There is more to life with TurboRenault.co.uk

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Alpine A610 Turbo - Megasquirt Conversion

Hehe... I didn't realise Adaptronic wasn't that well known - I was more concerned that no-one would have understood Megasquirt, it being DIY and all that!
Runcorn is 5 to 6 hours away from me. That makes for two very long days as I couldn't imagine driving there and back and getting the mapping all done in one!
 
I didn't realise Adaptronic wasn't that well known
Yeah tiny, its only been around 20 years and is distributed globally : ) (Thats sarcasm btw)

The only UK dealer is EFI, so maybe not huge in the UK but if you check the distributor list its got good global presence. Its all out there if you look.
 
Yeah tiny, its only been around 20 years and is distributed globally : ) (Thats sarcasm btw)

The only UK dealer is EFI, so maybe not huge in the UK but if you check the distributor list its got good global presence. Its all out there if you look.


Oh I love your sarcasm.....its almost as good as mine. Cock.
I'm well aware of its recognition outside of the UK but thats of zero help to me is it? Runcorn is considerably closer than Santa Fe, Argentina. Although there are now 2 more UK dealers, one in Birstall and the other in Stockport. Ive never heard of Birstall and Stockport sounds......well.....all a bit Northern to me.
 
thats of zero help to me is it?
The problem is more 'reputable tuners' IMO, we all know that Scoff and RST have excellent service and reputation... thats as much the reason I travel as anything else. Too many cowboys out there that will take you for a ride and do a shit job. If you have an answer for "who do I trust to tune my car" thats local, then by all means go for it. specially in your case the hard parts are done, base setup and cold start, transients, config like that. Its just a map and boost tune.
 
Another quick update. I think that running the engine with no exhaust is starting to annoy the neighbours a bit! I realised I couldn't tune the car properly without a working idle air control valve. So I decided to build the circuit for it.

The underside of the Megasquirt now looks even worse!

yhaRp4RgSmx7Sb_ENL9c016921458b7e76ee87a137ed5edffe.webp
Hmmm... not the tidiest, but it did work, eventually!

Two problems - the first was that I wired the valve up backwards with the opening and closing the wrong way round. I didn't have a pinout for the valve anyway, so I just guessed. Swapped them round inside the ECU as that was the easiest place to undo and redo it. Then I found that the ECU power was dropping out every time it tried to energise the valve. Finally it dawned on me that I'd used the wrong 12v source for the controlling circuit. I'd used "sanctioned" 12v, which is a sort of smoothed (but not regulated) 12v, but it appears not to allow much current. I switched to using 12v raw and it all worked!

I proved the valve was working and used it as a simple on/off device. That doesn't seem to work properly - I need to have the valve open a bit under normal idle conditions, so now I will try PWM closed loop idle with the valve. I ran out of time after fitting the exhaust to do any more tuning. I have a bad exhaust blow from somewhere near the turbo but I can't keep the engine running without feathering the accelerator, so that will need to wait until I've tuned the IAC.

I was also finding that the air/fuel ratio was a bit all over the place, generally tending towards the rich side, so I did some research on injector dead times. I originally used the manufacturer's specs, but maybe wrongly. I found someone else that used Megasquirt with the same injectors. They'd found the injectors had a much lower dead time than I was using, so I think that's been throwing off all my air/fuel calcs. The dead time is added to the injector pulse width, so the injectors open for the time specified. If the dead time is too long, they will be open for longer than desired.

Andrew
 
I proved the valve was working and used it as a simple on/off device.
So the valve you are using is a pain in the arse.

The two wire stepper type IACV that I mentioned is a very basic PWM setup that runs from one aux output. You feed 12v to one side, the other back to the ECU and then it regulates at 25hz.
04.webp

The rotary three wire takes 12v in, then the other two pins are the opposite of each other so they fight it out to balance the valve. Note the two aux outputs and that one is inverted. Im not sure if this will help you out and i'm not sure how you 'invert' a PWM signal but thats how it is :)

03.webp
 
Thanks Dave! I had another look - needs more investigation as MS software seems to only have settings for 3 or 4 wire stepper IACs.
Good news is that I enabled PWM control of the three-wire IAC and it works! Very nicely. I can set 0-100% and it regulates the idle properly. It was actually pretty simple. For now I have used open loop PWM IAC control. It progressively closes as the engine warms up. I'll make it closed-loop later, but the settings for that are quite complex and I may not even need it.
Screenshot at Apr 10 22-42-28.webp

Screenshot at Apr 10 22-43-01.webp
Screenshot at Apr 10 22-43-25.webp

The engine now idles properly from cold to warm (although it's hunting a little bit and the air/fuel ratio fluctuates quite a bit), but it's all coming together. I prefer the original IAC because of how it fits and is installed/where it goes.

Could the air/fuel ratio fluctuation be down to the fact that I'm using such large injectors? Apparently they are supposed to be very good at small pulse widths. Currently hovering at 1.5 to 2ms.

Now I have another major problem in that I can't seal the exhaust to turbo connections properly. It's made of small sections with "Bischoff" clips, whatever they are. Apparently they are supposed to be renewed whenever they are removed. Yeah. right...
 
I love MS! Everything is so well documented and the community is awesome!

The bad name comes from people thinking they could do it themselves while they have no clue, it's a great platform, you just need to install it properly!
 
Hehe... very funny. I am indeed doing my best. So far, so good. Just done another cold to hot tune. Idling not too badly now. Just need to get the car together properly so I can go for a gentle drive with Auto Tune doing its magic. NO WOT runs!!!

Yes, I am very impressed by MS. There is much to learn and I'm impressed by its feature set, the community and the wealth of info out there. I get the basic concepts but I am learning all the time.

Andrew
 
Yep. Using Lambda for fuel tuning. Guessing timing advance. There's absolutely zero info out there on it so I have no idea where to start. I will start safe and go up from there.
I dialled in a few extra degrees advance at low RPM to help get a stable idle... Yes, I know there's a big jump from 600 to 1100 RPM.

Thanks for the reminder on not tweaking the laptop. I know and I must be absolutely 100% about that.
Screenshot at Apr 11 11-58-23.webp


Andrew
 
Does this help? J-series 4cyl but i'd imagine it wont be a million miles away. Standard internals.

Why is your grid only 0-100 Kpa? Shouldnt it be 0-200 on a standard MAP sensor? Note mine is to 300 Kpa (2 bar boost)

01.webp
 
Thank you - yes it does.
And I just realised I made another big mistake! My fuel and spark tables only go up to 100kPa!!! I need to extend them to 200kPa. Darn. That means re-tuning my fuel at idle that I just spent ages doing. It's due to me assuming 100kPa would be the MAP reading at peak power and torque. Well, it won't. It'll be 200kPa - one bar boost. Again, same mistake as before, but I did generate the initial tables before I realised the significance of the MAP reading.
Suddenly, my AFR table goes from:
Screenshot at Apr 11 12-04-54.webp
to:
Screenshot at Apr 11 12-05-52.webp
And my VE table goes from:
Screenshot at Apr 11 12-07-41.webp
To:
Screenshot at Apr 11 12-10-18.webp
Updated spark advance map:
Screenshot at Apr 11 12-21-15.webp
For the spark map, I picked a few values then interpolated the rest. I am going to stay well away from boost until I get the base map dialled in. Think I'll peg the wastegate open.
Oh well, live and learn.
Andrew
 

Attachments

  • Screenshot at Apr 11 12-19-30.webp
    Screenshot at Apr 11 12-19-30.webp
    121 KB · Views: 6
Thanks for the tip on the wastegate.

Yes, I think I do. RPM x axis, MAP, y axis. Ignition and fuel are looked up on the grid, interpolated between cells. Fuel is base fuel required multiplied by the value in the cell, giving a pulse width for the injectors. It's a bit more complex than that but I have studied the Megasquirt docs on this in quite some detail. Here's the extract from the manual. I am using Speed Density as the fuel calculation algorithm:

The primary load input is the MAP sensor.
With 'Multiply MAP', without 'Include AFR target'
PW = DT + (ReqFuel * MAP * VE[RPM,MAP] * AirDen * BaroCor * corrections)
With 'Multiply MAP', with 'Include AFR target'
PW = DT + (ReqFuel * MAP * Stoich/AFRtarget * VE[RPM,MAP] * AirDen * BaroCor * corrections)
 
Back
Top