wastegate

Have put a task in for this, not sure when it will be done.

Few local dealers are advertising it and listed as in stock..
EKZ in Australia have it and listed @ $5150au
 
To be honest, it's a tough sell. People will still upgrade I'm sure, it's a easy upgrade if you are already on G4X an want extra IO or some features.

However, there is plenty of competition in the arena. 
 
I wish it had Gearbox control, even CAN based minimum and support gearboxes like DCT/8HP over CAN to the factory TCU. Don't expect it to have 6+ h-bridges to drive them directly. But supporting 3-4 speed autos directly would have been nice too.

I used ethanol as its the first parameter in the list.  It is just being used to generate a constant, since we are multiplying it by zero then adding an offset it will always output a constant no matter what the actual %ethanol is reading.  
 
They are not meant to turn off, the assist level controls the flow rate of the hydraulic fluid through the system.
 
Yes, that is the intent, you can put whatever axes you like on there.  Typically wheelspeed Vs a multipostion adjustment knob on the dash.

Can't do 2104136 at ~71.4hz and 1ae0092c at ~2.38hz, closest you can get on that with custom CAN would be 50Hz and 2Hz.
1ae0092c is a hex number and is '450890028' in decimal format, this will be an extended Id
I assume 2104136 is also a hex number based on the above code, this will be '34619702' in decimal format and will also be an extended CAN Id.
To get the fixed values in the streams use any runtime and give it a multiplier of 0 and an offset that matches the fixed offset as shown below for the 0x1ae0092c ID:
local data_0x1ae0092c = { 0x00, 0x00, 0x22, 0xe0, 0x41, 0x90, 0x00, 0x00 }

Note that I've done it as 2 16bit numbers instead of 4 8 bit numbers, 0x22, 0xe0 = 0x22e0 = 8298.

The following code says that the 2nd byte (assuming 0 list indexing in the above code) in the 0x1ae0092c id message changes each sent message.
slowRoll = (slowRoll + 1) & 3
data_0x1ae0092c[1] = slowRollTable[slowRoll + 1]
The first message will have '((0 + 1) &(bitwise AND) 3) + 1 = 1+1 = 2' in that byte
The second will have '((1 + 1) & 3) + 1 = 2+1 = 3'
The third will have '((2 + 1) & 3) + 1 = 3+1 = 4'
The fourth will have '((3 + 1) & 3) + 1 = 0+1 = 1'
The fifth will have '((0 + 1) & 3) + 1 = 1+1 = 2' which means the cycle has repeated.
The best way I can think of to replicate this given that you can't do a specific message counter in custom CAN is to use a timer with a max time that won't run out in any drive, the max value of 16,000,000 wouldn't be a bad choice, set the activation to always on and polarity to reset when inactive. Then use a math block setup as below:

Note I haven't tested this setup but the thinking behind it is multiply time by 2 so that you are effectively incrementing every half second (this matches the 2Hz message transmit rate), the '%4' is a modulus operation which should give values of 0, 1, 2, 3 and then adding one will make this value move between 1,2,3 and 4 every half second, then add the math block to byte 1 of the 0x1ae0092c message.
Note: I'm pretty sure the modulus operation isn't working in the current release firmware so you might need to wait for the next firmware release for that bit to work.
 
For the 0x2104136 use the same trick as above for the fixed values (data_0x02104136 = { 0xbb, 0x00, 0x3f, 0xff, 0x06, 0xe0, 0x00, 0x00 }).
and then just put a CAN Aux duty cycle in bytes 7 and 8 (width of 16, Least Significant (LS) byte first and control that CAN Aux duty using a gp pwm output.
 

From what I can tell, based purely on what I've looked into, the pump will "default" to 70% with no CAN input.
The keepalive is a heart beat signal from the ecu, so effectively turning on the PWM of the pump itself, enabling it to be controlled 0-100%. 
If the keep alive signal is not there, it will default to 70%.

Once the keep alive is active, the PWM can be changed. 
It "might" be as simple as giving the keep alive the rpm parameter, so as long as the engine is running, it has that keep alive signal, but then it might not and might be expecting a certain counter.. You just need to know what bytes within that frame need to change. I'm very novice on code and not quite making out the C++ of the Arduino code but should be pretty easy for a coder to say what bytes need changing etc from that code example. 

No, there is no way to control this 8 speed with any Link.

You have a couple of paths to follow if you are keen. I've been doing quite a bit of research into this last couple of years, only with life to fuck me over pullignt he trigger.. Here is what I've compiled. 


-- CanTCU -- https://www.canformance.net/product/cantcu-transmission-controller/  **Recommend if Street Car
Keep the BMW TCU in the mectaronics, you can use CanTCU with a Link and connect via CAN. The CanTCU will take the CAN information from Link like torque, map, tps, rpm etc and the CanTCU will communicate with the BMW TCU in the transmission itself and give it the correct info it needs. 
Pros: 
Less wiring, bugger all tuning
Near OEM drivability
BMW TCU flashable with xHP tunes.
Cons:
Some drag/drift fuctions not possible like clutch or transbrake.

-- HTG -- https://htg-tuning.com/
Standalone TCU, requiring rewiring of the mechatronics and removing the BMW TCU. Link communicates via CAN to the HTG TCU.
Pros:
Flexibility
Highly tuneable
Clutch by Wire
Trans brake possible
Cons:
Needs modification of transmission
Needs to be tuned
Somewhat buggy software
Constant firmware updates and some require retune

-- TurboLamik -- https://turbolamik.eu/ ** Recommend if Racecar
Standalone TCU, requiring rewiring of the mechatronics and removing the BMW TCU, Link communicates via CAN to the TurboLamik
Pros:
Flexibility
Highly tuneable
Clutch by Wire
Trans brake possible
Cons:
Needs modification of transmission
Needs to be tuned

-- Praetor PTC -- https://mectricmse.com/products/praetor-ptc
Standalone TCU, Link communicates via CAN to the Praetor, unreleased so far, due this year. Australian company. Hardware looks good, somewhat overkill really.
Pros:
Flexibility
Highly tuneable
Australian Company
Cons:
Needs modification of transmission
Needs to be tuned
Not released yet.
-- Sound German PCS TCM2800 -- https://www.soundgermanautomotive.com/pcs/
Standalone TCU, only supported with FCA (Fiat Chrysler Automobiles) 8HPs, limited CAN support (GMCAN, J1939).
Pros:
Highly tuneable
Cons:
US based
Limited Support
Small User base

 

Thanks Adam, seems to have helped! Idle settles on target nicely (as much as cams can). 

I have a side question.
Should my Throttle be resting at 5% while at speed under decel (or <5% APS). Idle I understand but no matter what the car is doing, cruising at 100kph or around the street, the TPS always shows minimum of ~5%. I would have thought during a decel event, TPS should be fully closed? 

Hi Everybody,

I'm in need of a little advice, trying to get my cold start and idle sorted.

Engine is a 2JZGTE VVTI with 272 Cams. Starts okish, just takes a few cranks, when it starts, the idle settles to 700-800rpm and is a bit cranky. Until the engine get warmer (over 40c) where it smooths out.

However, I have set the target RPM to 1300 for colder temps, but it never get to that.

I have Closed Loop enabled, I also have IGN control turned on with max advance 20 degrees. After start, the idle goes to 800rpm, and IGN is at 12 degrees, then about a couple of seconds, IGN control kicks in and pushes the IGN to 20 degrees to try and meet target, however all this does is drops the rpm about 100rpm then kinda goes up then down. Starts to smooth out once engine starts warming up.
I've captured a PC log of the cold start and tune file.
Log
https://drive.google.com/file/d/1ULMaLkHUhJiwwCnNxeLiLMFNIS-v-wfo/view?usp=sharing
MAP
https://drive.google.com/file/d/1Ud05ODTCVPjJLFeana5_2_wK-5VSMb9x/view?usp=sharing
I've tried to change the base position table from say 5 to 7.5 but doesn't seem to do anything. What am I missing here? Is it a fueling thing? Too much fuel during startup? Lambda is not enabled until 15 secs or so after and it's very rich. But if it's not rich, the idle is very lumpy.

Those fuel cuts are the overrun fuel cut.
 
Messed around a bit today, turned off CLL and can manage to replicate the issue while the car is stationary and just revving to 2000-2500rpm. Goes very rich @ ~0.7 lambda. Then by about 3000 rpm it comes good again. So played with the VE table and can get it a little leaner in that spot but it's a hole in the map deeper than a diamond mine. Something is certainly up.. I wonder if a injector is to fault here.. 
I will change plugs first and check them. Then get the injectors cleaned.

Certainly some things to try. In regards to vacuum, it's only an issue with me for the first time after startup with minimal brake asset. I'll have a look at those misfires. Change plugs etc. Running pretty low gap at 0.3 as I'm still running the stock coils. R35 coils are on the plate so i can increase gap again at the power end. 
Will have a look at those misfires, what time stamps are they? I do have a ignition retard on gear change.
 
Will give it a go Adam cheers

Trans is controlled by a Compushift Sport TCU.
TC is not locked up at that point. I log the TC lockup as well. 
It seems to be repeatable at that speed. Can certainly feel it, at first I thought it was going lean as that was what I was seeing on the dash, but it's going rich then closed loop catching it, then whatever is causing it to go rich stops and CL has to add fuel again. 
I have pretty low vacuum anyway even at idle due to my cams 272, 9.9mm lift. To a point I'm putting a vacuum pump on for my brakes to help.
Just seems puzzling that during that low speed cruise it happens pretty consistently. 
If it is a driveline issue and something in the trans binding up causing a little extra load I should see that extra load in the logs?

I had this a while ago, contacted Haltech and they corrected the map. It should have been rolled up in the newest ICC firmware.
Basically they changed the ecuchannels.json to correctly show the Generic Dash Gear Position numeration. 
  {
    "unitGroup": 12,
    "unitSelection": 0,
    "ID": 40,
    "Name": "Gear",
    "DefaultLabel": "Gear",
    "DefaultUnit": null,
    "MinimumValue": 0.0,
    "MaximumValue": 15.0,
    "Shortname": "",
    "Function": "Gear Position",
    "Unit": "Gear",
    "PullupPosition": "",
    "PullupStatus": 0,
    "ThresholdValue": 2.5,
    "SensorIndex": -1,
    "MuxValue": -1,
    "MuxByteOrder": "",
    "EcuCanID": 1404,
    "EcuByteOrder": "3-2",
    "OBDIICanId": 0,
    "OBDIIByteOrder": null
  },
 

What exactly stopped working with the Anti Lag in 6.22? is it not activating? can you describe or show what issues you are having with it?
 
can you attach a copy of your tune so I can see exactly what you are doing with your rpm limits please.
Have fixed this, fix will be in the next release.

Outside of idle conditions the throttle will go to whatever is in your E-throttle target table + idle base pos + dashpot offset.  If it went to zero during decel as soon as the RPM got low enough for idle control to engage there would be no air in the plenum and throttle is closed so the engine would stall as soon as you press the clutch in. 
You would normally have 0 right across the top row in your E-throttle target table, I noticed you have 3% above 2000rpm so that will hold it open even more.  

Thanks Guys, I tested the 3d idle out just now and certainly an improvement, not perfect as it oscillates once or twice before settling so will keep at it with your suggestions too Adam!

Thankyou

Cheers mate, that makes sense now.

Will give the idle base position table a play.

Thanks for your help!

If yours is reading 0psi at atmospheric pressure then it is a gauge pressure sensor which is the correct type for fuel and and does not need any math.
The original posters sensor is an absolute pressure sensor so would read something like 14.5psi with sensor removed from the rail.

Yes, I myself ended up with quite a bit left over. I've done mine to a 2JZGTE VVTI out of a JZS161 Aristo.

On the Link/Altezza header, you can free up the following as they are not used on a JZ and if you are using a Wideband like a Link CAN WB.
Header Pin - Function - Description - Link IO
A5 - ACIS - Variable Intake VSV - AUX5
A13 - IGT4 - Ignitor 4 - IGN4         ## Freed up for use later, rewired to connector
A28 - OX1A - B1S1 O2 - ANVOLT1  ## I'm using this for Cruise Control switch
B3 - HT1A - B1S1 O2 Heater - INJ5  ## repinned to existing in engine loom.
B6 - OCV2- - VVTI Exhaust - AUX2
B8 - LCK1 - AC Clutch Lock - DI8
B14 - GEX - CAM Pos Exhaust - DI2
B21 - VG - MAF Input - ANVOLT4
D10 - OX2A - B1S2 O2 - ANVOLT7
D11 - HT2A - B1S2 O2 Heater - INJ6  ## repinned to existing in engine loom.
E3 - FC - Fuel Relay - IGN5             ## I freed this up and rewired the FC Relay to AUX2 (B6) so I can use IGN5 for seq ignition..
E4 - FAN - Fan Relay - IGN6            ## As above, AUX5 (A5) now runs FAN relay so can use IGN6 for seq ignition.
E6 - PRE - AC Pressure Switch - DI7
I've wired in a Deustch connector for these freed up inputs 
IGN4
IGN5
IGN6
DI2    ## Connected to FLEX sensor
DI7
DI8
ANVOLT7
ANVOLT4
SGND - Sensor Ground if needed

Also have the Link Expansion Connector (EXP) which gives the following
GND
5V
DI10
DI9
ANVOLT9
ANVOLT8  ## Fuel Pressure
ANTEMP4
AUX8   ## Boost Control
 
Hope this helps you get started.

Sorry for the late replay @Adamw, yes getting that data on CF00203. Input shaft speed is listed at HGM CAN Byte 6 bit 0. Which to me looks like Link Start Pos 47, LS, 16bit length.
 
I've been going through the CAN messaging and doing a bit of a spreadsheet 
https://docs.google.com/spreadsheets/d/1yqeI4xfFQhsc9jpNo8mTtEORLC_8jf8TRkfLRoIFZko/edit#gid=650908763
I'm really not sure why HGM decided to use the full HEX for frames egress but you need to drop the "C" for frames ingress to the CS3. For example in order to give the CS3; RPM and TPS you need to set a transmit channel to DEC ID 15729664 (from their docs) which is HEX "F00400" but egress frames from the CS3, you NEED to add "C" to the HEX ID to get CF00203 and thus DEC ID 217055747.

I will test the shaft speeds out when I get my car back as wife stole it today. Can't do much with it on my G4+ but I have my G4X setting on my desk waiting and already put it in the config.

G4X is working properly now!  
changed the firing order to 300ZX and it started right up! 
Sorry to have wasted everyone so much time, and thanks so much for you and your team's help. I appreciate it. 
I had no idea the shop who built the car used a 300Z ECU until now, i don't know what the reason was ....
I changed the pins back to factory on the ECU connector after we confirmed G4X is working. 

Here is my data, from Golebys website, not sure if it's dead on but it is something.

 
 

Personally I think Adam and the team does a wonderful job of helping people get their shittas running. 

In fact I've seen a number of times Adam (sorry to single you out) has gone above and beyond to help.

You have to remember that these guys (non gender version) make and support the ECU. Getting the engine and systems working with the ECU comes down to who is putting in the numbers into the ECU. It's not on Link to get it running and tuned for you. 

Some people keep bring up this whole thing about them feel abandoned on G4+. It's EOL, new version now with better features and faster CPU. It would be like complaining to *insert brand* management that a 10 year old laptop can't do Machine Learning or run multiple VMs. I was a little bit on the same thought when the G4X was released. But then you look at the features of the G4X and realise it's better in so many ways. Nothing wrong with the G4+, the X is just more. 

I'm sorry you have had such a hard time with it. I spent 10 years in L1, L2 and L3 support before consulting for a multinational enterprise computer company, support is hard when you don't have the whole picture. You feel guilty about asking for more information, but what people don't understand is the more info you get, the better idea that support person has of fixing the issue first time for you. 

 

Personally I think Adam and the team does a wonderful job of helping people get their shittas running. 

In fact I've seen a number of times Adam (sorry to single you out) has gone above and beyond to help.

You have to remember that these guys (non gender version) make and support the ECU. Getting the engine and systems working with the ECU comes down to who is putting in the numbers into the ECU. It's not on Link to get it running and tuned for you. 

Some people keep bring up this whole thing about them feel abandoned on G4+. It's EOL, new version now with better features and faster CPU. It would be like complaining to *insert brand* management that a 10 year old laptop can't do Machine Learning or run multiple VMs. I was a little bit on the same thought when the G4X was released. But then you look at the features of the G4X and realise it's better in so many ways. Nothing wrong with the G4+, the X is just more. 

I'm sorry you have had such a hard time with it. I spent 10 years in L1, L2 and L3 support before consulting for a multinational enterprise computer company, support is hard when you don't have the whole picture. You feel guilty about asking for more information, but what people don't understand is the more info you get, the better idea that support person has of fixing the issue first time for you. 

 

Hey Guys,

I've just upgraded from a G4+ to a G4X TALTX and I'm just installing G4X PCLink EN 6.19.65. However, running into an issue were it fails to start.

Errors in a dialog box saying FTD2XX.DLL is missing. This is in a Windows 10 20H2 Virtual Machine on VMware Workstation 16.1

Uninstalling and installing 6.18.44 works fine.
 

Update, HGM fixed the firmware and now have gear change request. 

Thanks Adam on the torque reduction strategy, works a treat. Just need to fine tune it. I've reduced the IGN trim retard a little as I don't have 20 degrees of timing in some spots and it's getting retarded too much and seems to spike boost.