1971 Volvo 142E - MS3Pro Mini Build

[DougThe142]

Starting a build thread for the build I'm working on. I picked up a very clean running driving 71 142E last year that was very clean, but had some mechanical negligence over the past decade. I bought if from the children of the owner who had owned it for 30 years and passed away, that guy's name was Doug (who I named the car after). Doug was a Volvo fanatic who owned a few newer ones as well, but this one was his Sunday cruiser from the 90's. He had the car to drive it, not to look at it, which is my kinda car owner. It had 181,000 miles when I bought it.

Although it made the drive home ok from Colorado Springs to Littleton, CO, I had it almost to the floor just to make it up hills, and there were... noises... I'm no life long Volvo fanatic, knew very little about them coming into this, but similar to the 82 Subaru Brat I bought years back, when I saw it on marketplace I loved it, and when I started doing research on the history of the car and engineering behind it - I loved it even more. So, it ended up in my garage.

The first year has been spent doing maintenance. Replacing tie rod ends, engine/tranny mounts, fuel injectors (trickier task to source the correct injectors and o-rings than I anticipated...), starter, shocks, bushings, ball joints, tires, water pump, thermostat, window seals, having the radiator flushed and professionally repaired where leaking at the spigots, replacing vacuum hoses, fuel pump/regulator replaced, fuel tank dropped/cleaned... probably a couple other things.

Now that all that is out of the way, I've started with the more fun project: aftermarket ECU.

This may be a ridiculous way of going about this, but here is my plan. Because it's running pretty darn good now, I want to collect some data on how the 50 year old ECU from ze Germans is performing. I'm an engineer and I like data - so having a baseline dataset to see how well this open-loop contraption controls lambda as-is seems like it would be fun to review. It would also be nice to quantify the improvements as I take over with the megasquirt. SO - the first time I power up the harness, all the wiring for fuel and spark will be landed on connectors, but not plugged into anything. I want my initial run to have the original Bosch ECU controlling everything with the MS3Pro logging engine speed, throttle position, O2, coolant temp and intake manifold pressure. I thought it would also be cool to use one of the generic inputs to try to log injector pulse duration and spark timing - but I haven't investigated how I would need to wire this up to make that happen. I MAY have access to our chassis dyno's at work, so I may be able to get some really fun data here. Time will tell on that.

Once I've collected some data, I plan to take over fueling. Once that is tuned, take over spark. In a perfect world I'll be able to tune on a chassis dyno for each of these, but that will depend on a lot of factors with work.

I've been trying to take some videos as I go recently to share, I've posted them to this playlist and will continue to post as I go.
https://www.youtube.com/playlist?list=PLmRc6FgTUlUwu4uE63P0pKohtLSYqe6yS

Never posted to a forum before - so created a imgur account to try to share some photos of the car. Hopefully these work,

I am HOPING that an all-input no-output configuration is something that the megasquirt software will support. Reading through the manuals it seems like an infinitely-configurable software package. Anyone ever tried a ridiculous idea like this?

 

[JohnMc]

You could just install a wideband O2 sensor to see how well (or not) the D-Jet is doing.

It was a pretty advanced bit of technology back in the day, but the lack of feedback made it a bit more of a party trick in terms of passing the emissions tests of the day. Without live feedback, any changes in operating conditions (fuel pressure varies a little, timing gets off a little, etc) and it would be off of stochio. What they did was have an adjustment on the ECU itself, and at the factory, they'd get the engine warmed up and idling, hook an exhaust gas sniffer up to it, and twiddle the adjuster so the engine ran clean at idle, with all the manufacturing variances present on that individual car. Whenever the engine was off idle, it wouldn't use that adjustment, it would revert to the hard wired fuel tables, which could be on, or off, unknown really. But the US market tests of the time were only conducted on new cars at idle. So it worked for that specific test scenario. Step on the pedal and it was probably belching pollutants.

The big throttle switch on a D-Jet is not a position switch. Well, it has an on/off idle position switch. And a separate set of alternating contacts that fire extra injector firing events as the throttle opens, it's a low tech accel enrich feature. The injectors are wired in pairs, the ECU manages the fueling as two separate channels, and the distributor has two triggers, one for each channel. So instead of the ECU seeing the throttle open and calculating longer injector open times to enrich the mixture briefly, the alternating contacts just feed in extra pulses, extra injector fires. So you won't be able to read a throttle position using that. If you were successful in logging injector fires somehow (i.e. pulling tach signal from the injector fires instead of the coil?) you'd probably see weird RPM increases when the throttle opens.

All in all, it's probably an interesting exercise in deciphering what ze Germans hard wired inside the cake pan ECU, but really, it's probably more work to get some iffy data. And it's possible someone has already done that in the past, and you could jut peek over their shoulder. I'd just bite the pullet and pull the D-Jet out and replace it with the MS, and spend the time working on the MS configuration.

 

[DougThe142]

You could just install a wideband O2 sensor to see how well (or not) the D-Jet is doing.

It was a pretty advanced bit of technology back in the day, but the lack of feedback made it a bit more of a party trick in terms of passing the emissions tests of the day. Without live feedback, any changes in operating conditions (fuel pressure varies a little, timing gets off a little, etc) and it would be off of stochio. What they did was have an adjustment on the ECU itself, and at the factory, they'd get the engine warmed up and idling, hook an exhaust gas sniffer up to it, and twiddle the adjuster so the engine ran clean at idle, with all the manufacturing variances present on that individual car. Whenever the engine was off idle, it wouldn't use that adjustment, it would revert to the hard wired fuel tables, which could be on, or off, unknown really. But the US market tests of the time were only conducted on new cars at idle. So it worked for that specific test scenario. Step on the pedal and it was probably belching pollutants.

The big throttle switch on a D-Jet is not a position switch. Well, it has an on/off idle position switch. And a separate set of alternating contacts that fire extra injector firing events as the throttle opens, it's a low tech accel enrich feature. The injectors are wired in pairs, the ECU manages the fueling as two separate channels, and the distributor has two triggers, one for each channel. So instead of the ECU seeing the throttle open and calculating longer injector open times to enrich the mixture briefly, the alternating contacts just feed in extra pulses, extra injector fires. So you won't be able to read a throttle position using that. If you were successful in logging injector fires somehow (i.e. pulling tach signal from the injector fires instead of the coil?) you'd probably see weird RPM increases when the throttle opens.

All in all, it's probably an interesting exercise in deciphering what ze Germans hard wired inside the cake pan ECU, but really, it's probably more work to get some iffy data. And it's possible someone has already done that in the past, and you could jut peek over their shoulder. I'd just bite the pullet and pull the D-Jet out and replace it with the MS, and spend the time working on the MS configuration.

The recommendation to install a wideband O2 and see how it is doing is essentially what I am doing. I'm planning to do some datalogging, and I also have a pi set up with a 7" touchscreen running TSDash, this is what I'll be using as my gauge cluster while I'm testing it out. I have some long-term plans for how I'd like that set up - but that is tbd.

For throttle, my currently unplanned plan was to 3D print an adapter bracket that would allow me to run a TPS in addition to the existing throttle switch the D-Jet uses. I'm not entirely sure I'll be successful at this - but I have a vague idea of what I need to do to make this work. It may prove so cumbersome I give up, we'll see.

My other unplanned plan for measuring injector pulse was to tap off the negative side going to the 2 banks of injectors, figure out what signal conditioning i would need to do, and measure them on pins 34 and 35 of the MS3Pro mini. Those are generic configurable dig/freq inputs, so in theory I should be able to measure both if I could sort out the wiring. Realistically I will probably not get around to that. I remember seeing pulse duration data from 1800philes or jetronic.org, can't remember which but I have it saved in my project folder somewhere.

For engine speed I already have a mag pickup installed, just isn't wired to anything yet. Fabbed a bracket with some ugly welds and used the toothed wheel from diyautotune.

You're 100% right it is more work, and questionable payoff - time will tell if I stick to it or just rip out the D-Jet... but when it is all 'future work' it all just sounds fun and exciting, right? lol

 

[bobxyz]

AFAIK MegaSquirt can't measure spark advance or injector duration. You should still be able to monitor: WBO2, MAP, Idle Switch, RPM/Trigger Wheel. Without bothering to look it up, I don't know if MSpro Mini has provisions for traction control using VR wheel sensors. If so, you should be able to use it to monitor MPH (either off a wheel or off a driveshaft bolt head).

What emissions do you need to pass - IM240 dyno, or just stationary? Does it currently pass? (They've been tightening the requirements year-after-year.) Littleton you say - anywhere close to the Lucky Mutt?

 

[DougThe142]

AFAIK MegaSquirt can't measure spark advance or injector duration. You should still be able to monitor: WBO2, MAP, Idle Switch, RPM/Trigger Wheel. Without bothering to look it up, I don't know if MSpro Mini has provisions for traction control using VR wheel sensors. If so, you should be able to use it to monitor MPH (either off a wheel or off a driveshaft bolt head).

What emissions do you need to pass - IM240 dyno, or just stationary? Does it currently pass? (They've been tightening the requirements year-after-year.) Littleton you say - anywhere close to the Lucky Mutt?

For pre-1975 they do not test in Colorado, so none. This was part of why I went this route. I expect when I'm done it will have cleaner emissions than from the factory - but still ideal to not have to test.

Are you stalking me??? Yes, I live within minutes of the Lucky Mutt. Are you in this area?

 

[doucheNozzle]

Great photo

 

[bobxyz]

I made my own hardware logger for LH2.4 a few years ago. It might be adaptable enough for your setup. Have you done any work in the Arduino environment before? See: https://www.turbobricks.com/index.p...osd-logging-elm327.360020/page-3#post-6215757

To measure spark advance, it needs a reference pulse that's fixed to TDC. The MS tach output might work - I haven't tried it on the bench to see if it wanders as RPMs change. It also needs a logic level spark signal, for example, the LH2.4 EZK to ignition powerstage signal. It's not setup to work directly off the coil primary (not well protected against the 100volt+ flyback spikes).

I have friends near the Mutt and am often in the area. We'll need to meetup sometime.

 

[OttoB]

So is the idea record injection data and use that as a base map for new ECU? If so, DJet uses constant fuel pressure but now days fuel pressure is constant over intake pressure. So maps are different. But of course data for example cold start enrichment could be good to have. Then again, DJet uses cold start valve.

NA engines are rather easy to map, and ignition map from 085 distributor is rather close