Megasquirt Update

[luthier]

So yes I did have the MS manual. I got it to startup today but it failed to start.
I calibrated the TPS but everything else seemed to have less need. The IAT sits in a bottle up by the fuse box. The CLT is sitting connected to another sensor waiting for calibration. The MAP seems to self adjust?
There was no fuel at first but I found the correct terminal out of the Tachy relay plug and can clearly hear the fuel pump run when that circuit is activated. I wonder if that activates both pumps. But I smell fuel after cranking.
When cranking I found that the MS went offline with the load. I couldn't find an appropriate output wire from the ignition to avoid this so I ran a temporary wire direct from the battery with an on/off switch, for testing purposes.
A couple of times there was a small fire but only for nano seconds.
I checked for spark and it was great. But the plug was dry. Haven't cracked the fuel line to check the fuel pressure.
I studied your map and copied parameters to my new one because yes your map was not compatible.
So I'm wondering if the O2 sensor actually needs to go on now , whether it's actually in the exhaust or not.
I have an issue with the way it should hook up however.
It has 4 wires, red, black, white and yellow.
Red and black are to power it, Yellow is signal but white can be hooked into the headlight circuit or just earthed. I suppose that's the earth that goes to MS along with the signal? It's not exactly spelled out.
That's it for me today. Not surprised. I would have been astounded if it went first go.

 

[robs]

My mistake. I thought my Microsquirt was fully up to date, but checking at msextra I see there have been two firmware releases since mine. Yours will probably be one of those. Can you tell me what your TunerStudio window title bar says. Something like:

TunerStudio MS v3.1.06 - delarue ( MS2/Extra 3.4.2 release 20160421 11:50BST(c)KC/JSM/JB uS )

Does yours say 3.4.3 or 3.4.4 where mine says 3.4.2? Is your TunerStudio at 3.1.06? If I get mine synced up to your firmware version, I'll update the msq and you'll be able to use the whole shebang.

First thing to establish is that you have a consistent tach input. While you're cranking, does the TS tacho show anything at all? Does it jump around or reset? It's not worth worrying about anything else until you have a steady tach reading.

It'd help a lot if you posted a data log here. Set up plain text logging in TunerStudio:
Options->Data Logging Preferences->Data Log Format->ASCII Format(msl)
then generate a log file:
Data Logging->Start Logging
accept that, and crank the engine for a couple of seconds. Then stop logging. The log file will be in your project's DataLogs folder. Tons of info in that file.

Your specific questions:
MAP sensor requires very simple calibration: Tools->Calibrate MAP/Baro. This lists common sensors, including yours I hope. Otherwise you have some research to do. Sanity check is that sitting still your MAP should be around 100kPa.

CLT thermistor needs custom calibration. Tools->Calibrate Thermistor Tables. Values for the standard 505 sensor:
Temperature1: -10(C), Resistance 1: 9600 Ohms
Temperature2: 20(C), Resistance 2: 2500 Ohms
Temperature3: 80(C), Resistance 3: 330 Ohms

You'll need something similar for the MAT sensor, but the values depend on the thermistor. DIY should have provided that info one way or another.

For TPS, CLT, MAT, V Batt and everything else the TunerStudio BasicLoad Setting->Real-Time Display is a good sanity check.

Can't help with the fuel pump wiring. S2 wiring diagram should help.

You should probably power the MS from a relay. It uses the voltage at its input to determine important things like injector dead time. I didn't do this, but its ideal setup would take MS power from the +12 feed to the injectors.

You definitely don't need the O2 sensor at this stage. I ran mine for a few months without one. I then installed a narrow band for a year or so before getting a wideband. That fourth wire taking power from the headlight circuit sounds like it's to dim the gauge at night.

Have fun,

Rob.

 

[luthier]

OK I have spark and fuel pressure at the rail. But no pulse at the injector plug.
And you are correct, there's no Tacho input.
There's a wire labelled ignition but since this is a fuel only install I thought the Tach signal would come through the OPTO leads.
Maybe they are connected wrongly. Would that cause damage to the unit? Can I reverse them to see?
I can't view the Data as I haven't yet bought Megalogviewer and don't intend to until next week.
I could send you the file, but I think my problem is simply lack of Tach input.
Interestingly in the pinout list it says OPTO- to negative coil and OPTO+ to negative coil. Maybe that's the secret?
But then in the description it says OPTO- to positive coil and OPTO+ to negative coil.
I think I have very good reason to be confused.
What do you think?

 

[luthier]

I found out how it should be.
Positive OPTO goes to Negative coil and Negative OPTO goes to ground.
Got it started but won't idle , needs throttle constantly flicked to stop stalling.
Now for all those calibrations.

 

[robs]

Glad you've got it to fire up.

I'm sure you're suffering from information overload. Section 6.1 of the manual is quite clear that OPTO+ goes to coil -ve and OPTO- goes to coil +ve. I doubt it'll hurt anything to put the OPTO- to ground, but it's best to go by the book. Ignore what I say if you got your advice from someone like Matt Cramer.

You're right about the ignition wire. It's to fire a coil. Microsquirt has two of these to support 4-cyl wasted spark.

If you log data in ASCII as suggested, you can look in the file using wordpad or whatever. Not hard to load into Excel either I expect -- tab-separated rather than comma separated.

Anyhow, I want to help, but you keep ignoring MY questions. What firmware version? Is it 3.4.4?

If you want to try getting it going reasonably well by yourself I suggest:
1. open the thumb screw idle air valve (turn it anticlockwise) until it keeps running as is.
2. warm the car up.
3. increase req_fuel value in the very first menu and see if it improves or gets worse.
4. keep fiddling with req_fuel and the thumb screw until it idles reasonably about where the thumbscrew started.
If you've loaded the VE table I gave you, that should be OK for most things. Accel enrichment would be your next thing to work on.

Have fun,

Rob.

 

[luthier]

Yes Rob, it's MS2 Extra 3.4.4
Well it's very strange about the OPTO connections. I know it says that the OPTO- connects to Coil + but it had no signal that way.
But it did have signal with OPTO- to earth but ran very badly for awhile. But then it stopped showing a tach signal again and hasn't started since.
I'll hook it back on the positive side and see if it has got over it's tissy fit. Just hoping the unit was not damaged.
So I checked for spark and that was OK but I thought maybe the coil or the ignition module were playing up. So I swapped all that around to no avail and now it just doesn't want to play.
I've had enough for another day.
My next effort will be to clean the main power leads to starter and engine block as the starter has been intermittent and lots of power gets pulled at intermittent times. I may need a new starter. So I am basically going over the health of the car in an attempt to get things going.
I found that my plugs went from too lean to much too rich since running with MS.
The timing advance on the MS gauge was reading an excessive amount, starting round 15 and going up to 40 and there was no way I could increase the throttle slowly through the range as it would cut out at about 1/3rd throttle. Sometimes it would start with no foot on the throttle, other times it responded to a small amount. Obviously there's plenty wrong at the moment. I undid the air hose that goes through to the airfilter thinking there might be too much drag through the disconnected MAF but that made no discernible difference.
I find the fuel map thing very hard to manipulate in 3D form and not really sure what values should be entered.

I found adjusting the tables really the only way but I need to know what values to enter there. I'll check yours out in that way.
Thanks.

"If you log data in ASCII as suggested," can't work out how to do this.

 

[robs]

The OPTO input will count a trigger when the OPTO+ goes more than about 2V above the OPTO-. The coil negative should spike up to a couple of hundred volts when the spark is triggered. By comparing it with +12V rather than ground makes it much less likely to get false triggers, but it should still get all the true ones.

That it was very rich when running suggests you were getting lots of false triggers. High or jumpy RPM might have confirmed this. A log file really helps with finding things that you didn't spot at the time.

By the looks of this forum question it is possible to fritz the OPTO circuit. that guy having burnt out R20.

First thing to do is to get an ASCII log file. In TunerStudio (as mentioned above):

Options->Data Logging Preferences->Data Log Format->ASCII Format(msl)

then generate a log file:

Data Logging->Start Logging

Crank the engine for a couple of seconds. Stop logging. Look in the file or post it here (you may need to change its extension to .txt to post it). If the RPM is absolutely dead, it suggests your R20 (no, not a Renault) may have bought it.

If so, all is not lost. That resistor is only used in the OPTO circuit. You could try triggering from the reluctor sensor (bottom of the dizzy) using the VR input. It's lower voltage and has a better reputation. The worry with that is that when I used it, the Megasquirt put too much load on the signal and prevented the Bosch igniter from sparking. I believe the Microsquirt presents a smaller load, so this may not be a problem.

Don't worry too much about the MS gauge showing ignition advance. Your distributor is still in control of the actual advance. That the gauge got to 40 degrees suggests the MS thought the RPMs were pretty high. More evidence for false triggers.

I agree about the 3d table adjustment. Much easier just to type the values in to the 2d table. If you want to enter the tables I sent you, there are export and import icons above the table in TunerStudio. These let you save and load just that table.

I'll update my Microsquirt to 3.4.4 and generate you a complete .msq which should be very close for your car. It'll then be a matter of sorting out hardware problems -- triggering and sensors.

Happy to continue my "support desk" role here (or we could move to e-mail). You might prefer to join the msextra forum. Matt Cramer and other very experienced Megasquirters regularly help people there.

Have fun,

Rob.

 

[luthier]

Hi Rob,
The news today is I have no Tach input through OPTO but there is through my old dwell/tach . So good chance I have fried the R20 resistor.
And I had just worked out how to import your fuel tables so I was expecting a better day. Glad there's another option though maybe I should send the unit back for repair. I _might_ be able to replace it myself and I see on that forum there is mention of making a small circuit with diodes to protect MS there, which sounds like a possibility.
I will make one of those files which should prove whats happening.
Happy to email if you are and it looks like that forum is worth checking out.
Thanks so much for your support.
My email is kellawayguitars@gmail.com

 

[luthier]

Here's that log file, not sure if it works here.

 

[robs]

I have switched to e-mail which makes sharing files a lot simpler.

For any following along, the log file shows no sign of life from the tach input, so R20 may have bought it. Suggested using VR input from the distributor/ignition module instead. It's generally considered more reliable anyway. Also pointed out that the very low temperature readings in the log file show the thermistors aren't calibrated correctly which probably added extra fuel when it was running.

Have fun,

Rob.

 

[luthier]

After some fine tuition and bright ideas from Rob I managed to get the car running quite well today.
The initial tach input I used was the 'OPTO' wires which derive their input from the coil. After a series of ways to wire this I got a signal but that disappeared after some very rough running.
I then decided to split the output from the ignition module that send pulse to the reluctor in the dizzy. These were connected to VR+ and VR- on the Microsquirt and a good clean tach signal was obtained. I had also found a pinout for the Peugeot module here: https://www.pim-engineering.com/tiedostot/ignitionmodules.pdf [Ours is the one ending in 123.]
Very relieved to get it running and now itching to take it for a spin but much tidy up has to happen first. The way it revs out so freely is very inspiring.
Thanks Rob.

 

[lowpugV2]

I have just read through this, super interesting! Very impressed with the sharing of knowledge and luthier just having a go.

It did get me thinking though, you mentioned you have the GTi and the STi? Why not low boost turbo the auto STi? So nice revvy na GTi and effortless cruiser STi? You know, put these new skills to use?

 

[luthier]

I'm a little way from it being properly sorted lowpug, but steadily working through various issues.
Not idling properly yet but today it went from lumpy and pretty horrible to beautiful and smooth after a refill of 98 fuel.
That was weird because the old fuel must have had issues. Then I think that I should retry previous fuel tables that Rob supplied which seemed to just run way too rich. I generated new tables giving the standard specs for the engine in terms of HP and Torque and this smoothed it all out but doesn't have the wow factor that I experienced a few times with Robs tables.
So Rob suggested checking out the fuel pressure reg which could be causing issues. Also I'm fitting the resistor plug versions of the BP6ES which are BPR6ES as these make less 'noise' electronically.
O2 sensor gets fitted on Wednesday which should help with the tuning too.
It's taken a few days to get here, the worst being the first few of cutting the original loom and the uncertainty of what lay ahead after spending all that money. And then the no go at first which was expected but disappointing. But it's going to be so much easier next time once all the issues have been dealt with. Not sure about the turbo though, that's another leap in the dark for me. Maybe spark management instead of just fuel management could be a better result though the VE connection from the reluctor wires appears to be stable so far.

With the computer removed I managed to mount everything on that plate behind the glovebox, the fuse block, the relays, the MAP and the Microsquirt unit along the bottom and all the wires and the vacuum tube for the MAP so far have gone through the original rubber bung in the firewall with no mods.

The O2 sensor however has quite a bulky plug and loads of heavy wire so I've worked out it can come up next to the gearstick and then feed through the console at the bottom below the stereo in that funny bottom shelf which is where I'll jam the gauge.

Under the bonnet you'd hardly know anything had changed unless you really knew your 505's. There's an alloy plate I cut to adapt the Hyundai TPS. There's a bunch of new wires going across the firewall ledge which will be encased in a black concertina insulator , a smaller version of the one the big fat loom came in, and the VE shielded set which comes off the reluctor wire from below the ignition module and I'm leaving the black cover off the coil as when it was fitted the signal failed, no idea why, but as soon as it was removed I was back in business. It's either a dodgy crimping that got stretched or electrical interference from stuff crammed together too close, I suspect the latter.

Plus two new wires to each side of the battery. Hardly anything at all. It will become more obvious when I remove the old MAF and replace that section with a 4" hose. I'll try to find some black stuff though. Wouldn't want one of those bright blue thingos the young punks stick on their cars to impress their mates.
But as you suggested the most interesting thing is that a complete novice can be guided through this process by an old hand like Rob who has the patience of a saint and a seriously analytical mind for sniffing out issues and suggesting directions.

 

[robs]

Thanks for the kind words Dan. Of course I have been enjoying it too, and you're the one who gets the skinned knuckles, etc. It will be interesting to find out where all the extra fuel is coming from.

Strictly speaking, the plugs should be BPR6EFS, though I don't think the projected tip (F) makes much difference anyway.

Must go. I've got a halo to polish.

 

[luthier]

Just got back from a long day on the road. The car is now running exceedingly well! It's very smooth and has a lot of quiet power.
I had the oxygen sensor fitted last Wednesday and by the end of Thursday, with Robs help I had it properly configured. The effect was immediate. Suddenly it ran better and idled and started easier.
I made a modification to the original Mass Air Flow meter rather than ripping it out. Rob tells me it will still be more restricted than without it but what I did was take off the top of the working part, drop out the clip that retains the return spring, rotate the wiper around to fully open and then rotate the spring wheel around so the spring is fully opened up and holds the opening section fully open, then replace the clip to hold it there, then silicone the top back on so it all looks original.
To fit the O2 sensor I had a guy weld in the bung which he provided and claimed that the one that came from megasquirt was trouble.
He did a nice job and charged me $50. He's located in Billinudgel and he's old school. Good bloke.
Then I had to deal with quite a lot of wire that couldn't be cut and shortened.
I fed the plug from the sensor on the pipe up through the gearstick tunnel and then coming from the other way drilled a big hole using a dremil in the side of the bottom shelf below the stereo. I fed all the wires from the O2 gauge in through there and with it all clipped together that big pile of wire disappears in behind the console. Then the gauge head just sits down there blinking Air Fuel Ratios at you all the time.
Once the O2 is fitted you can then drive about with the computer plugged in to the microsquirt and enable the Autotune function. Every now and then you stop , save and burn the new figures acquired this way. Apparently the process is best done for a couple of weeks. It's important to drive smoothly to allow time for the setup to make the best assessment and the best changes.
I have never driven such a smooth 505GTI! And I'm finding the power absolutely lovely. It's not earth shattering but you leave most cars behind on big hills and it's kind of torquey almost like a good HDI.
The extra power kind of relaxes you to sit back and enjoy the ride.
By next week I'll probably feel confident to enable the other two functions that also help with economy.
The first is the EGO control which makes changes on the run but they are only temporary and don't get burned to the brain.
The other is the overrun fuel cut which turns off fuel during a drive when the throttle is off. It's a clever thing that is dependent on the revs so as you come down to idle fuel is returned so you don't stall at the lights. Apparently this function can return about 5% fuel savings just on it's own.
I can't thank Bob enough for his help and patience with me and I can now thoroughly recommend this mod to anyone who isn't completely phased by a handful of wires and a bit of computer setup.
I'll repost once I have economy figures .