the cat.
Something about emissions spiking vs just a high steady level during the driving test.
Old O2 sensors get sluggish so transients (step change) are worse.

[pentium]

Re: ECU's, interfaces and baud rates

Mileage is up there 311,000km so some burning oil could also make it high or damage the cat.

That test was done in the fall of 2010. As of 2017 that engine has 430000km on the odo.
I would not be surprised it the cat is bad but the assumption was that something upstream was running the engine a fraction too rich.

[eccerr0r]

Re: ECU's, interfaces and baud rates

I hate emissions tests. Where are they actually showing the public that the values are true? Unless something really is wrong, I think smog tests are just a way to do forced obsolescence of cars...

That being said, having bothe HC and NOx high usually is a mechanical or multiple issue. Running rich tends to reduce NOx.

[redwire]

Re: ECU's, interfaces and baud rates

pentium, good chance that your '92 Suzuki ECU has an emissions defeat device.

I was involved with '91 project using a Suzuki engine and we read out the ECU's firmware and diassembled it. It's been a while but I remember a Nippon-Denso ECU with a HD6301 family 8-bit MCU or a Fujitsu 6502-based part. We had to modify the fuel tables and calibration. Hardware trojan to get in there.

Anyhow, in the firmware we find strange routines and realize they are for detecting an EPA test cycle. Back then EPA it was drive at certain speed for say 30 sec. then another speed for 1 min. etc. very predictable. Once a test cycle was detected, the ECU would then retard ignition timing, turn up EGR etc. the engine became pretty gutless. I guess it was how Suzuki got the engine to pass EPA emissions testing. Their swirl combustion strategy did not work.
Some driving patterns can false trigger this and you notice the engine extra gutless for some reason.

No different than the Volkwagen scandal. I think EPA knew about it and then changed to random time tests. We didn't say anything because of getting in trouble for hacking the firmware.

[UHU]

Re: ECU's, interfaces and baud rates

The point to remember is your car once worked – it now either is not working because something is worn out and needs to be replaced OR something has abruptly broken and needs to be repaired or replaced – think about that - they are not the same failure modes – wear is generally more expensive to fix than an abrupt breakage or failure – so did the hydrocarbons gradually creep up over the years or was it suddenly through the roof one year of the emissions test?

It is easy to make most perfectly "healthy" cars fail an emissions test just cool down (engine cover open) the engine or do not follow the purge procedure to rev the engine - 3,500rpm for 30 seconds then measure – or test a stone cold engine - very few cars will pass that way and it means more money for the mechanic to fix the car at the testing station – that is why I never leave any car unattended for its annual vehicle inspection with emissions test (MOT in the UK) because the testers try and drum up business –do you leave your car unattended / alone / with the garage OR stay with it for the entire vehicle inspection / emissions test?

I digress but - having said that I had to go to a new (to me) MOT garage (the MOT garage I used for decades closed because the owner died) and the new MOT tester without asking “helpfully” topped up the screen wash and Brake fluid – before I could stop him and despite me watching – got home and took a look - they had poured in filthy cloudy black brake fluid (which I vacuum siphoned back out of the reservoir – [I change my brake fluid every 2 years so it is never that colour/state]) while the “helpfully” topped up screen wash erupted with bright green algae that summer – if the algae blocks the screen wash system that is an instant MOT failure – a couple of years later have nearly sterilised the screen wash back into being algae free. Never had algae in the screen wash of any of my cars before or since.

Anyway – your CO reading is listed as a PASS while you are failing on HC (why don’t they use % for CO – their CO reading is meaningless to me mixture adjustment has in my experience always been set to CO%) this generally means incomplete combustion which is not exactly the same thing as excessively rich mixture (too high CO) your CO is listed as a pass – which is noteworthy.

You have to now go further than the emission system when looking for a fault – look at your HC and CO emission graphs over the duration of the 4 minute emission test you have spikes every so often that cause your emission failure – is one of your cylinders intermittently not firing and passing neat fuel (which would be HC) – ignition fault? OR sticking valves not closing so no compression so incomplete combustion means high HC (possibly also gummed up piston rings). With your high mileage there will be some gum by now – any oil leaks or oil consumption?


You need to now do all the work personally yourself – otherwise you could have a mechanic claiming they have done whatever – you rely on their false statement and move no closer to solving the problem – re-clean all the earth’s personally – under the dashboard, in the engine bay - everywhere – is the ECU solidly mounted? They often earth through the case as well – next get a wiring schematic and do voltage drop tests with the engine running on all the power and earth connections on the back of the ECU multi plug – either disassemble the ECU wiring loom connector and measure inside there OR open the ECU and use miniature test probe grabbers to connect to the ECU connector going to the PCB – resistance tests are almost useless for this kind of diagnosis – are any ECU chips excessively hot? Touch them and find out while the lid is off – if the car is not regularly driven due to emissions failures you probably have stale fuel by now which can also cause high HC because stale fuel is very difficult to ignite and will not then burn normally– drain fuel and refill.

Engine wiring looms all have splices where multiple earths are linked together (same goes for power lines) there is normally no visible clue there are splices all over the wiring harness – voltage drop tests should show up most splice problems – remember most car engine actuators are wired to permanent live and the earth is switched by the ECU to control the actuator. Are any of your ECU loom wires shielded? Schematic will often tell you – a classic is the shield intermittently shorting the sensor wire inside the shield.

How clean are your battery terminals? Meaning not just terminal to battery but the various factory wires that are bolted to the terminals – the ECU will generally have an earth straight back to battery negative and a supply straight back to battery hot – sometimes you will also find another earth and hot pair going from the battery terminals to the engine ECU wiring loom.

Try fuel additive – the one in Texaco petrol (Techron) is quite potent – it is available as a bottle of concentrate too – use oil flush to free up sticky valves and piston rings idle with light intermittent revving max 2000rpm for a good 20 minutes on a HOT engine then drain oil DO NOT DRIVE THE CAR CONTAINING FLUSH - it is surprising what can be un-gummed and flushed out of the oil system – warning loads of people complain oil flush makes their engine leak – I have never met one of those people personally and I have used oil flush on hundreds of cars and nothing has leaked for me – at trade shows turbo remanufacturers come out in hives if you mention any oil additive – even flush that is only in the engine for 20 minutes then drained out – I have stopped oil smoking turbos with nothing more than a can of oil flush – which is probably why turbo rebuild companies don’t like oil flush! Odd how those same turbo rebuild companies nowadays supply a hefty syringe full of an oil additive to squirt into every rebuilt turbo’s oil supply line before its first start – hypocrisy?

If you get a flushing product you generally want to pick one in a metal not plastic canister – generally only the ones in metal cans have high enough levels of cleaning chemicals to work which is also why they have to be supplied in a metal can – plastic would be too porous, certain solvents which are crucial parts of its formula would evaporate while sitting on the shelf!

Get an aerosol spray of carburettor cleaner – clean out the EGR valve, Throttle body and idle air control valve.

Do a power balance test i.e. prevent the spark plug going off for each cylinder in turn while the engine is running the idle speed should drop if the cylinder was firing – find a cylinder with no difference means the cylinder is not firing.

Are there any Vacuum leaks? What is the vacuum reading at idle?

Valve clearances correctly set? Meaning you cannot get the next bigger size of feeler gauge in the valve lash/clearance gap.

Think of it this way - if the spark plugs (or oxygen sensor) are fouled replacing them does not find or fix the problem - you have to figure out why they are fouled in the first place and then fix the underlying problem causing the fouling – sounds obvious but that can be surprisingly difficult to do.

Sorry about the length of this post.

[pentium]

Re: ECU's, interfaces and baud rates

It's a problem that's literally been with the car as long as I owned it. Almost all the sensors have been replaced and the rest have been tested good. Fuel and vacuum lines have been changed, the injector has been switched, the plugs replaced, the fuel octane upped, the wire harness tested and cylinder compression checked. I cannot find the source. It just runs slightly more rich than it should....

so did the hydrocarbons gradually creep up over the years or was it suddenly through the roof one year of the emissions test?

I received several years worth of previous AirCare test results and yes the increase was gradual.
To make the vehicle pass we slipped the timing slightly off and it would just pass everything, then you adjusted the timing back. AurCare was also shutdown several years ago anyways so the emissions control are no longer enforced.

There are however oil leaks. The seal on the distributor has failed and needs to be replaced. I've been filling the car with high mileage 10W30 ever since I started driving the bottle of oil every month helps keeps the engine clean.
The head was resurfaced in 2012 to deal with an overheat years earlier. New valve seals were installed and no problems were found in relation to tolerances or cracks. That was also the last time the car had the valves set. The car is scheduled to go in for a complete tune-up as well as several seals I don't want to mess with replaced, plus it's time for a new timing belt and pump. The latter two I can handle at home.

Don't worry about the length. I've been lectured on what to check in this thing for years. I can't really deem it a lemon problem but man it's a strange one.

Anyways the adapter was stuffed into an extruded metal enclosure so it didn't look so hokey. I also installed a new ECU diagnostic port. All the wiring is shielded.




Meanwhile, our buddy at RhinoPower is having a total fit.
His website now has robots.txt enabled and Internet Archive has blocked access to his old snapshots which also means the free versions of his software are no longer directly available from him. Everything now funnels back to him wanting you to pay out the ass for his software and dongle kit.

Can't blame him for trying to make a buck, but hey if you want to hide your freeware behind a textfile and accusing people of stealing I'm sure someone will just host it elsewhere.

[UHU]

Re: ECU's, interfaces and baud rates

Personally unless I do 100% of the work myself I never know how much other "specialists" have REALLY done or tested - they are in the business of making money by billing hours which is not necessarily the same goal as fixing the problem! So the first thing I usually do is re-do everything that is meant to have been investigated / replaced.

Nevertheless - what fault codes and useful info have you discovered from connecting the laptop to the ECU - has that given any good new clues?

"Tested good" could mean anything unless you personally tested it - if it was only a resistance test the results (besides dead short or open) are not usually that useful - voltage under load and volt drop under operating load using a multi-meter are my favorites for discovering what’s really wrong.

It is also surprising how many "techs" look at the reading on a multi-meter and say "nahh can't be" ignore it and just say "tested good" - I have watched them do that loads of times - why they bought any multi-meter in the first place is a good question.

Were all the vacuum lines reconnected back to the right places? Were they even in the right places to begin with? Compare to manufacturers reference manual.

A slightly pinched 1/4" vacuum hose is a subtle but awful problem to find.

Cheap vacuum hose replacements can collapse under high vacuum conditions i.e. Coasting / engine overrun combined with engine heat - were the hoses replaced with the right stuff or generic that cannot handle the engine bay atmosphere?

Nothing above says the ECU has ever been tested or that its engine wiring looms have ever been volt drop tested while under operating load.

Has anybody investigated the EGR valve - does it work how and when it is supposed to - is it blocked? Are the lines/passages to it blocked?

Oil flush shifts things nothing else can - it is a kerosene and diesel base with (basically) a dry cleaning fluid solvent - there are even flushing oils you completely refill the engine with (change the oil filter first) that you can drive around on for a few days for stubborn cases - or use engine oil for a diesel engine in a petrol engine - the extra alkalinity of it can work wonders in a petrol engine.

The oil leaks are generally also a source of vacuum leaks which is why (depending on location) oil leaks when the car is parked on your driveway then with the car running engine vacuum pulls the oil in (stopping the leak) while circulation through engine lowers sump oil level and stops those leaks that way as well BUT those vacuum leaks will lean out the idle CO mixture so the ECU (if it is new or clever enough) will then enrich the mixture to get the right CO when measured at the Lambda sensor - so plug the oil leaks that also leak vacuum and the mixture will drop in my experience.

Do you mean engine timing belt timing or distributor timing?

Am fuzzy on emission tests in Canada – I understand how you were passing emissions re timing but if no more emission test in Canada then why bother fixing the fault?

In the UK if you fail emissions you fail the MOT and car is off the road - even worse since last year if you did the annual MOT a month early and failed it that immediately cancelled the remaining month of your current MOT so in theory your car would then be stuck at / confiscated by the garage until fixed - UK government thinks this is for your own good – I disagree....

Are you serious that the valves haven’t been checked or adjusted for 5 years? Does the manufacturer actually suggest a 5 year interval – if the car has hydraulic lifters then it is possible (all the hydraulic lifter cars I have encountered have had way out valve clearances) but with manual lifters a 5 year interval is way too long – remember valve setting = engine timing and altering the engine timing was how you were passing emissions....

Long valve adjustment intervals are similar to car manufacturers more recent con-trick of “sealed for life” Auto Gearboxes etc. – same ATF oil is used as before – no great jump forward in gearbox or oil technology but just leave the oil in there till it goes black – manufacturers know it is a great way of selling new cars – just get the owners to grenade the gearbox of their current car by never changing the oil!

It could just be a stuck valve causing high Nox – oil flush could free off that stuck valve – engine oil never will!

Let us know what progress you make.

[Viking]

Re: ECU's, interfaces and baud rates

Meanwhile, our buddy at RhinoPower is having a total fit.
His website now has robots.txt enabled and Internet Archive has blocked access to his old snapshots which also means the free versions of his software are no longer directly available from him. Everything now funnels back to him wanting you to pay out the ass for his software and dongle kit.

Can't blame him for trying to make a buck, but hey if you want to hide your freeware behind a textfile and accusing people of stealing I'm sure someone will just host it elsewhere.

Is there any progress on this project?

[pentium]

Re: ECU's, interfaces and baud rates

Whoops. Almost missed this reply.
At this point the project is complete. The goal was to make the RhinoView software communicate with the Suzuki/Geo ECU using the base design of the RhinoPower adapter and that was done and my work was published, or at least as much as Rhinoman wanted to let me.

The vehicle was sold in 2020 after ten solid years of service alsong with the adapter I built and I'm now driving and reverse engineering the computer in a 40 year old AMC Eagle. The project files were packed up into storage but here's a link to the archive, including two versions of the freeware software.

https://www.dropbox.com/s/r6hrxguq6t...oView.zip?dl=1
^Future viewers, this link may expire at some point.

[Viking]

Re: ECU's, interfaces and baud rates

Whoops. Almost missed this reply.
At this point the project is complete. The goal was to make the RhinoView software communicate with the Suzuki/Geo ECU using the base design of the RhinoPower adapter and that was done and my work was published, or at least as much as Rhinoman wanted to let me.

The vehicle was sold in 2020 after ten solid years of service alsong with the adapter I built and I'm now driving and reverse engineering the computer in a 40 year old AMC Eagle. The project files were packed up into storage but here's a link to the archive, including two versions of the freeware software.

https://www.dropbox.com/s/r6hrxguq6t...oView.zip?dl=1
^Future viewers, this link may expire at some point.

Pentium
Thanks for your prompt reply.
It is looks to me that the interface works only with older version of Rhino view, which could have some parameters mistake.
Please correct me if I am wrong.

[pentium]

Re: ECU's, interfaces and baud rates

Yes, it will only work with versions before 1.0.
Two reasons for this. They had not yet added an adapter whitelist which locked you into buying their specific USB FTDI based adapter, essentially turning the adapter into a software protection dongle and starting with version 1.0 they rewrote a number of software components and replaced a lot of reverse engineered code with licensed Suzuki protocol code. This is why it went from free to use and build to you had to buy the dongle to use the software. Now they have to pay fees on the protocol.

Between that, being harassed with DMCA's on the free builds because it cannibalized his profits and overall the car was getting tired I sold it in 2019 and now drive an AMC Eagle, whose computer has no protocols to deal with when performing a diagnostics. :P

[eccerr0r]

Re: ECU's, interfaces and baud rates

really want to get something that understands the dodge/chrysler pre-OBD2 protocol... or get the protocol itself...