Between work, kids, family, vacation, heat waves and more work, I actually found some time to devote to my coupe and get the 3.5 conversion one step closer to completion. You may recall I’m using a custom injection system running off of a GM ECU. Since my last report, I’ve finished hanging all the accessories on the motor and all of the engine electrical (starter, alternator, etc.). Got the 265 back in, exhaust hooked up, etc. That was easy. Now the hard stuff.
Since there’s no “out of the box” harness for the sensors and ECU I’m using, I decided I would build one and run it in the original locations starting over by the starter, running across the valve cover (picking up the injector and sensor wiring along the way) and then down through the frame rail into the passenger compartment. I bought a universal harness – basically a bunch of pre-marked wires – intended for Megasquirt conversions and set to work. Approximately FIFTEEN hours later, I ended up with a beautiful harness with all the connectors assembled using the correct professional crimper tools and pins – there are NO splices – I even used new connectors and pins for the 30+ wires into the ECU. In the engine compartment, it’s a dead-ringer for the original – everything is shrink wrapped in incrementally larger sizes as more wires are added to the bundle – very close to the original harness construction. I didn’t have enough wire to make it to the rear seat like the original D-Jet, so I drilled a small hole in the floor just below the firewall and brought the harness up through there. The ECU, relays and fuse board are mounted right above the glove box for easy access. So far, so good. I was getting excited!
That ended up being the easy part; the rest has been nothing short of challenging. In the past, I always used a separate chip burner which worked flawlessly, but created a time consuming tuning process (collect data, pull chip, erase chip, burn chip, install chip, repeat). This time, I am trying out some new hardware that lets me program the ECU on the fly while monitoring all the vital stats from the diagnostic port. Good stuff! But it wouldn’t work. After several frustrating hours, I finally got it to work with some help from the seller – my fault, of course – something about bits and parity settings on my COM port. Then I spent a few hours surfing the web to locate an appropriate .bin file to start programming with. I found several, and made appropriate changes to the injector constants and engine size to at least get it to start. The excitement is building.
Saturday night. Check all my wiring – run the fuel pump to prime the system – crank the motor over with no spark to get oil pumping through it – set the base timing as close as I can – fire up the laptop and connect it to the ECU – they are now communicating nicely – check everything again. Finally, after 16 weeks of parts chasing all over town (and the web), research into every sensor, wire and connector, late nights in the garage playing with cams, pistons, bearings, gaskets, hoses, etc. – its finally the MOMENT OF TRUTH – I turn the key, and. . . .
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, pop, pfitzzzzz.
Hmm. Not what I had hoped for. Check the timing, close enough. Crank over, same result. Try again, and again, and again. Same results. I went to bed.
Next day after the usual errands, kids birthday parties, dinner, baths, bedtime stories, etc. I’m back in the garage. Check the firing order. Reset the timing. Try again. Same result. I start thinking about it some more and suppose it could be the 180 degrees out of time. That could be it! Now, mind you, I really knew that’s NOT it since it installed the dang dizzy with the valve cover off and I KNOW cylinder no. 1 was at TDC on the compression stroke, but sometimes desperation has a way of twisting things in your head, so I tried that –
Grrr, pop, grrr, pop, grrr, pop, grrr, pop. Not even close. Pack it in for the night.
Monday at work. I’m obsessed. What could be the problem? It must be running way too lean. I gotta get home. After the usual family obligations, I’m back in the garage at 9 p.m. This time, I pull out one of my most useful diagnostic tools – the propane bottle. Huh? Yes, that’s right, propane bottle. If you drill out the orifice from a small bernz-o-matic torch and attach a hose to the end, it makes a very useful tuning tool. Its quite handy for finding vacuum leaks, but even better to figure out whether you’re running rich or lean. If its worse with propane, you’re already rich. If its better, then you’re lean. Easy and intuitive. I hooked it into the manifold and turned the knob. Went to start it and. . .
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, pop, pfitzzzzz.
WOW – progress! I took off the bottle and jumped back into the computer programming. My injector constant was set for 24lb/hr injectors (which is what I’m using) so I knocked it down to 19lb/hr to richen up the mixture. I know this isn’t right, but I figure I can deal with it later after its running. Same result as the propane. After three or four tries, I can keep it sputtering just long enough to get to the throttle, but as soon as I try to gas it, it dies. It’s late, I’m packing it in for the night.
Two days in San Francisco on business kept me away from the coupe, but gave me time to think. Why is it so lean? Why won’t it run clean and rev? I gotta get it running at 3k rpm with a good mixture to break in the cam. There’s just too much time and money in it to screw it up now. Hmmm. Wait! D-Jet runs at a lower fuel pressure and my injectors are spec’d at 3 bar! That means I’m not getting nearly enough gas. That’s gotta be it! I got home late on Wednesday and was too tired to even look in the garage.
Yesterday was a loooong day at work waiting to get home. This time, I coerced my son into “helping” me in the garage before dinner since there’s no way I could wait any longer – I know this is it! I adjusted the D-Jet regulator in all the way and ended up with 40 psi – close enough! Then I adjusted the injector constant back where it belongs – checked everything twice – contact!
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, pop, pfitzzzzz.
Arrrggghhhh! Take a break, eat some dinner, put the kids to bed. Back in the garage at 9 p.m. I try going way lean on the injectors. No improvement. I try way rich. Same result. I try a completely different .bin file. Even worse. I enjoy a good challenge every now and then, but at this point I’m pretty frustrated. Take a break. Have a glass of wine. Then, a flash of inspiration! I’m running an L-Jet dizzy with a VR sensor which goes straight into an MSD 6-A. Then the ECU takes its trigger off of the MSD tach signal. When I set it all up the first time, I had the VR sensor set so the MSD would trigger on the leading edge as the star wheel in the distributor passes the magnet. That made sense at the time, but maybe it really wants to see the trigger on the trailing edge. Maybe, just maybe, I have the VR sensor wires reversed. . . could it be that simple?
A quick wire swap, injector constant set back to 24lb/hr, double check everything again, turn the key, and ------
Vroooom!
Its actually running! I jump out and hit the throttle to keep it going and it actually catches! Revs to 3k and stays there while I’m holding the throttle open! Finally. Its surging, so I know its running really lean, and I shut her down for now. I’m going to swipe the wide-band O2 sensor off my Jeep and put the cooling system back on before I go any further. But it runs – and it should run well after several hours with the WBO2 and the laptop. What a relief! So now I just need to find some time this weekend to break in the cam, re-torque the head bolts, experiment with the tuning and hopefully it’ll be back on the road next week. Its been too long. Unfortunately, there’s a brake job on my wife’s Grand Cherokee in line for Saturday morning, then my mother in law’s window regulator, then the exhaust on my brother in law’s 535i . . . I should just open a shop.
Sorry for the long musings – but I had to share with people who may appreciate these trials and tribulations. I tried to share my excitement with my wife and got a simple “that’s nice – when is it done?” in response. I’ll get some pictures of the finished product posted when (should that be “if”?) its done.
Since there’s no “out of the box” harness for the sensors and ECU I’m using, I decided I would build one and run it in the original locations starting over by the starter, running across the valve cover (picking up the injector and sensor wiring along the way) and then down through the frame rail into the passenger compartment. I bought a universal harness – basically a bunch of pre-marked wires – intended for Megasquirt conversions and set to work. Approximately FIFTEEN hours later, I ended up with a beautiful harness with all the connectors assembled using the correct professional crimper tools and pins – there are NO splices – I even used new connectors and pins for the 30+ wires into the ECU. In the engine compartment, it’s a dead-ringer for the original – everything is shrink wrapped in incrementally larger sizes as more wires are added to the bundle – very close to the original harness construction. I didn’t have enough wire to make it to the rear seat like the original D-Jet, so I drilled a small hole in the floor just below the firewall and brought the harness up through there. The ECU, relays and fuse board are mounted right above the glove box for easy access. So far, so good. I was getting excited!
That ended up being the easy part; the rest has been nothing short of challenging. In the past, I always used a separate chip burner which worked flawlessly, but created a time consuming tuning process (collect data, pull chip, erase chip, burn chip, install chip, repeat). This time, I am trying out some new hardware that lets me program the ECU on the fly while monitoring all the vital stats from the diagnostic port. Good stuff! But it wouldn’t work. After several frustrating hours, I finally got it to work with some help from the seller – my fault, of course – something about bits and parity settings on my COM port. Then I spent a few hours surfing the web to locate an appropriate .bin file to start programming with. I found several, and made appropriate changes to the injector constants and engine size to at least get it to start. The excitement is building.
Saturday night. Check all my wiring – run the fuel pump to prime the system – crank the motor over with no spark to get oil pumping through it – set the base timing as close as I can – fire up the laptop and connect it to the ECU – they are now communicating nicely – check everything again. Finally, after 16 weeks of parts chasing all over town (and the web), research into every sensor, wire and connector, late nights in the garage playing with cams, pistons, bearings, gaskets, hoses, etc. – its finally the MOMENT OF TRUTH – I turn the key, and. . . .
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, pop, pfitzzzzz.
Hmm. Not what I had hoped for. Check the timing, close enough. Crank over, same result. Try again, and again, and again. Same results. I went to bed.
Next day after the usual errands, kids birthday parties, dinner, baths, bedtime stories, etc. I’m back in the garage. Check the firing order. Reset the timing. Try again. Same result. I start thinking about it some more and suppose it could be the 180 degrees out of time. That could be it! Now, mind you, I really knew that’s NOT it since it installed the dang dizzy with the valve cover off and I KNOW cylinder no. 1 was at TDC on the compression stroke, but sometimes desperation has a way of twisting things in your head, so I tried that –
Grrr, pop, grrr, pop, grrr, pop, grrr, pop. Not even close. Pack it in for the night.
Monday at work. I’m obsessed. What could be the problem? It must be running way too lean. I gotta get home. After the usual family obligations, I’m back in the garage at 9 p.m. This time, I pull out one of my most useful diagnostic tools – the propane bottle. Huh? Yes, that’s right, propane bottle. If you drill out the orifice from a small bernz-o-matic torch and attach a hose to the end, it makes a very useful tuning tool. Its quite handy for finding vacuum leaks, but even better to figure out whether you’re running rich or lean. If its worse with propane, you’re already rich. If its better, then you’re lean. Easy and intuitive. I hooked it into the manifold and turned the knob. Went to start it and. . .
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, pop, pfitzzzzz.
WOW – progress! I took off the bottle and jumped back into the computer programming. My injector constant was set for 24lb/hr injectors (which is what I’m using) so I knocked it down to 19lb/hr to richen up the mixture. I know this isn’t right, but I figure I can deal with it later after its running. Same result as the propane. After three or four tries, I can keep it sputtering just long enough to get to the throttle, but as soon as I try to gas it, it dies. It’s late, I’m packing it in for the night.
Two days in San Francisco on business kept me away from the coupe, but gave me time to think. Why is it so lean? Why won’t it run clean and rev? I gotta get it running at 3k rpm with a good mixture to break in the cam. There’s just too much time and money in it to screw it up now. Hmmm. Wait! D-Jet runs at a lower fuel pressure and my injectors are spec’d at 3 bar! That means I’m not getting nearly enough gas. That’s gotta be it! I got home late on Wednesday and was too tired to even look in the garage.
Yesterday was a loooong day at work waiting to get home. This time, I coerced my son into “helping” me in the garage before dinner since there’s no way I could wait any longer – I know this is it! I adjusted the D-Jet regulator in all the way and ended up with 40 psi – close enough! Then I adjusted the injector constant back where it belongs – checked everything twice – contact!
Grrr, grrr, grrr, grrr, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, bhlapt, pop, pfitzzzzz.
Arrrggghhhh! Take a break, eat some dinner, put the kids to bed. Back in the garage at 9 p.m. I try going way lean on the injectors. No improvement. I try way rich. Same result. I try a completely different .bin file. Even worse. I enjoy a good challenge every now and then, but at this point I’m pretty frustrated. Take a break. Have a glass of wine. Then, a flash of inspiration! I’m running an L-Jet dizzy with a VR sensor which goes straight into an MSD 6-A. Then the ECU takes its trigger off of the MSD tach signal. When I set it all up the first time, I had the VR sensor set so the MSD would trigger on the leading edge as the star wheel in the distributor passes the magnet. That made sense at the time, but maybe it really wants to see the trigger on the trailing edge. Maybe, just maybe, I have the VR sensor wires reversed. . . could it be that simple?
A quick wire swap, injector constant set back to 24lb/hr, double check everything again, turn the key, and ------
Vroooom!
Its actually running! I jump out and hit the throttle to keep it going and it actually catches! Revs to 3k and stays there while I’m holding the throttle open! Finally. Its surging, so I know its running really lean, and I shut her down for now. I’m going to swipe the wide-band O2 sensor off my Jeep and put the cooling system back on before I go any further. But it runs – and it should run well after several hours with the WBO2 and the laptop. What a relief! So now I just need to find some time this weekend to break in the cam, re-torque the head bolts, experiment with the tuning and hopefully it’ll be back on the road next week. Its been too long. Unfortunately, there’s a brake job on my wife’s Grand Cherokee in line for Saturday morning, then my mother in law’s window regulator, then the exhaust on my brother in law’s 535i . . . I should just open a shop.
Sorry for the long musings – but I had to share with people who may appreciate these trials and tribulations. I tried to share my excitement with my wife and got a simple “that’s nice – when is it done?” in response. I’ll get some pictures of the finished product posted when (should that be “if”?) its done.
