Is this M48 problem obvious?

[peterraymond]

I have an M48 and a 2 liter Cosworth YBM. At my last event the car died as a was driving slowly through the pits. Trying to restart it, it started to catch, but then nothing. A little later the engine started and ran fine till I shut it off, or for a minute or so. A couple hours later it did nothing other than crank as I ran the starter. At that point I had:
. Good battery voltage
. Fuel in tank
. Good flow after pressure regulator
. good fuel pressure
. The 2 fuel pumps ran continuously with the engine not turning, when they should shut down after a few seconds without oil pressure.
. No flow out of injectors
. No spark (not certain on this)
. fuses for ECU and pumps 1 and 2 are all OK.

It's now a couple days later and things seem similar, except that now the pumps don't run at all. I plugged in my windows 95 laptop with my RS232 cable and the engine seems to look OK, except of course it doesn't start. Diagnostics say that only sync is not OK, but that goes to OK when I crank the engine.

What should I look at? I could wire the pumps directly and look again for errors, but there seems to be a more fundamental problem.

[AdamW]

Does the RPM shown in the software look realistic when cranking the engine (200-300RPM)? Did you try the injector and ignition test functions? If you have logging enabled can you attach the most recent log file?

What type of triggers do these engines have? 36-1 on the crank I guess and a cam sensor??

[peterraymond]

The rpm while cranking does look reasonable at 240.

I've tried the injector and ignition test functions. The injectors sound OK to me. I could verify actual output, except of course the pumps aren't running. I unplugged the HT wire and I can see spark.

The original engine was sold by Cosworth with carbs, but the Motec was added by the engine builder Steve Jennings. The wiring harness was originally made to fit a 427 in a Cobra, but has worked for me for several years. The brackets for the sensors were fabricated by Jennings. The sensor looking at the cam sees a magnet that's mounted on one of the sprockets. The crank sensor looks at a 6 toothed wheel. 6 relatively narrow, widely spaced teeth. I don't think I'm getting errors from them.

I do have logging, but I'm not sure I can get the file off the old laptop that came with this system. I can look at screens and read data off them though.

[Holmz]

A squirt of ether into the intake will tell if it has a spark as will will run for second or two if it does.

A fuel pressure sensor would allow you log the fuel pressure.
If it is a bad pump, electrical supply, or regulator, then you won't get the right amount of fuel.

[AdamW]

The rpm while cranking does look reasonable at 240.
I've tried the injector and ignition test functions. The injectors sound OK to me. I could verify actual output, except of course the pumps aren't running. I unplugged the HT wire and I can see spark.

Well that indicates the crank trigger is ok, the ecu is doing what it should and at least some of the wiring & ignition system is ok.
If your engine has a distributor then I'm pretty sure it doesn't even need a sync to run so you dont need to worry about looking at the cam sensor much closer yet.

I do have logging, but I'm not sure I can get the file off the old laptop that came with this system. I can look at screens and read data off them though.

Oh yeah windows 95 wont make file transfer easy... (I bet your local computer store will laugh if you go in there and ask for a 3.5" floppy disc for your laptop... I would just suggest a general look through the log, check that injector duty cycle or pulse width looks realistic, temps, map or bap, throttle pos and if any errors are showing.

Your comments about the fuel pump not running sometimes and then running at other times when it shouldnt be sounds suspicious so I would start by temporarily jumpering that relay so it runs all the time to rule out that.

[peterraymond]

Tonight was interesting. I was back to the pumps running and not turning off, but this time I had the laptop at the car. I of course I tried to start the engine and it sounded like it was starting, but immediately died. This lasted for two or three tries and then nothing. Very similar to what was going on at the event. This time, after the engine refused to start, I plugged in the laptop and got ECU not connected. It looks like the ECU was not getting power. Right now, for some reason, it seems like sometimes the pumps get power and other times the ECU does, since when the pumps weren't running, the ECU seemed to be fine.

I'm guessing I'm having problems with the relays. But, it seems strange that no power to the ECU would cause the pump relays to turn on full time. It also sounds like I should mess around with the relays and check grounds and power levels. I assume I can figure out the pin-outs.

Looking at the Motec default wiring diagram, it indicates that turning on the ignition switch activates a relay that sends power to pin 1 and also to one side of a single fuel pump relay. The diagram also shows a second relay that would send power to the ignition system, but I believe I have my ignition coil/MSD directly connected to the ignition switch. There are three relays, but the third one powers the second fuel pump. The ignition switch relay goes directly to ground, but the other two connect to pin 36.

OK, without looking at the wiring diagram, that's all kind of confusing, but If you turn on the ignition switch, the main relay should turn on and power should go to the two relays. If power is not getting from the main relay to the ECU, perhaps pin 36 is default connected to ground and the pumps turn on. I think I need to check power to pin 1, the resistance to ground of pin 36 and the ECU ground at pin 25.

This engine does have a distributor and when the pumps are running I do have good pressure on a gauge. So long as I can hear the pumps, I'm confident that the problem is the injectors not firing.

I would just suggest a general look through the log, check that injector duty cycle or pulse width looks realistic, temps, map or bap, throttle pos and if any errors are showing.

As I recall, duty cycle was less than 3%, throttle position was zero and temps looked OK. When the engine is stationary and the ECU is talking, there are no errors and I could have, but didn't check to see if the TP changed as I opened the throttle. Today though, when it was briefly trying to start, the throttle seemed to working normally. Looking at the engine log, it looks like lambda was acceptable until the engine died.

I had looked at the logged errors and the were interesting. For some reason, every parameter showed two errors, with a few exceptions. The (ominous) exceptions were:

Battery, 59 times
Injector 1, 89
Injector 2, 35
Injector 3, 59
Injector 4, 106
No sync, 55
Sync error, 54

I will admit that seeing errors related to sync is not new, but in the past the engine went back to running fine once there was gas in the tank.

[SprinterTRD]

Power can feed back through the injector drives that can keep the ECU on but as soon as any load is applied, the ECU switches off. Check to make sure you have good 12V & grounds going to the ECU.

[peterraymond]

Progress I think. This morning I had inconsistent operation again, but the engine actually started at one point. I was also able to check and see that the TP went from 0% to 98%.

I started looking at the relays and with the ECU relay unplugged, the pumps both ran full time. I checked the resistance from power to the ECU relay and it was well under an ohm. I checked the resistance from lead to ground for the ECU power lead and it was around 2.3 ohms – not obviously wrong. Resistances from ignition to the first coil connection and from the second coil connection to ground were both low too. Then I checked the resistance through the relay coil and it was in the mid 600 ohms. Could that be reasonable? I pulled one of the pump relays and the coil resistance was something like 86 ohms. 600+ sounds bad.

The results I’ve seen seem consistent with intermittent power to the ECU. The high coil resistance in the ECU relay seems consistent with intermittent operation of that relay. The connections to and from the ECU relay look fine.

The ECU relay markings are:

tyco
VF28-61F14-P01
12193603

And, the ID on the fuel pump relay matches! The coil resistances should match and I could use the same spare for both relays!

However, from Motec I've found out that they probably no longer sell this relay and Tyco says the part is obsolete without a replacement. In the long run I probably should switch to what Motec sells today, but in the short run, my second low pressure pump that fills the surge tank only draws a couple amps. I think I will use the high pressure pump relay to run both pumps and use the relay that that frees up to replace the one that died.

[AdamW]

Then I checked the resistance through the relay coil and it was in the mid 600 ohms.

Yep, sounds like that relay is your problem. The relay coil is quoted as 1.6W so resistance should be around 90ohms.

However, from Motec I've found out that they probably no longer sell this relay and Tyco says the part is obsolete without a replacement.

I'm not sure if yours is the waterproof version or not but they are a fairly standard '280 footprint' relay, google 'iso 280 relay' or a similar term and you will find replacements easy enough, a four or 5 pin version will work for you.
Waterproof: http://www.ebay.com/itm/Mini-280-Relay- ... 1c309e95be
non sealed: http://uk.mouser.com/ProductDetail/Omro ... 96tn78lBJG

[peterraymond]

That was exactly what I needed to know. Thanks.

I found the same part number at http://stores.electronicdepotincga.com/ ... Detail.bok for $4.37 with, apparently, free shipping. I ordered 4 so I could go to 3 new relays with a spare.

I'm feeling much better. Now I just need to call the local Motec dealer and tell them that they can stop scrambling to get me going.

Thanks to everyone.