High Comp 4.6 Torque Monster

[Gorilla]

Must be telling you to add a turbo?

Nice work building the cleaner board and getting it going.

[mpgMike]

Thanks Gorilla. I have too high a compression ratio for a turbo. Might build another stroker for my other '95 ZJ with lower compression and a turbo.

"Why don't you just load the entire old tune?" I did that before and lost firmware updates. Spent a couple hours this morning verifying EVERY setting was the same as the old one (including RPM and Load values for the tables), found a couple differences and fixed them. Next I realized I loaded the wrong previous tune. I didn't use a conventional naming for saved tunes. I loaded the right one and it runs like before!

[mpgMike]

I'm taking an unusual approach to Fuel Table (VE) tuning on the rusEFI. I still have the stock PCM controlling the injectors. I connected a dual-trace O'scope (PicoScope 4-channel, using channels #1 & #2) to the stock Injector #1 Signal wire, and the rusEFI Injector #1 output (with a 10k pull-up resistor to B+). Again, the rusEFI is NOT controlling fuel at this time.

In Tuner Studio there is a Table for Injector Timing. I adjusted values so the onset of the injector signal matched the stock signal starting point. Next, I adjusted rusEFI VE (Volumetric Efficiency) cells so the Pulsewidth of both matched. I did this at idle -- no load, then revved it up a little, pulled it into gear with no throttle, gave it a little throttle... and so forth. I had to go back and forth between adjusting Injector Timing and VE Tables many times. It should at least idle when I connect the rusEFI to the injectors.

For me to take control of the fuel with the rusEFI I have to switch CTS and IAT from the stock PCM to the rusEFI*. I have connectors already in place for an easy transition. Next I have to unplug the stock PCM from the actual injectors and connect them to the rusEFI -- again, connectors are already in place. Finally, I have to remove the stock(ish) narrowband O2 sensor from the exhaust and install the Bosch 4.9 LSU in its stead and connect it. I need an earlier start in the day for that, and I need a day when it isn't raining. Today's work took place in the dry garage (yes, it was raining).

* The stock PCM has 4.7k pull-up resistors for the CTS and IAT sensors, while the rusEFI has 2.7k pull-ups. When both are connected, neither get accurate readings. Between disconnecting the temp sensors and injectors, I WILL get a CEL! I have a way of eliminating the injector codes and temp codes, which I'll probably mess with later. For the injectors it involves plugging in dummy inductors. For the temp sensors, it would be dummy resistors.

[mpgMike]

Swapped the O2 sensors (WB for NB), swapped injector & temp sensor connectors, got it fired up on rusEFI. Spent about a half hour just getting the AFR "close" parked in the garage. Tried to use the Auto Tune function...

Jumped online and read through some "Using Auto Tune" write-ups, and watched a YouTube video of someone actually doing it.

Made some changes (like enabling Closed Loop) and behold -- Auto Tune is absolutely fantastic! Took it out on the road (romping to hit those elusive table cells) and it was running better than ever! ...until I turned around to come back. Ran like garbage, pig rich and chugging on less-than-6-cylinders. Spent a couple hours pinning it down and the culprit was...

...wait for it...

...the rusEFI injector driver for cylinder #4!!!!!!!!!!!

Looked at the schematic, looked up the data sheet, and it's simply a MOSFET Gate Driver; not an injector driver. The part number in the schematic and what I have on the latest-n-greatest "Rev E" board didn't match, so I looked them both up. Unfortunately, it's massively over-kill for a MOSFET Gate Driver, but still rather anemic for an injector driver. Got new ones coming. They're surface mount, and I hope I don't kill the board de-soldering the old ones (2 checked out bad on the bench).

On the plus side; got the individual LS coils working extremely well, the added knock sensor functions, and the built-in WB O2 driver circuit performs as claimed.

[mpgMike]

Before I messed with anything, I rechecked resistance between the driver circuits' outputs and ground:
1 - 61k ohms
2 - 61k ohms
3 - 61k ohms
4 - 30.8 ohms (not k ohms, ohms!)
5 - 61k ohms
6 - 663 ohms

This means 2 of the driver chips were bad; for cylinders 4 and 6. Got the new "injector drivers" in today. To replace them, I first snip the legs with fine flush-mount side cutters.

rusCinch10.png

rusCinch11.png

[mpgMike]

This leaves the "feet" stuck to the pads. I use a hot soldering iron to gently loosen and remove them, leaving the pad smooth.

rusCinch12.png

I then apply soldering flux to the pads to promote bonding with the solder from the new part's feet to the pads. Before soldering surface mount components, I tin the tip of the soldering iron. Got the new ones installed and rechecked the resistance from the output to ground. The new ones were 58k ohms, while the one I left in place was around 61k, so I replaced that one while I was at it.

rusCinch15.png

Reinstalled the rusEFI in the ZJ and had to mess with Cold Start Enrichment to get it to start (I started with a warm engine last time). After it warmed to 40 degrees C (about 105 F), I engaged Auto Tune. It did a few minor tweaks to some of the cells, but nothing drastic. Then I pulled it into gear........

Suddenly the engine had a misfire AGAIN! It turns out injector #4 wasn't firing at all this time (better than hanging open like last time). Resistance between the driver and ground still checked out. I checked continuity between the injector connector and the rusEFI board -- to include everything I did in between; all good, and no shorts to ground. I checked resistance on my injectors to see if they might be the cause of failure:
1 - 14.35 ohms
2 - 14.28 ohms
3 - 14.52 ohms
4 - 14.43 ohms
5 - 14.05 ohms
6 - 14.29 ohms

Since the injectors were all fine, I broke out the schematic and noticed there is a 4.7k ohm resistor tied to the driver chip's Input pin from the processor. Checked resistance on all 6 circuits and all but #4 ohmed out reasonable. Number 4 was only about 77 ohms; it was never turning on to fire the injector. Back to the schematic, that 4.7k ohm resistor for injector #4 is labeled, "R32" (Resistor #32 on the board; a tiny SOB).

rusCinch16.png

I removed it and rechecked resistance to ground on the processor side and it was -- 80 ohms. I have a bad processor on my rusEFI board! I'm also posting this on the rusEFI Forum so Andre will see it. I'm going to ask for another free-bee as this was board #5.

[mpgMike]

A couple of days ago I decided to go fishing early morning. Went out to the Jeep and it wouldn't start. Checked for spark, nope. Reconnected the distributor and reverted to 100% stock PCM control. No answers yet as why, just a bit frustrated again.

[mpgMike]

Folks, I will soon be back in action! Checked the mail today and the most honorable AndreB had sent me a replacement UAEFI board (believe it or not, he sent 2!). I want to add flyback diodes (1N4001) between the injector driver outputs and B+ (will do that in the Breadboard section on the PCB). The data sheet for the injector driver suggests a 1k resistor between the Processor and driver chip, but there is no place for me to add it on the Rev E board. Will post more as I move along.

Rus6_1.png