[rusEfi] 1990 Geo Storm GSI - Advice/Suggestions? #18

[Spine]

Here is my TDC on the CAM:

TDC001.jpg (2.08 MiB) Viewed 16341 times

And crank timing marks:

TDC002.jpg (990.99 KiB) Viewed 16341 times

So, what's my next step? Also, the "test" buttons do not work. When pressing "test" for fuel pump or ignition, etc. Nothing happens. It shows in the log that it's "doing" it, but it doesn't actually do it. When it first powers on, the fuel pump does run for 5sec. so I know it's connected properly.

[AndreyB]

Set cranking ignition as described @ http://rusefi.com/wiki/index.php?title=Manual:First_engine_start#Confirm_TDC_position

Crank and see if you have the timing mark cranking and while Engine->Trigger->global trigger angle offset = 0
Crank and see if you have the timing mark cranking and while Engine->Trigger->global trigger angle offset = 20
Crank and see if you have the timing mark cranking and while Engine->Trigger->global trigger angle offset = 40
Crank and see if you have the timing mark cranking and while Engine->Trigger->global trigger angle offset = 60
Crank and see if you have the timing mark cranking and while Engine->Trigger->global trigger angle offset = 80

[AndreyB]

Do you have a MAP sensor - do you know the kind or calibration of your MAP sensor? Do you know your injector flow?

Correct me if I am wrong, but it looks like we do not know which cylinder is which so we can only do simultaneous injector or batched injection with different injection phase for different runs?

[Spine]

It does have a MAP sensor. I'm not sure what type though... Here is what it is:
http://www.amazon.com/Delphi-PS10075-Manifold-Absolute-Pressure/dp/B000CGI3Q0

Injectors 1 and 3 need to be batched together and 2 and 4 need to be batched together (per wiring on vehicle). As far as flow, they should be stock flow which I don't know. I'll have to research it when I get back.

Here are the injectors that I've installed:
142-449 KEM Parts Fuel Injector

[AndreyB]

http://www.summitracing.com/parts/dfp-ps10075/overview/ so it's used by GM so calibration must be available online somewhere.

The reason I am asking is to get http://rusefi.com/wiki/index.php?title=Manual:Software:Fuel_Control#Speed_Density running - that's one of the option for fuel control. If not speed density then you need a wide-band o2 sensor, do you have one? Technically you would need one for any algorithm, SD included.

[Spine]

Looking at the part numbers, I have a 1-bar map. This is what I found: http://injector-rehab.com/shop/mapsensor.html
I did a custom MAP, with the 1-bar settings from that website.

My o2 sensor is a 3wire, so it's not a wide-band. I'll be getting a wide-band in the next few months, but I have to work with what I have for now.

I'm still trying to find the flow for these injectors. I know it's the same as the 1990's Cadillac DeVills.

[Spine]

Here's the injector information...finally:
Brand:DELPHI
Part Number(s):FJ10007
Type:Multec1 TFP
Operating Pressure kPa:300
Static Flow Rate Lbs/Hour:16
Driver:Saturated
Alcohol Resistance:E10

[AndreyB]

I've added some info to the http://rusefi.com/wiki/index.php?title=Manual:Software:Fuel_Control page

It looks like you want to start by going Tuning->General->Fuel Algorithm: MAP which would allow you do control injection duration via the Fuel Table.

A longer route would be figuring out CLT, IAT and TPS sensors and going with Speed Density algorithm.

A narrow-band O2 sensor is of relatively low use to us the way I understand things.

[Spine]

Well, I got it running this morning. Runs a bit rich as expected. I had to set it at 300deg offset, and put it to 500rpms for cranking. Changed from SPEED DENSITY to MAP. Put injectors to 1.75ms, and 3.0ms cranking. I can only get to 2200rpms.

I've attached my log for any other input. Ideas on getting this running better?

MAIN_rfi_report_2016-04-21 09_59.csv

(1.36 MiB) Downloaded 636 times

Ah yes, I see your post now. I'll review it today. If you have any other ideas, let me know.

[AndreyB]

This makes you rusEfi #18

At this point TS log files are better since they have more data while running. I am making changes to the rusEfi console to write .msl files as well but as of right now TS log have more data.

[Spine]

I'll get it fired up again and get you a log from ts.

#18,that's awesome and without a phone call. Lol

Sent from my ONE A2005 using Tapatalk

[Spine]

Here's the TS log.

2016-04-21_11.23.42.msl

(204.04 KiB) Downloaded 507 times

[AndreyB]

We are at the edge of my skill set. Most obvious next step would be WBO. I do not remember if there is any sort of a default timing table in the firmware or if you need to set something yourself.

[stefanst]

A few observations from the log:

- It appears you're idling. But MAP is at around 60kPa. It should be around 30kPa or so. Your MAP sensor calibration may be wrong, or you're sourcing MAP from where you shouldn't. Or something else
- I usually get the type of oscillation you see at high MAP when I'm running lean. Those injectors you have are tiny. Maybe increase fuel by 20% in the cells you're using and see what happens.
- If you want to tune speed density you NEED a WB 02. Without it tuning is nearly impossible.
- Make sure you have a reasonable timing table.
- The CLT signal is noisy. Check your grounds.

What is the AFR signal we're recording? Is that noise, or is it hooked up to the O2 sensor?
Post your tables (VE and Ignition advance) here!

And congratulations on getting your engine running!

[Spine]

Ok, so I adjusted the timing, adjusted the injector timing (2.5 times), changed injector 3 and 4 to "none" for pin-out (3 is wired with 1, and 4 is wired with 2 - batch injecting). Ran much better. Still not "smooth" and seems rich => lean (low RPM => high RPM). So fine-tuning will obviously need done (just as any other project). Perhaps I'll wire each injector directly to the computer vs. use this part of the wiring harness - your thoughts on this? I'm looking at some WBO2 sensors on ebay. Here are some inexpensive ones that I've found.:
http://www.ebay.com/itm/NEW-BOSCH-OEM-5-WIRE-UNIVERSAL-WIDE-BAND-AIR-FUEL-A-F-O2-OXYGEN-SENSOR-/381476376022?hash=item58d1c38dd6:g:o10AAMXQQQZR5Gp4&vxp=mtr

and

http://www.ebay.com/itm/NEW-BOSCH-OEM-WIDE-BAND-AIR-FUEL-A-F-O2-OXYGEN-SENSOR-ACURA-HONDA-13680-/172067443337?hash=item281004d289:g:ctwAAOSw-dBTwMvW&vxp=mtr

Which would be better, or what do you recommend besides these? I'm not worried about plugs as I'd "cut them off" and solder my own connectors on it.

Also, I did find a ground wire disconnected. I connected it to a true-ground on the block. Here's my latest TS log file - I believe this one I also drove it around the block. Hope it looks better - I have no idea how to look at these or read these. Remember, I'm old-school. 1/2 glass of water, my ears, and my nose is all I ever needed to tune a car.

2016-04-21_13.20.51.msl

(1.25 MiB) Downloaded 663 times

[AndreyB]

What does not make total sense to me is how does your OEM computer run you injectors. With your even trigger signal shape there is no way for software to distinguish which signal is which, so even while you have two wires they must be either fires simultaneously or randomly or I do not get it. I am not sure to what extent batched or simultaneous injection is a problem, probably not the biggest problem at the moment.

See also http://rusefi.com/wiki/index.php?title=Wideband_Oxygen_Sensor

[Spine]

Yeah, I'm not exactly sure either. There was one setting in the custom trigger settings that gave me a larger trigger followed by 2 smaller triggers, but it only showed 3 total triggers. So, I changed it and this is what I have now. I do know that, based on the wiring schematics, injector 1 and 3 get tied together, 2 and 4 get tied together. Then it has a driver for injector pairs 1/3 and 2/4. I'm not sure how it all works.

As far as the WBO2, I'm assuming that "A" would be fine with controller, correct?

[stefanst]

Millions and millions of cars run happily on batch injection. Don't worry about it. Sequential injection is to get that last 1% out of the engine. This is NOT the problem you're looking for.
Edit: Removed previously posted WB info.

[AndreyB]

As far as the WBO2, I'm assuming that "A" would be fine with controller, correct?

Since there are at least two kinds of sensors it's safer to buy a kit.

[kb1gtt]

WO2's come in different variations, most common are LSU 4.2's and LSU 4.8's. I'm sure OEM's use different variants, or perhaps they meet one of the LSU specs. If you don't get the a controller that matches your WO2, it probably won't work well. Best to get a controller that specifies the type of LSU it supports, LSU 4.8 is slightly better than 4. 2, so I would mildly encourage LSU 4.8. This Canadian fellow at 14point7 has a $75 controller installed in the wire, which simplifies the wiring. Beware, he's kind of a pain as you commonly have to poke him to get him to ship product, but from what I've seen it's physically a good design.

http://www.14point7.com/products/spartan-lambda-controller-2

The OEM distributor would distribute the spark to the proper cyl, and was driven off CAM, or geared to half from Crank. Then the pulses going to the ECU were just to know RPM, and allow fuel calcs. It didn't care if it squirted at the exact crank angle. Squirting at the exact crank angle is just for getting better throttle response, slightly better emissions, things like that. You know that last 1% mentioned above.

[Spine]

Ok, so here's what I purchased:

Spartan Lambda Controller 2 Spartan Lambda Controller with LSU 4.9 Wideband O2 sensor ($125.00)

Now, how easy is it to wire into rusEfi?

[AndreyB]

That would be really easy - just hook up the "analog output" wire to PC3 as default pin-out says.

[Spine]

Will TunerStudio be able to "auto-tune" with the analog output?

I've also "bridged" injector outputs 1/3 and 2/4 via jumper wires on my breakout box. I then re-added injector 3 and 4 back to stock pinout. Changed injector timing to compensate for it (shorter injector times). Was able to rev over 6,000 rpms. I was actually able to drive it around yesterday after doing so. Easily got up to 40mph. It's still running rich on idle, and lean on drive - I haven't messed with any fuel tables yet. Timing sounds and feels good.

[AndreyB]

Full (paid) TunerStudio - yes, once you the three smaller sensors working for Speed Density.

Plain MAP sensor it should be possible to hack TS to tune as well but I have not tried that yet so there are no instructions for that. Hopefully I would have time this weekend to try this & post my findings.

[Spine]

Full (paid) TunerStudio - yes, once you the three smaller sensors working for Speed Density.

I'll need to look into this. So far my IAC isn't functioning or wired in. It's a 4 wire and I'm not sure how to get it wired in. It has IACV lo, IACV hi, IACV hi, and IACV lo. Two of each, hi and lo.

IAC001.png (108.64 KiB) Viewed 16251 times

The rusEfi only has one output: PC13

[AndreyB]

Couple of lines added to http://rusefi.com/wiki/index.php?title=Manual:Software:Idle_control

[Spine]

Parts have been ordered for the IAC. I can't wait to get this thing running right. I've been driving it around as it is, and it's running good. I have noticed that if I shut it off and the temp drops approximately 10deg before I start it again, it starts missing on 1-2 cylinders. It doesn't clear up till I blow out the plugs. It's definitely running too rich (on idle). Based on the gauge in TS, fuel ratio sits right around 11(+/- 0.5). I also have an oscilloscope ordered (DSO203) and can double check everything once it's in.

[Spine]

I must have a missing wire or something...the vBatt is 0. Is this related to the O2 sensor, or is it something else? I noticed it in the TS AutoTune tab.

[AndreyB]

Based on the gauge in TS, fuel ratio sits right around 11(+/- 0.5).

wait a second, there is no sensor so this is probably not true?

[AndreyB]

I must have a missing wire or something...the vBatt is 0

your fully assembled board & default configuration should have it. Please post analoginfo command output