More Data

[TheMonkey]

I'm scrambling to get more data with the sequencer.

The noise is persisting on my Pontiac. I thought I had the wiring set up to where I wasn't getting any more noise, but then I had a hard start and backfire on my last start. It is manifesting itself in the form of crazy ignition ignition signals.... I can't track it down, but the one clue I have is that it seems to not be as susceptible to noise when starting cold. After it's warm, it seems more susceptible. Strangely, my MS2 noise problems had a similar characteristic, but totally different unit and that was fixed with the 'Krisr mod' where an additional transistor was inserted in the VR circuit to emulate the sequencer diagram. A little kooky... I wish it was easier to track down.

One thing that makes this difficult to track down is that I have a 2 channel scope and there are 8 ignitions to monitor. I might be monitoring cyl 1 and it's firing okay, but I might not know if another cyl is the one with kooky timing or dwell period.

But.... I have a plan to gather more data while we are still waiting for others to get fired up; Bruce & Al agree this should be good data to gather.

I have a 1980 BMW 635csi with working L-jetronic. That road has been travelled a lot to convert to MS, and should prove a better and easier platform for the sequencer testing to where I can introduce it in stages to try to pin down if and when noise is introduced. Plan is to have stages:

- fuel wasted only in bare bones fashion
- ignition in wasted with MS still overlayed over existing wiring (sequencer requires COP; i'll use more D585)
- introduce cam sync
- introduce IAC, fan control, etc.
- remove existing wiring & l-jet, clean up new wiring loom

I'm new to BMW, but this looks remarkably easy to convert. Good news is that Peter Florance and many others are very familiar with this system, so I'll lean on them a bit, but I'm determined to sort through this.

[TheMonkey]

Donor L-jet went into surgery today. cables going into it remind me of the wires going into backs of the heads in the original 'The Matrix'. got a crank wheel today and all sensors. should have data shortly, but have a few knuckles to bust after head gasket arrives. shoulda been here a couple days ago. original L-jet is behind glove box, and there is a nice spot right next to it for SEQ to nestle.

& Yes.... I did cut the traces between batch fired pins.

since these are low-z injectors, i bumped up each INJ wires to ampseal to 14g, and ran 6 grounds from red side each 14g. if there is anything i hate, it's insufficient grounds.

[TheMonkey]

after sorting through some comms issues, plugged this in and car fired right up. running fuel only in wasted.

- it is really rough. the tach drops out a bunch. haven't even had a chance to try and tune yet. o2 reading is kooky (either from injectors dropping pulses, or because the power relay is switched by fuel pump signal pin from Seq).

- the 3amp fuse to the sequencer blows when ignition shut off at key - every time. does not blow if car dies on it's own. does not blow fuse unless car is running and shut off at key. while motor was running, i tried just pulling fuse to Sequencer to kill car without blowing fuse, but car keeps running if seq fuse is pulled (?!?!?). bumped fuse up to 5amp (actually ran out of 3amps) and same behavior on 5 amps.

- VR signal is huge. at idle, the DVOM reads > 20v AC, and the missing teeth on 60-2 wheel looks great. see pic:



See Datalog in next post.

Bruce- would you suggest the mod to the power supply you were talking about? Where should I be probing to find probs?

[TheMonkey]

seems to idle okay (i think), but can't rev past 1k. just falls flat and cuts out.

i scoped 10 seconds of the VREF from SensorGround at TPS while it was idling. Voltage was rock steady at 4.85v +/- .05v the entire time. even when recording VREF through a stumble when trying to rev past 1k RPM, VREF seems to be fine.

also... with a 10amp fuse to Seq, behavior changes. if it is running, and i turn key off, car keeps running with key out (5 amp fuse would have blown) (also, remember running fuel only; ignition is from distrib). so, with 10amp fuse, need to pull fuse and key to shut car off. with 5 amp fuse, need to pull fuse and then key or else the fuse will blow when key is pulled. i think i'm going to wire in a new relay to power sequencer and o2 sensor instead of tapping into factory relay.

look at this new datalog..... it really seems to idle fine. i very slowly started to crack open throttle, and at 1,200 revs it just loses it. i suppose this could be a VR circuit issue, or tooth decoding issue. it went 65 seconds without any trigger +/- until it cut out at 1200 revs.

[Matt Cramer]

It's getting tach dropouts. Check the sensor polarity, and if that's good, try putting a 10K resistor in line with the sensor.

[TheMonkey]

thanks Matt. i think polarity is fine. take a look at VR signal above. sequencer is like microsquirt where the raw VR signal is in phase with the square wave to chip, so i set the sync to falling edge.

i put the 10k resistor inline, and shazam... it worked. took it for a short drive.

here's a datalog of when i returned, after it warmed up, of a slow increase in revs to 3k: http://www.mediafire.com/?veo2myzxxmg

i'll start working on the tuning.

[TheMonkey]

I just ran a couple errands in it. Zero tach dropouts. It runs better than it does with original ECU, but something was not right with that... it was running super lean, and that's why it blew a head gasket.

I need to clean up a couple wiring things, then should be able to sort the fuel map and start getting ready for ignition swap. At least I can take a look at pulses on IGN wires to see if they are behaving right before coils are hooked up. ignition is what tripped up things on my Pontiac with this Sequencer.

I'm leaving Sunday morning for Baja 500, so probably won't have much progress for a couple weeks.

[Peter Florance]

I just ran a couple errands in it. Zero tach dropouts. It runs better than it does with original ECU, but something was not right with that... it was running super lean, and that's why it blew a head gasket.

I need to clean up a couple wiring things, then should be able to sort the fuel map and start getting ready for ignition swap. At least I can take a look at pulses on IGN wires to see if they are behaving right before coils are hooked up. ignition is what tripped up things on my Pontiac with this Sequencer.

I'm leaving Sunday morning for Baja 500, so probably won't have much progress for a couple weeks.

VR problem is interesting. Is the sequencer built like the microsquirt schematics? On the V3 BOM R45 R46 are 10K

On the microsquirt the single resistor is R26 is 1K 1/2w which is probably why the 10k resistor helped. I've never had to use that on V3 boards but I've always built them as the BOM shows.

[TheMonkey]

Peter-

My VR situation might deserve an asterisk.... at first, my VR signal was pretty messed up at the MS harness, but right at the VR pigtail looked great. M90 is laid out similar to M30; the grey shielded wire was right behind the aux air valve, tucked under the big loom which has all injector wires. That location was messing it up. When I moved that wire in front of the oil fill hole, the signal cleaned up. Also.... the B35 crank sensor signal is huge. Maybe the big amplitude caused this? (although I suspect you have a lot of experience with this same sensor).

Scott.

[TheMonkey]

i have put quite a few miles on this with just fuel. it's working really well.

i'm not a fantastic tuner, so i'm still working on the fuel map & accel enrichments, but it's running much better than the stock ECU.