Some great general troubleshooting tips - Miata Turbo Forum -Boost cars, acquire cats.

Welcome to   Members

Insert BS here A place to discuss anything you want

LinkBack Thread Tools Search this Thread
Old 04-21-2013, 10:23 AM   #1
Elite Member
Thread Starter
Join Date: Jul 2005
Posts: 6,314
Total Cats: 68
Default Some great general troubleshooting tips

I've seen many fall into the traps he describes

Andy's General Troubleshooting Rules

I don't want this to sound like a rant (but it probably will) but it seems that a lot of people could benefit from the basic principles and processes that I follow which allows me to solve problems for others quickly - and why when I go to a workshop that has had trouble with an ECU on a car, I can usually diagnose and solve the problem quickly. They might think it's because I know Adaptronic better than they do, and while that's true, the main difference is that I have processes for solving problems, which is what has allowed us to develop piggyback ECUs that don't throw fault codes where plenty of others have failed (to the point where it was declared impossible for some cars).

There's this Zen story; I can't find it on google so I'll try to remember it.

Zen master: There was a fisherman who dropped his keys off the side of his boat. He then spent the next hour looking for them in the bottom of his boat.
Student: Why did he look for them in the boat, if he knew he dropped them into the water?
Zen master: Because he could not swim.

I'm going to start off with traps I see people fall into all the time:

Trap #1: Assuming the problem will be something that you already know about, rather than diagnosing the problem correctly
RX8 ECU shows check light and oil metering pump fault code straight away. Clear the fault code and it comes back immediately. Mechanic, rather than checking for damaged wiring (knowing the engine had been replaced in the car), spends half a day sourcing a replacement oil metering pump and fitting it, only to discover it didn't fix the fault. I told him to check the wiring first - because doing that you will be able to find an open circuit in the wiring OR an open circuit pump, and furthermore if the fault comes up immediately, it probably means the ECU has detected an open circuit. It can't know how much oil is being delivered (there's no sensor for that) so the only other possibility is that the ECU is moving the motor but the sensor isn't showing a change in position (pretty unlikely compared to damaged wiring). But mechanical work is his comfort zone, wiring is not... so he looked in the boat for the keys.

Trap #2: Using expensive repairs as a diagnostic method
You see this all the time in factory workshop manuals. They give a symptom, and then say "replace part A". If that doesn't fix it, replace part B. If problem persists, perform operation C. In some cases this may be cheaper than the time to diagnose the problem, and certainly in a dealership environment where you may have access to these parts to test, as well as less skilled mechanics than you would find in a tuning environment (no disrespect, but let's be honest - the OEMs make the procedures as simple to follow as they can, so that they can just recruit normal mechanics - not outstanding ones). In aftermarket world though, swapping out parts like ignitors which may not be standard for that car because of different engines and so on, when you don't have them sitting around takes a lot of elapsed time while the parts come in, and can cost a lot in parts as well.

Trap #3: Making errant assumptions
I once got involved with a 2JZ engine not running on all cylinders. In the first case, it was presented to me as two of the ignition channels not working. I proceeded to show them that all 6 channels were working, so they said "Well, they must be at the wrong time then". I said "Look, I find that really unlikely, you're using a plug-in loom with a standard harness, I don't see how any could be swapped around, but let's test it". So we used a timing light to check that all 6 were firing, in the correct order (against the cam pulley, because it's direct fire), which they were. Then we found that two of the injectors were not connected properly (poor adaptor cables). To prove this, we disabled all the injectors and ran it on aerostart on all 6 cylinders. Doing an injector test initially would have shown the disconnected injectors; and I'd assumed that was already done. They wasted a lot of their own time on this by incorrectly jumping to the wrong diagnosis, and I wasted time as well by assuming that their diagnosis was correct. Fortunately they hadn't yet fallen into Trap #2 and started replacing ignition coils or the ignitor.

Another trap is thinking that the timing is correct on a 4-stroke direct fire or distributor engine simply because it looks correct on the crank pulley. The timing could be out by 360 crank degrees (180 cam degrees) and you would not know if you're only looking at the crank pulley. Again, errant assumptions.

Trap #4: Not believing the evidence
So many times I have seen people come to conclusions that are inconsistent with the data they've collected, and when confronted about this, they usually defend themselves by saying they don't believe the data, or making up some even more implausible theory as to why the data would be wrong. Perhaps the best example of this is Fleishmann and Pons who believed they had created nuclear fusion at room temperature in a test tube; neither of them was a nuclear physicist so the fact that they would publish their claims without review by an expert in the field is bold enough, but when some nuclear experts started saying things like, "hang on, with the amount of heat you claim to be generating, it should be producing X quantity of radioactive hydrogen, and Y quantity of neutrons - in fact enough that you'd be dead by now if you were in the same room" - and they responded with the claim that this must be a new nuclear reaction that no-one knew about and didn't produce any radioactive byproducts or neutron radiation! (The truth was that they didn't create nuclear fusion.)

A recent example was where a tuner was trying to run decent boost (27 psi) on an engine with inductive ignition. It was an aftermarket ignition coil which hadn't been tested in this application by anyone we knew (apart from the usual claims from the manufacturer - "it's good for 1000 hp"). The tuner had what appeared to be an ignition breakdown problem - a load dependent misfire. He had already reduced the plug gap to 0.5mm. He had the bright idea of turning off the lights in the dyno room and getting someone to watch in the cam valley as he ran it up - and he said it was a lightshow of arcs running down the outside of the spark plugs when it hit high boost. I spoke to another tuner about the situation and he said "You need to go CDI to run that amount of boost". I said "Thank you... but I don't believe you - there's enough spark energy from the coil - it's just that it would rather travel down the outside of the plug than jump the gap in the cylinder", and he said "Yeah but if you're trying to run that amount of boost with inductive ignition, you're kidding yourself". I passed this on to the first tuner, but I said that I disagreed and we need to look for different plugs that are pointier inside, or something that will create a stronger field so that the spark will want to jump the gap. That or coat the outside with silicone grease or whatever people do to reduce chance of tracking. He took my advice rather than the other tuner's, got some Splitfire plugs and it was all good. I'm not saying that we wouldn't need to go to CDI at 28 psi, but that going to CDI when the spark was already arcing down the outside wouldn't have made a difference. Fortunately in this case he didn't force the customer to invest in a full CDI system.

In another case recently, at Time Attack I was talking to another entrant who was running a non-Adaptronic ECU in his RX7, and he was telling me about the misfires he was having. He said that the tech support guy from this ECU manufacturer turned up the filtering on the crank angle sensor input, and I said "hang on, were you getting the misfires at part throttle or just at full load?" and he said it was a load dependent problem. I said that usually changing the filtering only helps with triggering problems and that if it's fine at light load, he probably has an ignition problem or something else. He said that yeah, after turning up the filtering the problem remained.
JasonC SBB is offline   Reply With Quote
Old 04-21-2013, 10:25 AM   #2
Elite Member
Thread Starter
Join Date: Jul 2005
Posts: 6,314
Total Cats: 68



Trap #5: Not measuring things properly
Several times I have seen tuners mis-classify a problem because they just haven't tested it properly. I've had RPM dependent, load-independent problems that were described to me as a load-dependent misfire ("it only happens at full throttle" - within the first 30 seconds of looking at the car I reproduced it with no load, and always at the same RPM) - which means that the person goes off searching for an ignition problem instead of a triggering problem. The circumstance above with the 2JZ showed someone thought they had an ignition problem when it was actually a fuel problem (I can't remember what led them to believe it was an ignition problem initially).

Trap #6: Not doing due dilligence
There are things that good tuners do before they tune:
1) Make sure the sensor calibrations are good
2) Check the base timing
3) On a multi plug engine, make sure all the spark plugs are working (and firing in the correct order; the engine might still run smoothly if the trailing plugs are swapped around, but as soon as you get it under load they go kerplink)
4) Make sure that then they're tuning, they have no corrections in place (coolant trim, ignition trim etc)
5) Make sure that the secondary injectors are working before running the car up on boost if it has staged injection
6) Check the fuel pressure

So here are my rules:

Rule #1: Follow the scientific method

Thanks to Sir Francis Bacon, everyone learns the Scientific Method in year 7 at high school. But it seems that almost everyone forgets it when they leave school. I'm not talking about writing up your experiments with Aim, Method, Procedure, etc - but what I mean is that you have to form an hypothesis, design an experiment to test the hypothesis and draw a conclusion from that. So we want to work out why this 6 cylinder engine is only running on 4 cylinders; the first requirement is to work out which 2 cylinders are not firing, by disconnecting injectors or ignition coils one by one, or measuring the EGTs using a pyrometer. Then, we need to work out if we're missing spark, fuel or compression, so each of those is an hypothesis in itself (hypothesis 1: we are missing spark. Experiment to test - check with a timing light. Result: etc). Swapping out an ignition coil, before testing if you actually have a missing spark, is silly, because you don't actually know if you have a missing spark, and even if you do, you don't know it's a bad coil, or that there isn't some other problem as well. So if that fixes the problem then you're sweet, but in the other 90% of cases, you don't gain any information (it doesn't tell you what the other problem is, and it doesn't even tell you that the original coil was OK).

Rule #2: Don't discount any evidence
Just because it doesn't make sense to you at the time, don't discount it because it affects your self-esteem by questioning your understanding of the universe. The universe is bigger and more powerful than you; your job is to understand it; it will not bend to you based on your feelings. So instead of saying "that can't be right" like a child in denial, a more productive reaction is: "that's interesting..."

Rule #3: Write it down
Write down all the results of your tests. People think they can keep it all in their heads, but when you test things properly it quickly becomes too complicated to remember everything, and you forget the exact setup at the time when a particular test happened (did we do that test before we changed this, or after?). So write down the tests you do, the results, and any changes you make between them. You might think it takes time but it's very easy to blow a lot of time repeating tests, or worse, replacing parts or other corrective action that turns out to be wrong because it wasn't diagnosed correctly initially. It's also amazing how many things that don't make sense at the time and many people would ignore - when you write them down, they become pieces of the puzzle that actually help you find the real problem later. This also means whenever you are doing something with an ECU, save the ECU file at the time of the test and always log things. I'm going to rewrite the software so that it's always logging for this reason (storage is cheap, and the number of times people can't give me a log file or have to go back and reproduce the problem because they weren't logging the first time...)

Rule #4: Don't tear down the fence until you know why it's there
Especially if you're modifying someone else's tune file, you must remember that whatever is in there is there for a reason. It might be they have no idea what they're doing, but it might also be there to get around some other problem. For example, I saw an ECU file recently which had -10 degrees master ignition trim. I knew the car had a crank angle sensor, so the timing should be correct (no adjustment on the car, and the settings in the ECU looked correct). So I removed it (the customer is in Melbourne) but I said to the customer that the timing needs to be checked with a timing light, and you need to listen for knock; it's possible that your ignition map is too advanced and that's why they put -10 degrees in the master trim. I still don't know the full story there but at least the customer knows to listen for knock and check it with a timing light - it's about to be retuned. Also sometimes you see funny shapes in ignition maps, dips or bumps in fuel maps where the VVL comes on, ramps in the fuel map in the top end to get around a fuel pressure problem... all these things are part of a bigger system and they need to be considered in context.

Rule #5: This is a very orderly universe
Things don't just happen; if we don't understand an effect, that doesn't mean it's an effect with no cause; it means it's a cause we don't understand. If we don't need to understand it, fine, but that doesn't mean it's magic. There is always an explanation, even if it's beyond the budget of the customer to understand it - after all, tuners are running a tuning business, not a research organisation.

Rule #6: Normally it's 90% diagnosis, 10% solution
In most cases, the value is in the diagnosis. Once you know which wire is damaged, it's usually simple to repair it, for example. In some cases the repair is costly, for example if it's an engine-out job like a broken apex seal (unless you're a guru who can replace apex seals through the exhaust ports - yes I am joking), but in most cases the things that people seem to struggle with the longest turn out to be an ECU setting problem, a wiring error or some faulty component which are easily fixed once identified. But if you start fixing things without the correct identification process, you're just creating work for yourself.

Rule #7: Make life easy for yourself where possible
So many times I see people do things the hard way, because it seems easier at the time. For example:
1) Using non-standard parts, so that when someone else has to fix it, they can't source a replacement
2) Doing things against the manufacturer's recommendation, and then expecting their help to do it the hard way (eg using milliseconds tuning mode rather than VE and then saying it's hard to get the correct mixture at idle, or to cure a rich-lean idle hunt)
3) Ignoring the advice of people that know what they're talking about (I have fallen into this trap myself on my race car, because I'm not an engine builder or suspension / setup person, but I have received lots of contradictory advice from everyone about every aspect of my car, because everyone has an opinion and they're going to disagree, and I won't repeat here whose opinions have turned out to be rubbish and whose have always been right regarding different aspects of the car - but needless to say the people who are right aren't necessarily the people who are the most convincing; they're the people who have been there and done it).

I know these are all very general points and they're no good unless you're conscious of them when you're actually facing the problem. But time and time again I hear about workshops spending days solving what I'm sure they expected to be a trivial problem because they didn't diagnose it correctly initially. Also I find it hard to tell people general points like this when helping them with a problem because it sounds patronising, which I don't want to be. But hopefully this helps someone!
JasonC SBB is offline   Reply With Quote
Old 04-21-2013, 12:33 PM   #3
I haz no trubo
iTrader: (2)
Oscar's Avatar
Join Date: Jan 2007
Location: Bolton, UK
Posts: 2,946
Total Cats: 99

It's no coincidence that someone used to the POS that is called adaptronic has a plan for troubleshooting
Oscar is offline   Reply With Quote

Related Topics
Thread Thread Starter Forum Replies Last Post
WTB MP62 (Hotside) (NB2) Rick02R WTB 3 01-03-2016 07:18 PM
Raleigh: Fiberglass headlight scope, 1.6L ECU, AFM, Hardtop latches, more bigmackloud Miata parts for sale/trade 9 11-07-2015 11:58 PM
Moroso Air Oil Separator Catch Can Aroundcorner Miata parts for sale/trade 2 10-01-2015 03:20 PM
Time to start learning and play with tuning The Gleas MEGAsquirt 3 10-01-2015 09:30 AM
Missing/knock at idle, can someone point me in the right direction? Mikel MEGAsquirt 4 09-28-2015 04:46 PM

Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On

All times are GMT -4. The time now is 08:08 PM.