Vaughan
-
Posts
2,103 -
Joined
-
Last visited
-
Days Won
74
Content Type
Profiles
Forums
Events
Gallery
Blogs
Posts posted by Vaughan
-
-
2 hours ago, Climer97007 said:
Does the Voodoo ECU support self-learning or do the Maps have to be hard coded and updated manually
G4X and G5 ECUs do have Closed Loop Lambda with long term trims if that is what you mean by self learning, it is always recommended to do a decent amount of tuning first though before relying on Closed Loop Lambda.
-
-
Might be that the original tacho output was from a coil negative and it was looking for a big voltage spike with a threshold of ~14V. If so you will either need to modify the factory tacho (common for MR2s and Celicas when going ot individual coils) or a tacho booster wired inline.
If you want to confirm the signal from the ECU is still operating correctly you will need to put an oscilliscope on it, the ECU pulls low and has a weak pullup to battery voltage, if the ECU isn't actually pulling up to battery voltage when the tacho is connected (scope with tacho connected, scope without, check height of square wave) then maybe a stronger pullup will fix it.
-
Are you using a solenoid from Link or generic solenoid from elsewhere? You'll want your solenoid to have the runner from the turbo feed connected to the gate when un powered and blocked off when powered and you want the gate to be connected to the housing reference when unpowered and connected to atmosphere when powered.
Are you 100% duty (gate connected to atmospheric pressure) when spooling?
Do you know if the gate is opening at all while spooling (can usually be viewed while car is on dyno)?
You could also play antilag, launch control or ignition timing games to try and get it spooled earlier. If the gate is opening when it is connected to atmospheric pressure then you will need heavier springs or a 4 port setup to feed pressure into the other side of the gate to hold it shut.
-
6 hours ago, Sigch said:
What Kind of EVO map should i use and will IT realy start with a four cylinder map qnd serial Lambda and Mass Air flow Sensor
Doesn't really matter which as you are only using the ful and ignition values, not the cylinder count, firing order, ignition mode etc.
I would always strongly recommend adding a CAN based wideband when fitting an aftermarket ECU if you want to have any success with tuning and trouble shooting it.
I would recommend using the MAP signal rather than the MAF as your load source and on your fuel and ignition table axes.
If you haven't already it might be a good idea to download the software and have a look through the various settings and read some of the help manual which explains how everything works. You are going to need to do some wiring which the help manual has some information on as Link doesn't sell a plug and play ECU to suit 3000gt's.
-
22 hours ago, Blue said:
Does the rate at which the blue LED light flashes on the ECU mean anything if that changes?
Blue LED on the ECU flashes the fault codes. No flashing means no faults.
-
You are sitting on your Min Effective Pulse Width value, doesn't matter how much you reduce the VE numbers it will not inject less than your Min Effective Pulse Width Value.
-
The plugin for this was tested on a local car, trigger pinout will be based on that car. The link wants the reluctor signal to go high first and then uses the zero crossing on the falling edge to sync, wouldn't surprise me if the Honda ECU wants it to go low first and then uses the zero crossing on the rising edge to sync. Correct polarity can be confirmed by doing a trigger scope.
-
4 minutes ago, Gtr_lee said:
I did do the VVT angle test when I first got it up and running but I will try it again in case I did not do it correct. I did notice the angle never really changes from 9 degree advanced regardless of target.
You can select the Inlet bank 1 cam in the cam angle test to check the actual angles it is seeing (angle of each tooth shows up in the 'Cam Angle #x runtimes'), maybe try adjusting the duty cycle output on Aux 1 using a gp pwm to see if you can make it move at all, odd for it to be consistently sitting somewhere other than 0deg especially after a successful calibration as that implies that when you do the cam calibration it moves to 0 and then moves back to the 9deg after.
-
Will also crank over faster and get better spark with the spark plugs out making it easier to get it timed before the battery runs out of juice
-
Sweet, I didn't see anything obvious other than the VVT not working correctly so I would start by running the calibration on that in VVT Control -> VVT Setup -> Cam Angle Test -> Calibrate, it'll turn itself off after it's done, you can watch the status of the cam in the VVT tab of the runtime view (F12 or R to bring the runtime view up), this needs to be done while above the VVT engine speed and ECT lockouts.
I assume you've set the timing with a timing light?
I've also attached a copy of the last tune I have from my car before I sold it, was a G4X Fury (so different pinout than the plugin) on a 1.8 VVT motor with factory coils, 36-2 crank wheel, G19 turbo manifold, TD04HL turbo and 700cc injectors. Note it was a modelled tune so the fuel table numbers won't be particularly relevant.
What injectors are you running out of curiosity?
last copy of tune 9-5-2023.pclx
I did find when tuning mine that I had difficulty accounting for how heat soaked the intake manifold gets when you've been sitting idling, would be fine while running and driving but would come to a stop and the pre throttle IAT wouldn't noticeably change but the lambda would despite all else being equal, this might have been solvable by fitting the IAT sensor to the manifold instead of the pre throttle pipe so that the heat soak in the manifold is seen by the ECU
-
17 minutes ago, Rally Mazda said:
I've just pulled the plugs and yeah I've soaked them fml! Looks like it'll be a winch onto the trailer until my tuner sets her up for me
Injection is usually turned off during timing calibration for this reason.
-
-
9 hours ago, JMP said:
what particular computer-specific items cause the glitches? As I've experienced them on all the different notebooks I tune with, and don't see the same issues on previous G4 stuff on the same systems.
The issue is to do with graphics drivers, the reason you see it on this particular build and not on previous builds will be because the issue is related to the graphics libraries as they were when that version of PCLink is compiled. If these libraries are updated between PCLink builds regardless of any change in our PCLink code then these issues can occur, we don't write the graphics drivers, just code that uses them.
We are planning on doing a new PCLink and firmware release soon so hopefully the graphics issues will be fixed by that.
6 hours ago, Laminar said:I guess I'm interested in understanding the upside of a nag box popping up every time I launch the app. I don't understand who it helps - is it your R&D team or some tuners with multiple ECUs live at one time? From a one-person one-ECU point of view there's no upside.
The change was made so that people with the new G5 Voodoo Pro ECUs with Wifi could find and connect to them. It also means that instead of PCLink sitting there trying to connect to everything it can find, which is what it did previously, you can now direct it to connect to a particular device.
There is a slight change coming for the next release where if Connection Port is set to USB and only one device is found during a regular connect attempt then it will try to connect directly to that. Shift + F3 will still do the full search.
-
Can't really help without a copy of your tune and a Pc log of the issue happening
-
You're trying to trigger the AC Clutch and Condenser Fan relays with a high side signal from inj5&6 which can only do low side control. Same issue with Thermofan on Aux 3. Aux 7 controlling the starte solenoid can do high side but you generally would try and use low side triggering for this.
I would also recommend moving the coil power for the fuel pump relay and isc solenoid relay from the ignition switch signal to the ECU power side of the relay switch so that if the main relay fuse blows it doesn't power up those devices while the ECU is not powered causing all manner of weird back feeding issues.
If it were me I would run Coils, Injectors, ISC Solenoid and ECU all off of one 30A capable relay and fuse, I only see separate relays for those things as added complications and added room for issues to occur with no real benefit. I would also have the power side of the coil for all other relays (fuel pump, starter, fans etc) coming from the ECU power on the ECU side of that relay.
Your pinout also seems to be setup for an Xtreme not a Fury as you are trying to use B plug pin 30 as a digital input when it is one of the dedicated wideband pins on a Fury and you have a CAN based wideband instead of wiring the sensor directly to the ECU.
-
-
Your PC Log shows that it is trying to fire the coils and injectors. To get the trigger scope runtimes you need to use the ECU Controls -> Trigger SCope function and click capture while the engine is cranking.
-
Have you set the ignition timing with a timing light?
-
Have you done the APS calibration? Can you take a PC Log of you moving the pedal between 0% and 100% and attach it here please.
-
4 minutes ago, don17 said:
is it possible reaching max ignition?
Doesn't look like it's flatlining, don't have your tune but assume you're max is set to 45deg.
5 minutes ago, don17 said:So the so the lambda is richer than gauage, I don't think I have the gauage ground wire to and actual sensor ground it is grounded to body.
Which gauge do you have? Best way to connect it to ECU is CAN if that is an option, gauges with separate grounds for the ECU signal than the power ground are not terrible, gauges with no separate analog signal ground will always be terrible.
-
How accurate is the Lambda reading compared to the gauge? Seems very rich everywhere.
Also the ECT Ignition trim is advancing the ignition angle not retarding it (positive values not negative).
-
Graphical issues are compiler related and so not something we can change in the code, more of a look to see if they have been fixed since by various updates to the build computer. It's also not something I've seen on my own computer so isn't necessarily going to be the easiest to test if it has been fixed.
No current plans to remove the ECU selection box, it does work better in Manual mode as auto mode will try and fail when you power cycle ECU or disconnect and so lose the selected ECU.
27 minutes ago, Timboj said:Also, another minor issue I've had with PC Link G5; the software (at least my current install) doesn't discriminate between what you're saving/loading when it comes to the last used folder. What it means is that if I started by opening a map for example, when I go to open a log file, I have to change folders. And vice versa.
There are some setting around directories in the options menu with tick boxes and editable directory paths. Do you have the boxes ticked or not (ticked means always use that folder, unticked means go to last place you opened that type of file from).
I don't think that setup changed at all on the G5 update stuff.
-
Which engine do you have? Quick google shows the RS3 uses the Bosch HDEV5.2 which we have the current and voltage data for. The RS3 high pressure fuel pump we don't explicitly have data for yet but a quick scope with a current trace will show it's peak and hold current values.
You will also need to measure your high pressure pump cam lobe shape and find the angle of the top of one of the lobes relative to tdc of cyl 1 when cam is in it's rest position.
Hardware wise we have 4cylinder box on the way with built in fuel pump peak and hold driver but I don't have any specific timing as to when this will be available to purchase.
I would also assume that the XBow has a CAN bus system which will need reverse engineering so that we can replicate what the factory ECU transmits and receives.
Electronic Boost Control Routing
in Engine Tuning
Posted
If the gate isn't physically opening and you aren't leaking exhaust gases past the gate then the only way to improve spool (without adding something like fresh air antilag) will be to play with ignition timing.