rusefi.com

The most advanced open source ECU
https://www.rusefi.com/forum/

NA Miata EcotecMiata Race Car

https://www.rusefi.com/forum/viewtopic.php?f=2&t=2294

Page 1 of 3

NA Miata EcotecMiata Race Car

Posted: Sat Mar 26, 2022 3:35 pm
by mtmotorsport
This is our shop race car. It is an NA Miata with some mods to go a bit faster than SpecMiata.

GM LSJ Engine (Saab B207R)
T56 F-body Trans with Quicktime Bellhousing
7163 EFR Turbo

Is pictured as GT4 class, but we broke out and it has been bumped to GT3. For a speed reference, this is 1:31 on 275 AR1's at Mosport. Car is easily capable of going faster but that is the breakout for GT3. Once you move to GT2 you are playing with the big boys in tube chassis cars. Best guess is maybe a 1:29 on AR1 and a couple more seconds on Hoosiers

This car was on MS3, with a separate DBWX2 box for the throttle. This winter we swapped it to MRE (due to the success I had with the MRE on previous projects, and a mid race failure of the MS3PRO ecu). The car has two PDMs, PodiumConnect, AEM VDM, RaceCapture dash, Blink button panel, MK60 ABS, Toyota EPS, so lots of Can devices. Last year we had a teensy based controller to manage the Can messages to control the PDMs, buttons, etc. With Lua being added to MRE I thought this would be a great way to get rid of a device and integrate the control into the ECU.
275771035_1003567607234720_5089399390941070050_n.jpg
277068682_360617136073608_3999355101647392482_n.jpg

Re: NA Miata EcotecMiata Race Car

Posted: Sat Mar 26, 2022 3:53 pm
by AndreyB
Thank you for posting this, feels great to see MRE being useful :)

Please let me know once/if https://github.com/rusefi/rusefi/issues/3970 becomes a blocker for you.

Re: NA Miata EcotecMiata Race Car

Posted: Sat Mar 26, 2022 5:09 pm
by mtmotorsport
Further to the Lua Scripting thread, here is the complete code as it sits on the car right now. Everything is working as intended with one exception. I would like to grab the brake pedal pressed message from the can bus. If it could be set as the brake input for DBW safety that would be great otherwise if I could just relay the brake state to a cantx message that would work as well.

ID = 351211806 is the message that contains the brake state. If byte 1 is 128 and byte 4 is 16 the pedal is depressed. If byte 4 is 0 the pedal is release. This message is continuously broadcast, requires no handshake or tx request.

I would like to pass the state of the brake switch to Rusefi for DBW and also to the txCan message with ID 351215889. This txCan message is already being broadcast and the other functions contained in the payload are working as expected.

Code: Select all

d = 0
df = 0
w = 0
wp = 0
t = 0
tl = 0
h = 0
hz = 0
br = 0
local cm = 0


canRxAdd(418381601)
canRxAdd(351211806)


setTickRate(1)

function onCanRx(bus, id, dlc, data)

	if id == 418381601 then
		bn = data [4]
		bs = data [5]
		if bn == 3 and bs == 1 then
			t = (t + 1)
			if (t % 2 == 0) then
				tl = 0
			else
				tl = 127
			end
		end
		if bn == 2 and bs == 1 then
			w = (w + 1)
			if (w % 2 == 0) then
				wp = 0
			else
				wp = 127
			end
		end
		if bn == 1 and bs == 1 then
			d = (d + 1)
			if (d % 2 == 0) then
				df = 0
			else
				df = 127
			end
		end
		if bn == 5 and bs == 1 then
			h = (h + 1)
			if (h % 2 == 0) then
				hz = 0
			else
				hz = 127
			end
		end
		if id == 351211806 then
			bks = data [1]
			bk = data [4]
			if bks == 128 and bk == 16 then
				br = 0
			else
				br = 127
			end
		end
	end
end



function pdmSetup()
	txCan(1, 351215121, 1, { 0, 1, 255, 0, 255, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 2, 255, 0, 254, 255, 4, 255 })
	txCan(1, 351215121, 1, { 0, 3, 255, 0, 255, 255, 6, 45 })
	txCan(1, 351215121, 1, { 0, 4, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 5, 75, 1, 252, 255, 5, 255 })
	txCan(1, 351215121, 1, { 0, 6, 75, 1, 252, 255, 5, 255 })
	txCan(1, 351215121, 1, { 0, 7, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215121, 1, { 0, 8, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 9, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 10, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215121, 1, { 0, 11, 255, 0, 255, 255, 6, 173 })
	txCan(1, 351215121, 1, { 0, 12, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215121, 1, { 6, 110, 110, 110, 46, 109, 109, 255 })
	txCan(1, 351215121, 1, { 7, 110, 110, 110, 28, 44, 110, 255 })
	txCan(1, 351215889, 1, { 0, 1, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 2, 255, 0, 255, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 3, 255, 0, 255, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 4, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 5, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 6, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 7, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 8, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 9, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 10, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 11, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 12, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 6, 110, 110, 110, 44, 44, 44, 255 })
	txCan(1, 351215889, 1, { 7, 46, 44, 44, 44, 110, 110, 255 })
end

function radFan()
	local fan = getOutput("isFanOn")
	if fan == 1 then
		rf = 127
	else
		rf = 0
	end
end

function fuelPump()
	local fuel = getOutput("isFuelPumpOn")
	if fuel == 1 then
		fs = 127
	else
		fs = 0
	end
end


function onTick()
	if cm == 0 then
		pdmSetup()
		cm = 1
	end
	fuelPump()
	radFan()
	txCan(1, 351215121, 1, { 4, rf, 127, 0, 127, wp, 0, 63 })
	txCan(1, 351215121, 1, { 5, df, 127, 127, 0, 0, 127, 63 })
	txCan(1, 351215889, 1, { 4, tl, br, hz, 0, 0, 0, 63 })
	txCan(1, 351215889, 1, { 5, fs, 0, 0, 0, fs, fs, 63 })
end

Re: NA Miata EcotecMiata Race Car

Posted: Sat Mar 26, 2022 10:06 pm
by mtmotorsport
AndreyB wrote: ↑
Sat Mar 26, 2022 3:53 pm
 Thank you for posting this, feels great to see MRE being useful :)

Please let me know once/if https://github.com/rusefi/rusefi/issues/3970 becomes a blocker for you.
Does this mean available Lua space can change dynamically, while the engine is running? Eg: if my script fits in available space with engine off, it might overflow once engine is running?

Re: NA Miata EcotecMiata Race Car

Posted: Sat Mar 26, 2022 10:12 pm
by AndreyB
No-no-no

This is about us not using 427/429 fully

Re: NA Miata EcotecMiata Race Car

Posted: Sun Mar 27, 2022 4:26 pm
by mtmotorsport
AndreyB wrote: ↑
Sat Mar 26, 2022 10:12 pm
 No-no-no

This is about us not using 427/429 fully
Makes sense.

I wonder if someone could help with this section of script. It seems to pull the 351211806 data correctly, but does not pull the 418384601 data. This also causes a Lua out of memory condition when a 418384601 message is received. The following error is printed in the Console.

2022-03-27_12_12_59_026: EngineState: LUA CAN RX error not enough memory

If I comment out the 351211806 section then the 418384601 part works as intended with no other changes to the script.

Code: Select all

canRxAdd(418381601)
canRxAdd(351211806)

function onCanRx(bus, id, dlc, data)
	id1 = id
	if id1 == 418381601 then
		bn = data [4]
		bs = data [5]
		if bn == 3 and bs == 1 then
			t = (t + 1)
			if (t % 2 == 0) then
				tl = 0
			else
				tl = 127
			end
		end
		if bn == 2 and bs == 1 then
			w = (w + 1)
			if (w % 2 == 0) then
				wp = 0
			else
				wp = 127
			end
		end
		if bn == 1 and bs == 1 then
			d = (d + 1)
			if (d % 2 == 0) then
				df = 0
			else
				df = 127
			end
		end
		if bn == 5 and bs == 1 then
			h = (h + 1)
			if (h % 2 == 0) then
				hz = 0
			else
				hz = 127
			end
		end
	end
	if id1 == 351211806 then
		bks = data [1]
		bk = data [4]
		if bks == 128 and bk == 16 then
			br = 127
		else
			if bks == 128 and bk == 0 then
				br = 0
			end
		end
	end
end

Re: NA Miata EcotecMiata Race Car

Posted: Sun Mar 27, 2022 11:47 pm
by AndreyB
I've made the change I was talking about - please take latest firmware https://rusefi.com/build_server/rusefi_bundle_mre_f4.zip in about 15 minutes from now, this would double Lua RAM if your unit has proper microcontroller.

luamemory rusEFI console command is your friend, that would show how of available RAM you've used.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 12:06 am
by mck1117
AndreyB wrote: ↑
Sun Mar 27, 2022 11:47 pm
 luamemory rusEFI console command is your friend, that would show how of available RAM you've used.
But as he's discovered, it can fluctuate up and down as the script runs. It's best to have significant headroom available to "weather the perfect storm" of memory usage.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 12:26 am
by mtmotorsport
I think there is something wrong with how I'm calling the canRx. If I just call one can ID it uses 65% of available space. Adding the last if statement in the function oncanRx causes 2000% memory usage.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 12:35 am
by AndreyB
please confirm that you've updated the firmware

please provide exact luamemory outputs for different scenarios to give us a bit more visibility

ideally please provide two full versions which behave too differently, "if I change xxx" is too vague to reproduce for me.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 2:32 am
by mtmotorsport
Just loaded this new firmware build.

Everything works as intended in the script below with the following luamemory result

2022-03-27_22_18_50_615: EngineState: Lua memory heap 0: 16828 / 60000 bytes = 28.0%
2022-03-27_22_18_50_615: EngineState: Lua memory heap 1: 0 / 12000 bytes = 0.0%

This luamemory situation is odd because this is the lowest memory utilization I have ever seen. I'm not complaining, obviously, just weird that even prior to this stage when I had fewer functions it was still consuming more memory. This gives me hope that I can code in the PDM current feedback to illuminate the button panel backlights as driver feedback that outputs have been enabled.

Code: Select all

d = 0
df = 0
w = 0
wp = 0
t = 0
tl = 0
h = 0
hz = 0
br = 0
local cm = 0


canRxAdd(418381601)
canRxAdd(351211806)


setTickRate(20)

function onCanRx(bus, id, dlc, data)

	if id == 418381601 then
		bn = data [4]
		bs = data [5]
		if bn == 3 and bs == 1 then
			t = (t + 1)
			if (t % 2 == 0) then
				tl = 0
			else
				tl = 127
			end
		end
		if bn == 2 and bs == 1 then
			w = (w + 1)
			if (w % 2 == 0) then
				wp = 0
			else
				wp = 127
			end
		end
		if bn == 1 and bs == 1 then
			d = (d + 1)
			if (d % 2 == 0) then
				df = 0
			else
				df = 127
			end
		end
		if bn == 5 and bs == 1 then
			h = (h + 1)
			if (h % 2 == 0) then
				hz = 0
			else
				hz = 127
			end
		end
	end
	if id == 351211806 then
		bks = data [1]
		bk = data [4]
		if bks == 128 and bk == 16 then
			br = 127
		else
			if bks == 128 and bk == 0 then
				br = 0
			end
		end
	end
end



function pdmSetup()
	txCan(1, 351215121, 1, { 0, 1, 255, 0, 255, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 2, 255, 0, 254, 255, 4, 255 })
	txCan(1, 351215121, 1, { 0, 3, 255, 0, 255, 255, 6, 45 })
	txCan(1, 351215121, 1, { 0, 4, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 5, 75, 1, 252, 255, 5, 255 })
	txCan(1, 351215121, 1, { 0, 6, 75, 1, 252, 255, 5, 255 })
	txCan(1, 351215121, 1, { 0, 7, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215121, 1, { 0, 8, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 9, 255, 0, 254, 255, 0, 255 })
	txCan(1, 351215121, 1, { 0, 10, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215121, 1, { 0, 11, 255, 0, 255, 255, 6, 173 })
	txCan(1, 351215121, 1, { 0, 12, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215121, 1, { 6, 110, 110, 110, 46, 109, 109, 255 })
	txCan(1, 351215121, 1, { 7, 110, 110, 110, 28, 44, 110, 255 })
	txCan(1, 351215889, 1, { 0, 1, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 2, 255, 0, 255, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 3, 255, 0, 255, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 4, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 5, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 6, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 7, 255, 0, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 8, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 9, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 10, 255, 0, 255, 255, 6, 255 })
	txCan(1, 351215889, 1, { 0, 11, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 0, 12, 255, 1, 252, 255, 4, 255 })
	txCan(1, 351215889, 1, { 6, 110, 110, 110, 44, 44, 44, 255 })
	txCan(1, 351215889, 1, { 7, 46, 44, 44, 44, 110, 110, 255 })
end

function radFan()
	local fan = getOutput("isFanOn")
	if fan == 1 then
		rf = 127
	else
		rf = 0
	end
end

function fuelPump()
	local fuel = getOutput("isFuelPumpOn")
	if fuel == 1 then
		fs = 127
	else
		fs = 0
	end
end


function onTick()
	if cm == 0 then
		pdmSetup()
		cm = 1
	end
	fuelPump()
	radFan()
	txCan(1, 351215121, 1, { 4, rf, 127, 0, 127, wp, 0, 255 })
	txCan(1, 351215121, 1, { 5, df, 127, 127, 0, 0, 127, 63 })
	txCan(1, 351215889, 1, { 4, br, tl, hz, 0, 0, 0, 63 })
	txCan(1, 351215889, 1, { 5, fs, 0, 0, 0, fs, fs, 63 })
end
Thank you guys so much, this is so cool.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 2:38 am
by mck1117
Lowest percent or lowest bytes? We just made the available memory bigger so your percent used went down for the same byte usage :)

Also, some amount of that memory is used by the Lua virtual machine itself - a blank script will still have non-zero memory usage.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Mar 28, 2022 2:49 am
by mtmotorsport
Lowest percent for sure due to the 60k being available vs 25k before. I don't have time right now to load the previous firmware and reload the lua script but I seem to recall it being higher bytes than this iteration as well, into the 20k region.

Re: NA Miata EcotecMiata Race Car

Posted: Wed Mar 30, 2022 8:09 pm
by BeerMoneyMotorsports
Looks rad. Do you think the ecotec swaps would be a good candidate for a pnp ecu?

Re: NA Miata EcotecMiata Race Car

Posted: Wed Mar 30, 2022 8:17 pm
by AndreyB
BeerMoneyMotorsports wrote: ↑
Wed Mar 30, 2022 8:09 pm
 Looks rad. Do you think the ecotec swaps would be a good candidate for a pnp ecu?
i.e. we have a solution (rusEFI) we are looking for a market :)

Re: NA Miata EcotecMiata Race Car

Posted: Fri Apr 01, 2022 1:42 am
by mtmotorsport
E67 plug and play?

Would have to be waterproof, as ecu lives underhood.

stock ecu is $25 from junkyard and operating system supports 5 bar map sensor, so you would have to have a really compelling case to use it. We use standalone because closed loop wideband and built in boost control. OE ecu will not do either of those.

Real good idea is E39 plug and play. Direct injected ecotec variants have very poor HP Tuners support and can never be tuned to run properly in a standalone configuration. Current aftermarket DI ECUs are $1200+ and require at least modifying the stock harness. E39 also applies to HF V6 engine family which is a common miata swap as well.

Re: NA Miata EcotecMiata Race Car

Posted: Fri Apr 01, 2022 1:46 am
by AndreyB
Thank you, noted!

E67 is currently in the works
https://rusefi.com/forum/viewtopic.php?p=44864#p44864

E39 noted, GDI is currently in the works. Any chance you would be interested to some early testing of GDI prototypes?

Re: NA Miata EcotecMiata Race Car

Posted: Fri Apr 01, 2022 1:58 am
by bill
nice setup! any pictures of the engine/ drivetrain??

Re: NA Miata EcotecMiata Race Car

Posted: Fri Apr 01, 2022 5:17 am
by nmstec
I'm working on E39 on a Tahoe.

Re: NA Miata EcotecMiata Race Car

Posted: Wed Apr 06, 2022 3:53 am
by mtmotorsport
AndreyB wrote: ↑
Fri Apr 01, 2022 1:46 am
 Thank you, noted!

E67 is currently in the works
https://rusefi.com/forum/viewtopic.php?p=44864#p44864

E39 noted, GDI is currently in the works. Any chance you would be interested to some early testing of GDI prototypes?
I can help with testing if needed. I have a DI Ecotec with additional port injectors (8 injectors total) currently running on Motec that I could easily make a patch harness to rusefi.

I suppose I should get my hands on a Proteus for testing?

Re: NA Miata EcotecMiata Race Car

Posted: Wed Apr 06, 2022 9:41 pm
by AndreyB
mtmotorsport wrote: ↑
Wed Apr 06, 2022 3:53 am
 I suppose I should get my hands on a Proteus for testing?
Yes please.

As of today please start wiring Proteus to BMW 7506280 on the right. Hopefully in a few weeks we will have firmware for rusEFI GDI4 (on the left) which is pin compatible.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 2:41 pm
by mtmotorsport
Just had the car on the dyno with good results.

Stock LSJ engine
B207 intake manifold
LNF exhaust manifold
Stock location EFR 7163
Bosch 1000cc injectors
17psi flat to redline. Will add some reverse taper to help torque up top but it really doesn't need it.

Rev limiter set to 7200, with soft at 7000rpm.
MRE dyno 17psi.png
Link to short vid

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 3:13 pm
by AndreyB
Can you please upload your tune to http://rusefi.com/online unless it's really confidential?

Wow thank you for the sweet rusEFI bragging material!

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 5:48 pm
by mtmotorsport
Done. Nothing confidential, haha. The tune isn't really polished yet, just calibrated the cell we could reach on the dyno.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 5:55 pm
by AndreyB
I see "sequential" on https://rusefi.com/online/view.php?msq=946&dialog=injectorConfig and CAM input is selected by greyed out on https://rusefi.com/online/view.php?msq=946&dialog=triggerConfiguration

It's greyed out since cam wheel shape is not selected on https://rusefi.com/online/view.php?msq=946&dialog=triggerConfiguration_IO

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 6:04 pm
by mck1117
AndreyB wrote: ↑
Mon Apr 11, 2022 5:55 pm
 I see "sequential" on https://rusefi.com/online/view.php?msq=946&dialog=injectorConfig and CAM input is selected by greyed out on https://rusefi.com/online/view.php?msq=946&dialog=triggerConfiguration

It's greyed out since cam wheel shape is not selected on https://rusefi.com/online/view.php?msq=946&dialog=triggerConfiguration_IO
That's actually OK, I run in the same mode. Unless you really care about emissions, that 50/50 chance of when your injection timing happens doesn't really matter. Sequential gets you one firing per cycle instead of two, so any errors from deadtime etc are reduced by half.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 7:12 pm
by mtmotorsport
This is actually due to a bad cam signal noise issue we are having. I had to disable it and turn ignition to wasted COP. Normally these engines have a Bosch quick start (4 tooth) cam sensor module in the rear of the cylinder head. We removed that because we pushed the engine back about 4" from its normal position. In place of that sensor I installed a hall sensor in the valve cover, attempting to get a single cam pulse off one cam lobe. This works flawlessly as long as the ignition coils do not fire. As soon as ignition is enabled you can "see" each coil fire on the composite tooth logger. I suspect some shielded cable, or better noise suppression will fix it, but right now it runs fine.

Re: NA Miata EcotecMiata Race Car

Posted: Mon Apr 11, 2022 8:30 pm
by AndreyB
mtmotorsport wrote: ↑
Mon Apr 11, 2022 2:41 pm
 Just had the car on the dyno with good results.
what kind of gas did you use? street premium or something fancier?

Re: NA Miata EcotecMiata Race Car

Posted: Tue Apr 12, 2022 12:41 am
by mtmotorsport
91 octane, pump gas. Nothing special. Car will run 1-3 hours endurance races and 25 minute sprint races. Fancy fuel is too expensive. Might get a hold on some E85 and try that. This was the original reason for the 1000cc injectors.

Re: NA Miata EcotecMiata Race Car

Posted: Thu May 12, 2022 3:57 pm
by duhafnusa4
any updates?
All times are UTC
Page 1 of 3

Powered by phpBB® Forum Software © phpBB Limited