Help with building your Speeduino, installing it, getting it to run etc.
User avatar
By Kinetic Research
#42905
Hello all,

Starting a project here as I will soon be a fairly regular member due to the nature of my setup. I have an unconventional build layout that is using 2 x Honda J35Z2 V6's transversely mounted in a mid engined car. For the last few years I have been thrashing out the mechanical detail (the build log is here - https://www.pistonheads.com/gassing/top ... 827105&i=0) and am now at the point of applying for a 4B build authority from the LVVTA here in NZ.

Over the next few years I will be constructing an engine testbed (not a dyno unfortunately) and getting the J35 running without load. Initially the exoframe testmule will use 2 speeduinos - each controlling its own standard production engine - but that drivable test mule is still a couple of years away due to paycheck limitations! So gives me plenty of time to get my head around the electronics side.

I have not seen anyone setup the J35 yet, so will be posting all my learnings here.

Longterm goal is to pick up enough tuning skills to regrind the base circle of the stock cams and increase the rev limit and push the top end on these wheezy motors - short term is to just get the test motor up and running clean and revving happily. Off to RTFM a few times before asking any FAQ's.

Cheers,
Marcus.
Attachments
ergonomics.JPG
ergonomics.JPG (119.55 KiB) Viewed 7165 times
Lockdown progress.jpg
Lockdown progress.jpg (167.01 KiB) Viewed 7165 times
Last edited by Kinetic Research on Thu Dec 16, 2021 2:43 am, edited 3 times in total.
By ric355
#42913
I've just finished reading your PH thread. Very interesting project and looking forward to seeing it develop further !
User avatar
By Kinetic Research
#44948
Thanks Ric, much appreciated.

A major goal for this project is getting the engines to make decent power up to 7500 rpm, which with current gearing will give an honest 200mph. The aerodynamics of the body are proven up to 240mph, so even with 2 stock 200kW engines it should not be insurmountable. The speed runs that Mclaren conducted were done with a non production engine that was down on the final 627hp, so that helps build confidence at the practicality of reaching the double ton. The transmission is from a 2016 Corvette C7 Z06. The suspension and brakes are from a C6 Z51 corvette.

Attached are the snaps of the CMP and CKP sensor trigger wheels.

The cam signal (CMP) is taken from some bosses on the back of the cambelt pulley on the front cylinder head (highlighted yellow). Easy to access so next weekend I will pull off the cover and take some photos of the real installation to make sure they didn't reuse images in the service manual - assumption is the root of all F-ups. It should be easy to then work out the timing for the decoder. I am really pleased Josh is working on ardustim soon, I will be needing it to test this pattern shortly. Life is all about timing.....

The crank signal (CKP) is taken from the 60-2 pulse plate at the output end of the crank.

At this stage I think the stock sensors are hall type - the wiring diagram shows a power feed, ground and signal wire for each.
Attachments
J35Z2 pulse plate CKP.jpg
J35Z2 pulse plate CKP.jpg (141.25 KiB) Viewed 7168 times
J35Z2 front pulley CMP.jpg
J35Z2 front pulley CMP.jpg (103.88 KiB) Viewed 7168 times
User avatar
By Kinetic Research
#45109
The manual was correct, the cam pulley has the signal blocks protruding from the backface, which the hall sensor reads.

There are 46 teeth so each pulley belt tooth covers ~7.83 degrees cam ~ 15.65 degrees crank.

TDC mark is the detent in the housing near the top mark. 10 teeth later is the centerline of the Hall sensor ~39 degrees.

Using this 39 degree offset when the cam hits TDC on cyl 1 then the block stamped 2 will be "on", where 39 degrees after TDC on cyl 5 will be "off", so those will be the synch signals.

The missing teeth on the CKP sensor appear before TDC, but I could not accurately measure it, need a degree wheel or a trace of the cranking cycle.
Attachments
J35Z2 Cam signal.jpg
J35Z2 Cam signal.jpg (348.27 KiB) Viewed 7034 times
By theonewithin
#45115
Don't worry about trying to figure out where TDC is.

Once you have Speeduino setup you use a timing light to verify timing then.

At this point you can then change trigger angle until commanded matches actual.
User avatar
By Kinetic Research
#45129
Thanks, a good tip.

My challenge is to capture the decoder pulse train positions so that a pattern can be created for this motor type. Its tricky due to not owning an oscilloscope or having a starter motor connected to the test engine yet, so trying to determine the CMP signals relative to the crank angle and CKP signals. Luckily the engine uses hall sensors so a degree wheel and multimeter across the 5V signal wires should be enough to work it out (will of course have to connect sensors to 12V power and ground).
Last edited by Kinetic Research on Thu Nov 17, 2022 3:35 am, edited 1 time in total.
By ric355
#45133
Kinetic Research wrote: Mon Aug 31, 2020 2:53 am Thanks, a good tip.

My challenge is to capture the decoder pulse train positions so that a pattern can be created for this motor type. Its tricky due to not owning an oscilliscope or having a starter motor connected to the test engine yet, so trying to determine the CMP signals relative to the crank angle and CKP signals. Luckily the engine uses hall sensors so a degree wheel and multimeter across the 5V signal wires should be enough to work it out (will of course have to connect sensors to 12V power and ground).
Get yourself a logic analyser. They are available very cheaply and the software is free. You can then feed the hall sensors with a 5v signal and turn the engine over to log the patterns.
Example: https://www.ebay.co.uk/itm/24MHz-8-Chan ... 3662291815
They are only 5v tolerant so you have to be careful how you use them, but are very useful during the implementation phase when you're wondering if something is working right or not.
User avatar
By Kinetic Research
#45140
Thanks for the help, electronics are not part of my skill set - yet. Always happy to add another tool or instrument to the collection, especially when the cost is low. Little wisps of blue smoke are bad right?
User avatar
By Kinetic Research
#45523
Since the Auckland harbour bridge was damaged and commuting has been a chore I took the scenic way to work this morning. Lots of quiet country roads to allow recording of the ignition advance curve for the stock Honda ECU at 100% throttle, up some reasonable inclines to give a decent resistive load and thus a longer sampling window, using an OBD2 scanner/logger. It seems the ECU reverts to these values once more than 25% throttle is used. Interesting to see the effect of the VTEC on the advance required.
J35Z2 ignition advance curve.jpg
J35Z2 ignition advance curve.jpg (67.79 KiB) Viewed 6675 times
By infinityedge
#45582
Cool project!

Have you considered using GM LFX V6s instead of the J35Z2? The LFX is lighter and produces more power, plus the header is integrated into the head making exhaust routing in tight spaces much easier.

I uploaded logs and last tune and dyno plots look[…]

you could set 'skip revolutions' to 3, then it wou[…]

Hi, I am trying to assign Signed values to the x-a[…]

blitzbox

I've finally figured out why MAP and Lambda weren'[…]

Still can't find what you're looking for?