PDA

View Full Version : First track event in Portland went as expected.



magnusfeuer
August 23rd, 2015, 11:26 AM
On Friday I was attending my first track event at the Portland international raceway, intent on establishing British dominance once and for all.

And how dominance was established! Once I figured out the basic lines and started honing my brake points, lesser automotive beings were dispatched at a rapid pace. The checkered flag fell as ovulating women were exposing their breasts for me on the ledgers.

All for naught. Halfway through the cool down lap the engine died.

In utter humiliation I was towed back to the pits while now menstruating women hurled used feminine products on me, turning a glorious event into just another daily experience.

Fault search is in progress. Here is the rundown, so far:

No error codes apart from barometric air pressure giving an incorrect signal.
Battery good.
Crank sensor ok - RPM registers during crank.
Cam sensor has been disconnected forever, so 360 mode.
Injectors fire - I took them out and sprayed the garage for a bit.
Ignition relay clicks.
No spark - I've checked all plugs (coil on plug), and they are dead.

While fault searching, I did find a loose ground strap that I tightened down. This means that while the ECU was grounded through its own wire, the engine and chassis did had a flaky ground connection when the car died.

I've just emailed Giles to ask if the ignition driver could have been damaged by the lack of chassis ground?"

slomove
August 23rd, 2015, 02:24 PM
Sorry to hear that. As long as all the moving parts are in their proper place you should be fine....I can tell you nothing beats the excitement of seeing a valve stem through the plug hole :)

That said, the ignition driver should not need any ground on the engine side. The coils run from +12 to the ECU driver output. But I am not sure what happens with the COP coil high voltage side and how the spark circuit is closed. E.g. with the Zetec waste spark separate ignition coil it would not matter as well. Did you check the drive side with a scope? And nothing there, none of the coils? If nothing, is it high or low dead?

magnusfeuer
August 24th, 2015, 08:40 AM
Undocumented modifications were made to the harness when it was adapted to coil on plugs, so I will need to buzz the harness to figure out the 12 volt feed to the coils.

If the coils do have power, the T6 is blown. I will then fly down to California, sneak into Doug's house and switch ECUs with him. I'll keep you posted.

/Magnus F.

Doug Liedblad
August 24th, 2015, 12:22 PM
I thought Giles could test and repair them?

What software engineer did the undocumented changes?

How loose was the ground strap? Falling off loose?

magnusfeuer
August 24th, 2015, 12:42 PM
Giles can repair them, although I can probably ask our hardware designer at my office to do it as well. I've borrowed a scope to do some checks.

Adi did the changes, and I was too lazy to update the drawings.

The bold holding the strap to the engine was about 3 turns out, making the strap rattle. Fixed, with Loctite.

/Magnus F.

magnusfeuer
August 24th, 2015, 12:43 PM
I also borrowed a $60(!) oscilloscope to check for signals to the coils.

/Magnus F.

slomove
August 24th, 2015, 01:39 PM
I also borrowed a $60(!) oscilloscope to check for signals to the coils.

/Magnus F.

If this is a cheap and slow digital scope like the little handheld ones, you may not see the typical voltage spike every time when the drive output turns off. That baffled me for a while when I tried to find the reason for ignition problems a few months ago. This pulse (when the spark happens) is so short, a slow scope may just lose it between samples. Was a good excuse for me to buy a new scope.
But you should still see the drive "on" pulse.

If the loose ground strap interrupted the spark high voltage circuit (hard to believe though), the energy stored in the coil's magnet field can not discharge over the well defined plug gap and will backfire into the drive side, leading to a much higher voltage pulse on the drive. I have heard that may damage the ECU.

magnusfeuer
August 24th, 2015, 02:05 PM
Excuse my ignorance. But are you talking about scoping the high-voltage pulse itself, or the signal that triggers the high-voltage pulse?

I would guess that measuring the high voltage bit would fry the scope pretty immediately.

/Magnus F.

slomove
August 24th, 2015, 05:44 PM
Excuse my ignorance. But are you talking about scoping the high-voltage pulse itself, or the signal that triggers the high-voltage pulse?

I would guess that measuring the high voltage bit would fry the scope pretty immediately.

/Magnus F.
Yes, this is a good guess. I mean the drive side that is connected to the ECU. (Exception: if you have a separate ignition coil you can scope the high voltage side with an inductive clamp but that won't work with COP).
But for what I meant:
- When "off" the drive should be at 12V
- When "on" for the specified coil-on-time it should be under 1V
- When turning off at the end of the coil-on-time you should see a very short spike of maybe 200-400V that corresponds to the spark (which may have 30,000+ Voltson on the plug side). But as mentioned a slow scope can miss that pulse.

https://www.picoauto.com/images/uploads/screenshots/chrysler-cop-fig3.png

magnusfeuer
August 31st, 2015, 08:47 AM
Scoped the ignition during the weekend.

12 volt feed is ok and follows ignition.

Signalling is completely dead at 0 volt. I used the very clear 12 volt feed setting as a start and then decreased timing to the micro second range, but there was no indication of life.

I've emailed Giles at Zentec to check if he can service my unit, or if he has a spare one. If that doesn't work I have a hardware designer on my team that may be able to help.

/Magnus F.

slomove
August 31st, 2015, 03:29 PM
Well, does not sound good. But before you do something radical, I would check if you have an old fashioned short to ground in the cable or if the coil itself is burned out (no reasonable resistance). Although, thinking about it...you would need shorts to ground in all 4 cables or 4 broken coils which is kind of unlikely. But then again.... if it is not the cable you would have to burn out all 4 drive transistors in the ECU. Sounds also unlikely but it only means I have no clue what is going on. :icon_ambivalence:

magnusfeuer
September 2nd, 2015, 08:03 AM
Found the fault. See pic.

I've handed the unit over to one of our hardware developers at Jaguar Land Rover to see if he can fix it.. There are also 5 newly repaired T6 units coming back from England in a week or two. I will probably buy one or two of those to keep as spares.

/Magnus F.

http://i.imgur.com/CggPm8q.jpg

Doug Liedblad
September 2nd, 2015, 08:26 AM
Those look to have been on FIRE, not just melted.

All from a loose ground? I will go check mine now.

magnusfeuer
September 2nd, 2015, 09:20 AM
The theory is that the gate shorted out and fed full-current 12 volt into the ignition drivers.

/Magnus F.

Doug Liedblad
September 2nd, 2015, 10:49 AM
What is the gate?

slomove
September 2nd, 2015, 10:58 AM
Wow...you did blow all 4 transistors. Which means whatever caused this happened to a common part of the circuit or all 4 of them simultaneously. All cylinders failed at the same time I guess?

I don't have a good explanation except the earlier mentioned possibility that the spark circuit was not closed due to the ground problem and the coil backfired its energy into the drive exceeding the voltage rating of these transistors. Another far-fetched possibility is that the dwell time somehow became way too long and the current (rises continually over the dwell period) became too high. In both cases it should have blown the ignition fuse, though (if properly sized).

If you don't find and fix a clear root cause you need to check all relevant wiring and maybe simulate the ecu drive with a momentary button at the ECU connector (push for a fraction of a second) to see if the signal on the coils is correct and the spark fires. Otherwise you risk ruining a brand new ECU the same way.

magnusfeuer
September 2nd, 2015, 11:02 AM
The root cause was that the engine - battery ground strap was loose, as in rattling around.

The ECU still had a direct ground connection to the battery, providing a return path through the harness from the spark plug.

The ground strap bolt has since then been loctited.

/Magnus F.

Doug Liedblad
September 2nd, 2015, 09:26 PM
The smoke got out.

Maybe some of this will help.

magnusfeuer
September 4th, 2015, 08:36 AM
I am definitely out of smoke. I'll get some on amazon.

/Magnus F.

Doug Liedblad
September 8th, 2015, 03:00 PM
Magnus

Check your throttle cable. Before my T6 I had a short to ground and my engine ground strap was loose. It did not hurt any of my electronics but the only path to ground for the engine was thru my throttle cable. It melted the Teflon liner. It did not function well after that.

magnusfeuer
September 9th, 2015, 10:26 AM
Will do. The ECU had its own ground cable to the battery, and the current probably went through that. No burn marks in the connector or on the circuit board ground leads me to believe that the harness is ok.

/Magnus F.

magnusfeuer
September 28th, 2015, 03:33 PM
Quick update: A refurbished T6 should arrive this week (together with a fresh O2 sensor), and hopefully my car will be on the road by the end of the week.

/Magnus F.

Michael Murphy
September 28th, 2015, 05:11 PM
Magnus, sorry to hear about your T6 troubles, but with all due respect, can you get back on thread and talk about PIR? What's the track like? Is it worth a trip north? Do you have any video - your driving or the appreciative ladies, either would work.

magnusfeuer
October 6th, 2015, 08:43 AM
Update:

I got a new T6 ECU, which I flashed. However, still no ignition, which indicates that the coil on plugs are fried as well. Getting new voils is easy, but now I am a bit worried that the potentially shorted coils have blown the new T6 ignition drivers as well.

I took a small handheld scope yesterday and tried to find a signal, but nothing part from some baseline noise. However, looking at Gert's signals in this thread, I may have been at the wrong voltage resolution. I generally suck at scoping, and the scope I used was a very basic unit.

I will probably just ship the ECU with a new sets of coils to Zentec (who used to distribute Pectel) and ask them to fix it.

/Magnus F.

magnusfeuer
October 6th, 2015, 08:53 AM
As for the PIR: Short but fun. Fairly technical. The convenience of having a track 15 minutes from my house cannot be overestimated. I can just drive over there in the Caterham, race, and drive back.

Below are my experiences from a single session; I will surely modify my lines as I get more laps under my belt.

The straights is transportation and brake point timing for C1.

The C1 entry determines your C2 exit speed and defines the whole run up to C4.

I never got the hang of on C4/C5/C6, so I will not describe them.

Exiting C6, you apex at C6A, but try to keep it somewhat to the left, setting you up for a late apex C7 exit.

After that, you floor it all the way to C10. C9A/C9B has a right hand wall, but your suspension is under zero stress so it doesn't pose much of a risk.

I never figured out a good brake point for C10, but you try to flow through C10 and C11, setting yourself for an early start / late apex on C12, thus defining your speed entering the straights.

The track is absolutely flat, for better and worse. Things are more predictable with less risk to dying doing maneuvers like the Buttonwillow magic mountain mid-air rotation, but that also means a little less excitement.


http://portlandraceway.com/_image/page/track_map/10/130117130755/1080h.jpg

magnusfeuer
October 6th, 2015, 09:03 AM
Michael: If you want to come north, you, car, and family are more than welcome to stay at our place.

/Magnus F.

Michael Murphy
October 6th, 2015, 03:04 PM
Sounds good Magnus. Plans are fermenting (fomenting?).

magnusfeuer
October 7th, 2015, 10:44 AM
Fornicating.