A question for the MS experts on here - I've got my 190 project to the point of trying to start it and get it running. I've followed the MS manuals and getting started guides as best I could. I'm running a fuel-only implementation right now, ignition system from pre-MS conversion is untouched. Now I can almost always get the engine to catch for an instant while cranking, and fairly often it will run for a second or two or maybe even five. But I can't get it to idle steady, or run steady at any rpm if I give it some gas. The fact that it will run for even a little bit suggests to me that the wiring is okay (along with the fact that all the gauges for inputs on TunerStudio look reasonable).
I've been steadily increasing the cranking pulse % on the Idle setup screen, and also the ASE and WUE values. On the cranking pulse and the ASE tables, I'm starting to get up around the values where the little help button says are usually the max an engine might need. For cranking pulse I think it says something like "an engine will usually need 100% to 400% of the fuel during cranking" and I'm closing in on 400% , but it's only ~70*F, not -20*F which is where I'd expect you need the really big numbers there.
Since it seems to die even when I try to give it some gas, I'm wondering if I should increase the fuel everywhere (ie, the main table) at least for now, and not just the crank/idle settings? Since the crank, ASE, and WUE are all %'s based on the main table, if the main table values are too low that will skew the start settings, correct? I've pulled a plug a few times during the process, and it's never been wet.
I have a WBO2 wired in, but some previous discussion suggested it shouldn't be turned on until something like 5-20s after startup. So, I've not got there yet. I have closed loop idle off, so the WBO2 doesn't seem like it should matter, yet.
Thoughts? I'm enjoying this whole process, but I'll enjoy it more when it runs! Much thanks for any ideas!
Turn the WB on, and wait to start until it reads something. I'm betting that you are way too lean.
I’m not all that well versed in MS, just tuning in general. But if there is a way to change the injector size, that would be a fast and universal way to change the fuel. Make the injector something like 10% smaller in the calibration.
Generally you can get engines started with the default tables, if you have reqfuel set correctly. Double check injector size and the presence of fuel pressure (3-4bar ideally).
A quick check is starting fluid, if it runs longer on that, more fuel.
As mentioned above, easy button is to raise req_fuel manually (you have to key off/on for it to take) and see if it runs longer. If so, add fuel either in VE table, warmup or ASE... generally I cheat to get them started quick by just fudging req_fuel and leaving the rest, get it warmed up, then start adjusting warm VE then go back and tweak WUE and ASE.
Ok, cool, some stuff to try. Thanks for the ideas, will give these a shot.
Does Micro/Megasquirt have a setting for what the nominal fuel pressure is? Or does it just assume it's about 43psi? That seems like a really fundamental setting, but I haven't found anything where it lets you enter anything about the fuel pressure.
NorseDave said:
Ok, cool, some stuff to try. Thanks for the ideas, will give these a shot.
Does Micro/Megasquirt have a setting for what the nominal fuel pressure is? Or does it just assume it's about 43psi? That seems like a really fundamental setting, but I haven't found anything where it lets you enter anything about the fuel pressure.
Assuming that the fuel pressure is regulated to manifold pressure, the MS shouldn't care- as the injector flows are determined at that delta pressure. So you are putting in the fundamental fuel pressure when you enter the fundamental injector data.
Megasquirt doesn't care what your fuel pressure is (unless you are using it to control pressure), you just have to tell it what the injector flow rate is in the at what ever fuel pressure you are running. If you aren't running the injector at the pressure they are rated at, correct for the change in pressure to get the updated flow rate. https://www.deatschwerks.com/fuel-calculators/flow-vs-pressure-calculator
Enter that in the "required fuel" window under Basic/Load Settings > Engine and Sequential settings
All of the starting and running fueling values are based on that millisecond pulse width value that gets calculated in required fuel. If your required fuel value is half of the actual, for example, everything (VE, ASE, WUE, etc.) will be double in the settings. The suggested value ranges are only applicable if you have required fuel set correctly.
You can get everything to run right with in incorrect value jammed in for required fuel, it's just easier to check for realistic numbers and use the tunerstudio help if you have the value correct from the start.
Making progress! Cranked the req_fuel number up a bit and got it to run for a bit longer today, not great, but long enough to get the coolant up to 60*C, which it hasn't seen since before I started this project. I also got to wondering if there was a setting for the tach in, since I'm doing fuel only and using the coil as input (no trigger wheel on this car). Looking at the Ignition settings, I had completely ignored these and originally it was set to a 60-2 trigger wheel I think. Did some reading and reset the "Spark Mode" setting to fuel only. However, it's not entirely clear to me what the correct settings for "Skip Pulses" and "Ignition Input Capture" are. I assume "Number of Coils" should be 1, since that's all I've got.
This changed seemed to help, sort of. It ran for a bit longer now, but it was hunting between about 1700-2300 rpm according to the car's tach. However, the virtual tach in TS was going crazy, bouncing between 0, reasonable values, and stuff like 10k and 20k. So that's clearly wrong. Reading more, in Section 6.1 of the Microsquirt Hardware Manual, the "coil negative" approach shows the coil (-) running to the Opto + input (pin 30) and coil (+) running to the Opto - input (pin 31), which strikes me as a bit strange. It doesn't help that the wiring pinout (pg 16) says "Coil negative tach in" for both of those inputs. I switched my wires at the coil and can't say I saw much difference.
I haven't hauled out the scope yet, but I'm guessing the signal coming off the coil is super noisy and that's causing all the erratic behavior. My initial thoughts are a capacitor across the coil terminals, making the pair of wires going from the coil terminals to the Microsquirt a twisted pair, and/or running a shielded pair.
Thoughts?
I did log some data, but based on my 3 minutes using the MegaLogViewer, I absolutely hate that thing, so i'll have to look at that stuff with my usual tools for this (I basically did this sort of thing - looking at telemetry via plots, etc. - for the first ~15 yrs of my career so I'm pretty picky on how I want to play with big data sets).
In reply to NorseDave :
LOL- I'm always converting other people's data to use my work plotter. It's super nice to use.... I wish I could use it at home.
It does at least sound that the engine speed input is the core problem.... Not that I can help fix it. Do you have an ignitor box (like and MSD) you can use? You can use the tach out signal on one of those.
