Modern Vespa

Edit: I have no idea if this will work, but figured someone might find this thread entertaining/useful.
Here's a link to the Microsquirt: https://www.diyautotune.com/product/microsquirt-engine-management-system-w-8-39-wiring-harness/
I will be using this as my guide: https://www.efisource.com/docs/MicroSquirtHDWManualRS.pdf

The TPS sensor is finally mounted & MAP sensor temporarily mounted with zip ties. I'm using a TPS kit for a Weber DCOE, with a custom bracket & lots of filing. Standard GM MAP sensor will be used, with the vacuum line run to the port on the intake manifold originally used for the charcoal canister.

I also already have a wide-band o2 sensor mounted in the typical location, but was never able to fully satisfy the OEM ECU. I have a PLX converter box outputting a narrow-band signal, but never figured out the correct load resistor value needed in order to completely remove the original sensor. One idea I had was powering the new sensor from the original circuit... but I have no idea if the wide-band & narrow-band power requirements are the same.

Partial to do list:
- actually purchase the Microsquirt ECU.
- mount thermostat (Koso BF550023 with 22mm adapter)
- obtain fuel injector plug
- figure out if controlling the ignition timing is possible
- identify crank position signal. There's a plug near the front of the motor that should have it.[url][/url]











Last edited by AnnDee4444 on Fri, 14 Aug 2020 03:47:09 +0000; edited 1 time

Question for anyone:

I'm concerned about the OEM idle air control valve messing with the tune at idle. I'm not planning on controlling this with the Microsquirt, and I've always felt that it was very sporadic during warm-up. With the OEM ECU, the motor would frequently stall when I believe it was attempting to close the IAC as soon as it got to operating temperature.

Would it be a good idea to 'fake' the temperature sensor value going to the OEM ECU, so that the ECU would believe the motor is already at operating temperature? There are a few things that I have considered, but would like other opinions:
- I believe that the Microsquirt should be able to control idle without an IAC
- I think the fan sensor is separate, so it should still operate as normal. I can always have the Microsquirt take over fan control also.
- Since the Microsquirt won't be controlling ignition at this time, the ECU would be locked at a fixed point in the ignition table. I'm using an aftermarket exhaust, so I'm not really worried about warming up a catalytic converter.

Another thought I had was to adjust the throttle stop (with something temporary). The IAC is supposed to only be active when the throttle angle is exactly 5.7 degrees: https://cheekythoma3.wixsite.com/itsme/idle

Small update:

Received the Microsquirt yesterday, did a bench test with a 9V battery. Got the MAP & TPS sensors to respond appropriately. I'm considering re-fabricating the TPS sensor adapter bracket out of steel, as the aluminum's flimsiness is not something I want to gamble on.

To-do: Determine the crank trigger wheel pattern. I believe it's on the flywheel. Luckily, I have a spare that I can check. Per the manual, an alternative is using the negative wire of the coil, but this will only work for a fuel only conversion, so it will be a last resort.

More to-do: I'm currently running a PLX wideband O2 sensor, with the narrow-band output to trick the OEM ECU. It has an output that will be fed to the Microsquirt when the time comes. The issue I have now is that the original heater-circuit does not have the proper resistance, and it's triggering a CEL (possibly tuning changes as well). I've tried various resistors, but haven't found the correct value yet... but now I have a new idea: Power the wideband controller directly from the original heater circuit. There's a few unknowns:
- The PLX controller requires 30W max (2.5A @ 12V). Can the OEM heater circuit support this?
- A wideband's current draw varies depending on exhaust temperature & run time... not sure if the OEM circuit is going to consider the changing loads as satisfactory.

Just checked the flywheel/crank trigger wheel: it is a 24-2, which somewhat standard. The Microsquirt should have no problems with this.

The Koso temperature sensor doesn't have a "common sensor value" in TunerStudio, so I had to measure the impedance at three different temperatures and create my own curve. Here's the results from my test:

38° F = 130,000 Ω
77° F = 43,250 Ω
196° F = 4,400 Ω

Still need to work out the air temperature sensor, but I'm thinking about a standard GM/Delphi DETS10072, due to the ability to easily mount in the rubber intake tube.

I'm loving reading your project reports but sadly it is way beyond the limits of my understanding or comprehension. Please report if/when to manage to achieve what it is you are trying to achieve.

I should have written up a "why" when I started this thread... The purpose of this project is that I'm not satisfied with the tuning options available for the GTS 250 (which has the MIU1 ECU). I need to change the tune due to the extensive modifications: polished V4 head, bigger bore, Polini Cam, port matched PM Tuning exhaust without baffle, higher flow air filter.

I've tried multiple solutions, but none have worked for me. Here's what I've tried: Malossi ForceMaster 2, other tuning boxes that modify the lambda sensor or fuel injector, 3-4 ECU tuning programs, various fuel injector sizes.

If I had a GTS300 (or at least the MIU3 ECU), the solution would be a lot simpler & cheaper, but I didn't think it was worth trying to swap the ECU & wiring harness, and probably the gauge cluster and many other things I'm not considering.

The solution I came up with is to build a second 'piggyback' ECU from MegaSquirt, leaving the OEM ECU in place. The MegaSquirt (actually the MicroSquirt model) can take over the fuel injector and/or ignition. To do this, you need to either tap into the OEM sensors or add new ones. The TPS was the biggest hurdle (besides actually building the tune), as it required figuring out how to physically mount the sensor. It would have been easier to source a different throttle body, but I wanted to retain the ability to swap back to the factory fuel injector plug if something goes wrong.

To be honest, it's a little beyond my understanding as well. I feel that posting about it forces me to understand it better... If I can understand it enough to explain it, I should be able to do it.

so far the only thing I see as a possible mistake is breathing through the stock throttle body. there is some extensive tuning talk on the aprilia forums. there was data of running the larger bv350 throttle body on the 300 engine. the guy had dyno graphs and the TB gave two hp over a large rpm span. it was a 30hp build. 16hp baseline. when the larger TB was added it went from 24 to 26hp if I remember right.

I am aware of Zirri's success with the BV350 throttle body, and have actually attempted to purchase their manifold & camshaft that were fabricated for a while. The biggest challenge with the BV350 throttle body is that it uses the MIU3 ECU, where as the GTS 250 uses the older MIU1 ECU. All of the 300 ECU's are MIU3 (I think), making the conversion to the 350 throttle body easier (still need to change the ECU pigtail, re-pin, & reprogram). I believe that the MIU1 & MIU3 gauge clusters are incompatible with each other also, meaning that to use a BV350 ECU at a minimum I would be looking at:
1. new throttle body
2. new gauge cluster
3. possibly new wiring harness, which would need to be modified anyway
4. unsure if the new ECU would function correctly without ABS & traction control
5. still would need a custom tune & the hardware to mount the throttle body, probably some specialized software to make the immobilizer happy

If I can get the MegaSquirt ECU to take over the fuel injector & ignition, I could theoretically eliminate the Piaggio throttle body & ECU completely, and use any throttle body that I can figure out how to mount. MegaSquirt also gives some interesting tuning options that wouldn't be possible with the Piaggio ECU, such as real-time auto-tuning, barometric correction, launch control, ability to take advantage of <1bar MAP sensors for forced induction, nitrous oxide ignition retard, flex fuel tuning, configurable traction control, etc.

As I mentioned in my PM. I will be attempting something similar on a 155 3v. I have yet to find a head polisher yet as that might be cheaper than buying a used head and doing it myself. I have the cam, exhaust, etc.. already just have to find a house and a scooter first.

I'm just going the conventional piggyback route and leaving the lambda system open loop. No ignition changes for me as that seems to complex for a grocery getter or chicken chaser as we call them here and would also involve a dyno which will be 750kms from the potential house.

I will be following your project with great interest! Barometric correction is an absolute must even if the machine never sees a mountain. Density altitude is an ever present variable!

I know how much work goes into these projects. Just saying I would Have started with final size TB. tuning is everything in the end. there a good bit of work to redo if you start out with the tiny stock TB then switch later. and ITS HP man. dont leave that on the table.

I have a good bit of tuning experience and success with turbo 4g63 cars. tuned AEM version 1 from scratch with perfect results. I never liked the whole tuning industry. tuners get you on a expensive and likely far away AWD dyno do some WOT pulls. make some nice numbers for you. but you leave with no drivabilty tuning. and thats where all the time is if you want a nice driving vehicle. I could never take someones money and give them something I wouldnt enjoy.

I dont see much if any talk about real tuning on scooters. Its frustrating. I like what you are doing here and will possibly take similar route. I really want to turbo a scooter something fierce. I just dont want to spend a mound of time doing it. there just a big whole in the market when you start researching small engine ecu tuning. much less with a turbo.

It's really not that difficult to polish the head yourself, just do some research and take your time. I didn't think porting was necessary on the Malossi V4 head, and only re-shaped the exhaust port slightly. I'm not planning on any dyno tuning at this time, and all work will be performed in my garage. I'm not quite sure how well I can get MegaSquirt's auto-tune feature to work... there are quite a few challenges I will be facing (such as: how will it do with the CVT? My guess is OK, but I can't find any documentation of this being done before).

What's interesting is that the engine can be run with various speeds & loads while not moving (on the center-stand) by modulating the throttle & rear brake. This should allow for getting a rough tune before I even ride it, which wouldn't be as easy with a car. Agreed. I would bet it's even more important than ever when dealing with a small displacement highly tuned engine. Honestly, my goal isn't about maximum power, which is why I didn't decide to change the trottle body. My end goal is to prove to myself that I can make it function, and to use my experience learned here for other future projects. I don't really care about it being the fastest/most HP scooter, because there will always be something better. I mean... it's a scooter. If I wanted to go fast, I would have started with something faster.

That's not to say that I wouldn't love to attempt forced induction, but I will be doing it for the irony rather than the racing aspect. The idea of a NOS bottle mounted to the battery cover has been floating around it my head for way too long.

With today's modern casting methods I wonder if porting is really necessary? OK smaller ports can give better torque etc. so maybe the OEM designers optimize for that and size accordingly but porting to rectify original manufacturing?

Blend exhaust and inlet I can see to avoid eddys but as you say they are scooters and we are not racing. Or if I can locate some larger valves just for a laugh.

My motives are the same as yours but on a lower scale.

I am also going to tune mine unloaded on the stand using a piece of 60's optical technology know to any geezer who has had SU carburettors. On the 3v engine the place of the 4th valve is conveniently taken up by the spark plug which is accessible through the cylinder head access panel.
There will be no load but I can make adjustments based on RPM and let the stock ECU/TPS figure out the throttle opening etc.

Progress is slow due to the weather.

I've purchased & installed a GM IAT sensor.

Also measured the resistance of the fuel injector, it is 15Ω which MegaSquirt defines as 'high impedance'. This is good, as the fuel injector can be connected directly to the MicroSquirt without any extra resistors.

Still need to work out the crank trigger type. This is going to take some time to research.

I purchased a super-cheap oscilloscope from Amazon to measure the crank trigger output. It It appears to be a VR type sensor, which I determined from it's AC output. The peak voltage at cranking speed was around 4V. I have yet to determine if this means the high RPM voltage will be beyond what the Microsquirt can handle. I'll probably install a shunt resistor just to be safe.

https://www.youtube.com/watch?v=drh8Ebhobg8



The signal wire is orange on my GTS 250. I'll be building a harness so I don't have to modify the OEM wiring.

No notable progress has been made in the last few months, however a new product from SIP has been released that allows for uploading new tunes.
Unfortunately, it looks like you're at the mercy of the SIP tuners and can't actually modify the tunes... but it's better than any other option for the 250 right now.

https://www.sip-scootershop.com/en/product/fuel-injection-module-sip-chip_51416551

Bumping this thread in case anyone is watching... you will be interested in this: Megasquirt on a GTS 250 engine... In a Genuine Buddy 125.

I am looking at the timing maps. I am perplexed at how aggressive the part throttle advance is on these things. I guess I am used to looking at turbo timing tables which are vastly more conservative.