Finally... Microsquirt'd MB 190E has commenced

Well, I'm still plugging along but progress seems to have slowed recently, due to a variety of things.  And the stuff I have been doing doesn't really warrant much in the way of pics.  The wires going to the MS are terminated, as are the wires for the fuzes.  I have a temporary wooden box setup for the MS and the fuzes.  Once I've got it running, I'll go back and fab up some sort of metal box / tray to contain the battery as well as the MS and fuzes.  I was thinking I'd take a trip to the JY in the near future and find a nice fuze/relay panel cover off a Benz and build the box so I can use that as the cover.  Should look fairly trick. 

I'm now starting to go backwards through the engine and put it back into "ready-to-start" shape, which means installing gaskets, taking the tape off the intake ports, etc. 

As part of that, I did undertake the task of putting a "proper" bead on the Al tube I have in my intake tract now.  It's just a run of the mill 2.5" dia, .065" Al tube.  Not having a bead roller, I was going to have to do this the GRM way.  So, first up, a trip to HF for a pair of $5 pliers.  I welded on a "hump" on one side, and ground away a roughly similar spot on the other side.  

Unfortunately, the Al tubing just laughed at my initial attempts to crimp a bead onto it.  All I did was leave a mark on the inside of the tubing.  So, I had to put my thinking cap on, and I managed to come up with this Rube Goldberg-inspired setup.

Much to my delight, this worked quite well.  Glad I put the welding gloves on.  The result:

A few minutes w/ a flap wheel on a die grinder took the sharp edges off, and it was ready to go.  

I need to do a final wiring pin checkout, but after that I think I'm ready to try to talk to the MS.  All bets are off once I get to that point.

After tying up some loose ends (some of which I'll have to redo once I'm confident everything is wired up and operating correctly), it was time to start trying to talk to the Microsquirt.  The laptop I'm using for this is old enough that it has a serial port (!) so I could plug directly into it, and after installing TunerStudio, it connected right up without issue.  Walking through the New Project screen, I immediately hit my first issue - TunerStudio recognized the uS as having 3.4.1 firmware, but it didn't have the .ini file for that version, so the first round ended quickly.  

Some puzzling and Googling later, I was armed with both the 3.4.1 and 3.4.2 releases on a thumb drive (the ancient laptop in no way ever gets connected to a network).  TS was able to move forward once I pointed it to the 3.4.1 files, but the common wisdom seems to be to update to 3.4.2, so next up was a firmware update.  Got that done without issue, so I could finally start messing around.  

First, TPS cal.  Seems I've swapped two of the pins, so it reads a high count on closed throttle, and a low count on open throttle.  TS asked me if I was sure about this, but it seemed to accept it just fine when I hit yes.  If any MS experts know if running like that is a bad idea, please let me know!  After that were the MAP and temp sensor calibrations.  I couldn't find a coolant temp cal in the library that was remotely similar to what I was measuring, so I had to do my own calibration curve.  I got a ~25* F reading with this approach:

So that's where I am now.  Exciting to be moving into new territory, but I'm certainly not making any guesses as to when things will start happening!  As odj pointed out when he was down helping the first time we applied power, after about 15min of 12V power with all the fuses in "nothing is smoking or burning so that's a good sign."  Indeed!

Cool stuff!

My time will be coming soon as well with regard to jumping to MS or µS in my 190E. If you haven't been over to 190Rev, definitely check it out. Activity has definitely dwindled in the last few years, but plenty of info is still there.

Update!  

First, I had significant teething problems getting the Microsquirt to play nicely with the RPM input coming off the coil.  Actually, I never got that to work.  Two units died within about 60 sec of first getting that signal.  So, for anyone else reading this later, I'd suggest avoiding the "negative coil in to MS optoisolator OPTOIN" route completely.  

After some close reading of the 190's wiring diagram, I discovered that the tach signal was actually easily available on a screw terminal next to the "X11 diagnostic socket" that they put on '80s MBs.  That was alarmingly easy to access, and a quick check with the scope showed a nicely behaved signal, perfect for feeding the MS via the VRIN pins.  

So for the last month or so, in between all the other stuff I've had going on, I've been trying to chip away at this.  First was to sort out the odd idle valve behavior which was a result of it working opposite what I thought.  If you read back a bit, you'll see I wired in a combo IAT/MAF sensor, and while I'm not currently using the MAF sensor data to run the engine, I enabled it to get a air flow reading, which was crucial in sorting out the idle valve.  

Somewhere in the middle of all this, I got out of my "messing with the car" mode and went into full "engineering test mode" that I have often earned a living with, and the progress has been steady since then.  

I've now got it idling beautifully in closed loop, and it will start without touching the throttle with coolant temps from about 40*C and up.  So, not fully there yet, but getting close. 

Now that I'm this far, I have to do some more fab to get it drivable, and then I can start with tuning at load. 

So far it's been incredibly educational and enlightening, but at the same time I can totally see how these standalone projects do not turn out well for a lot of people.  

I'm not sure where the last 3 months went, but things are moving forward.  Once I got the tach signal sorted out, the Microsquirt itself became very reliable (knock on wood), and the car would start every time without fail.  Since I do not have a dyno at my disposal, I had to take a step back and start doing a few things to make the car actually drivable.  

For example, for quite a while I did not have an throttle pedal - I literally laid a throttle cable across the back of the engine bay so that I could reach it with my hand.  Not ideal, clearly.  Eventually I figured out how to make the cable end that I needed (there's another thread on here for that).  So, throttle cable, check.

Next up, somewhere to mount the O2 controller.  For the initial go-round, it was literally laying on the floor.  Between the shape of it and the length of the O2 sensor wiring, there was just no convenient place to put it that was out of the weather.  After some poking around, I chose to mount it in the passenger fender, behind the wheel well liner.  It should be nicely protected from everything in there, and with the ECU sitting just above it where the battery originally lived, it made for a short run of wiring.  I put my Binky hat on and made up a bracket for it, since the controller is a weird cigar-like shape with absolutely no mounting provisions.  I had to extend the original O2 harness to reach - for now I hardwired in an extension, put the whole thing in a fire sleeve, and p-clipped it to the backside of exhaust heatshielding.  

And finally, somewhere to mount the Microsquirt, a few fuses and a relay.  Originally I hacked together a wood (yes, wood) half-box just so I had a place to mount it.  My plan was to integrate a new box onto the original battery tray and shift my lawn tractor battery to the edge of said tray.  I took a trip to the junkyard with a piece of cardboard in hand roughly the size I wanted the box to be and set to looking.  Turns out a mid-90s S-class Mercedes had a perfect latching cover, so I snagged that.  After some cutting, bending, drillling, hammering, shaping, dollying, welding, and probably other stuff, I ended up with this little setup.  I have to admit I'm pretty happy with the way this came out. 

Apparently I didn't take a pic of the final version of the box, which includes a place for the harness to come in on the side visible in the pic.  Then I finally put everything back together, reattached wires, wired in the relay, etc.  

Ignore the purple wire, I was testing something.  I'd still like to clean this up a little inside, and I need to add a pad of some sort on the bottom to protect the wires from chafing (thus the blue rag).  But in general I'm reasonably happy with how that turned out.  And much to my pleasure, it started right up.  

I drove it around the parking lot last Sunday without issue.  This Sunday I'm hoping to actually drive it on the road.  We'll see...

I like the box setup. Very clean!

And the 190 has returned home!  Got up super early Sunday so I could drive it home from the space where I work on it without a zillion cars on the (DC) beltway, then went out with a friend later in the day and did some tuning work.  Took it all like a champ.  I finally figured out my vacuum leak only becomes obvious under braking, hopefully its the janky hose to the booster, I've already ordered a fix for that. 

It's still running pretty rich, so I'm gradually changing the VE table to approach the appropriate AFRs.  And it bucks when holding or just backing off the throttle.  Very noticeable in 1st, less so in 2nd, pretty much gone by 3rd.  The tuning we did Sunday improved it, but it's not gone yet. 

Otherwise, stoked to have it back on the road!

I am so ready to ditch CIS-E... so ready...

Just don't have the time to take the plunge. I've got a gnarly vac leak as well, and my attempts at making my own smoke machine have ended in everything but success.

Vac leak only when braking, and using a vac booster, sounds like the booster itself and not the hose.

Nice! I've efid a couple of my cis cars. It was worth it. 

Sounds like you're still running speed density. 

Is your ISV hose plumbed to your post MAF intake hose? If not, when you change over to MAF fueling, your idle air will not be metered by the MAF. Tuning mayhem will ensue! 

bradyzq said:

Nice! I've efid a couple of my cis cars. It was worth it. 

Sounds like you're still running speed density. 

Is your ISV hose plumbed to your post MAF intake hose? If not, when you change over to MAF fueling, your idle air will not be metered by the MAF. Tuning mayhem will ensue! 

Also a boost dump valve should dump to the turbo inlet so it is not dumping measured air.

bentwrench said:

Also a boost dump valve should dump to the turbo inlet so it is not dumping measured air.

There's that word again, Dave.  See, it's just so obvious that people assume the car has one :D

pres589 (djronnebaum) said:

Vac leak only when braking, and using a vac booster, sounds like the booster itself and not the hose.

Yeah, the hose is the easy button and needs doing anyway, so starting there.  The booster seem much harder to find - all the usual Euro parts places have no stock.  

bradyzq said:

Sounds like you're still running speed density. 

Is your ISV hose plumbed to your post MAF intake hose? If not, when you change over to MAF fueling, your idle air will not be metered by the MAF. Tuning mayhem will ensue! 

It's plumbed that way, but I've no plans to run MAF-only at the moment.  If I get really adventurous I might fiddle with using the MAF as a secondary load sensor or however they word that in the Microsquirt world.

odj said:

bentwrench said:

Also a boost dump valve should dump to the turbo inlet so it is not dumping measured air.

There's that word again, Dave.  See, it's just so obvious that people assume the car has one :D

I assumed boost was involved because he was putting beads on the air tubes....

Up until somewhere last year, RockAuto offered a rebuild service for the boosters. A quick check 2 minutes ago revealed that this was not available anymore, just like the crate M102's they used to sell... I ran into a booster issue myself 2 years ago and I ended up getting a good used one from eBay. I found out there are 2 models; a slim one with a large diameter and a thicker one with a smaller diameter. I have the slim one and I remember the thick one was way easier to find.

Supposedly Autozone has the one I need.  Ship to home is the only option, so not in a local store, but hopefully that's correct.  I have the slim one.  ATE I think.  

Otherwise, put about 80 miles on it yesterday doing normal car stuff.  I need to do some ASE and WUE tweaking, but it's running well.  Fun to have it back.  

Hey I stumbled on this thread from another 190e build thread. I'm curious what the MAF is doing in your setup? Also if I'm reading correctly you're using the RPM signal to trigger fuel only? I'm looking to convert my 190e to megasquirt and trying to follow the path of least resistance because I'm new to all of this as well.

In reply to NorseDave :

Hello Dave!! Me and a buddy are currently in the process of MicroSquirting his 190e.  Do you mind elaborating on which VRIN (+ or -) you connect to the stud next to the X11 diagnostic socket for the RPM signal? Thanks a bunch!!