Scott

Update;
 
The car went in to be tuned at Prestige Tuning and Motorsport. It made 110kw at 14psi. The clutch started to slip and as predicted, the injectors maxed out but overall, a great result. Thanks everyone for your help with my project.
 
Hamish

After doing a bit of research recently into running a tablet as the primary dash for my car here's some thoughts...
 
Hardware:
For the older Windows tablets using Micro usb you should be wary of any claims of being able to charge and have a device connected, a lot of descriptions of the adapter cables are wrong as it totally depends on the tablet, so make sure you do your research first, if in doubt it likely doesn't support both at same time.
However I believe the new tablets which support USB C may be better in this regard, again do your research first.
Performance of PcLink on lower speced windows tablets (1.3ghz quad cpu, 1gig ram etc) is actually very good, PcLink will be very slow as the software initially connects to the ecu, and the display wont respond for a good 10 seconds, but once it syncs up it works perfectly. 
Here's my 8 Inch acer tablet with Windows 10 and PcLink with different layouts
Because I could not charge and display data at same time I just ran it on battery, it lasted for a good 5-6 hours which was impressive.
To me there's no point in having a tablet that can't charge and display so I replaced the above with a Lenovo ThinkPad Tablet 2 (2 usb ports, 10 inch) with Windows 10 and PcLink
This charges and runs at the same time which is perfect.
 
Setup:
Setting up PcLink for the tablets is a bit of a pain as PcLink doesn't support touch very well.
The easiest way is to have another pc with PcLink running and do the setup on that then transfer the layout files across.
Your best to set the resolution of the display on your laptop to match the tablet as otherwise positioning of elements is very hit and miss (especially with high res displays) the ThinkPad I brought has a pen which made resizing and moving elements actually doable on the tablet.
Note there is an issue with PcLink and some tablets (My acer one doesn't have this issue) where they throw a 
"Access violation at address 00000000 in module 'PCLink.exe'. Read pf address 0000000." same as this thread
 http://forums.linkecu.com/index.php?/topic/6144-access-violation-at-address-00000000-in-module-pclinkexe-read-pf-address-0000000/
 
 It appears to be to do with the 3d chart display and it's very frustrating as PcLink is unrunnable as it triggers the error when ever you open PcLink and when it occurs it stops PcLink from working correctly. 
 
 You can work around this by running PcLink on a normal pc, getting rid of the default layout with a simpler one then copying the entire folder "C:\Link G4\PCLink G4+" directly to your tablet (this means the software wont try and load the default link layouts which cause the crash). Once you have done this then you can just import layout files as normal (just make sure you use non 3d gauges). Hopefully they fix this at some point (even if it's just disabling 3d charts rather then crashing).
As an aside PcLink is inconsistent with how it handles limits on displays, i.e analogue vs digital are configured in different sections of the app, Once you realise this then configuring it's not to hard, however I still haven't found a way yet to get rid of the decimal point on the km/h. 
PcLink has options for automatically going full screen, and auto connecting which are easily turned on through the menu.
Then in windows I have set it as a start up app so it opens automatically, and have changed the user account to go straight to desktop rather the sign in screen (this is a bit more work involving regedit, guides can be found online for both).
 
Realworld Performance:
You can see me using my 8 inch tablet in a race here (fully on battery).
It performed flawlessly, however as you can see in the video glare is an issue with these tablets (visibility was ok from the drivers seat but could be better), so if using as your main dash then a hood/cover is something you want to look into fitting.

Finally after far to long we have some recordings off the G4+ KnockBlock
These recordings are off our well abused 1UZFE VVT test engine.
 
 
Engine background noise.m4a
Mild Knock with bad tune.m4a
Severe Knock with bad tune.m4a

Hi Bram,
To answer your questions:
Q1. Yes, here is how we do it in our WRX V10 (2.5L e-throttle) plug-in ECU:

Aux Output 6 (FP Speed) goes to pin B27. Pin D30 is not used by our plug-in ECU.
Q2. Left hand is as your are sitting in the car. I'm not sure why the engineers chose that cam for the Trig 2, the pattern on both cams look to be the same, but offset. Maybe it is something to do with how the cam patterns are offset in comparison to the crank pattern.
Q3. 
Pin C17 is An Volt 10 which is Accelerator Pedal Sensor - Main Pin C28 is An Volt 9 which is Accelerator Pedal Sensor - Sub Pin C15 and/or C16 should be +5V to the sensor Pin C34 and/or C35 should be sensor ground to the sensor Q4, This may be possible, but I don't know anyone who has done it before.
Q5. Yes, this should be possible, you will need to use the ECU's Hold Power function to keep the ECU alive after the key gets shut off. See PCLink Help > Wiring Information > Power and Ground Wiring > ECU Hold Power Wiring. Scroll towards the bottom of the page and you will see two diagrams for wiring the Thunder for ECU hold power. You will want the ECU hold power timer to be set to longer than you want to run the oil scavenger pump for. You will also need to think about what condition you want to activate the scavenger pump under, maybe engine speed = 0 and ignition switch = off. You will have these conditions activate one of the ECUs timers and then have the timer activate an aux output that controls the pump.
Scott

The cable to a knock sensor should be standard automotive cable that is shielded. If you have a 2 pin knock sensor the shielded cable should contain two wires.
The narrowband sensor fitted to the engine from the factory is selected to match the expected frequency that knock will occur at in the engine. Engine modifications can shift the frequency that knock will occur at, especially modifications that change the bore of the cylinder. For this reason it is often better to use a wideband (donut) type knock sensor, and then use the ECUs selectable frequency filters.
In terms of setting up knock control, apart from selecting the frequency filter the process is the same for a narrowband or wideband type sensor. So I wouldn't say one is more user friendly than the other.
Here is some guidance on knock sensor location from the G4 KnockBlock manual:

Scott

We will have pre-release testing available, but are still working on the next PCLink and firmware release, so will still be a while yet. Send us as email on [email protected] if you are interested in helping with the pre-release testing.
Scott.

Hi,
Sorry, I was using a pre-release PCLink to test it before. It works in this as the engineers have fixed it. When this is released it will be available for you to use.
Scott.

Just found out bout the quake

Hi Nigel,
To be honest I don't think the engineering team will modify the DisplayLink firmware as it is now some years old. 
Another solution (that I've already done most the work for) is to make the 'Generic Dash' stream file adjustable by users. A month or so ago I did this for one of our dealers, but have not heard back yet if I created the file correctly.
Here is an instructional video I made. You can follow this, and go to frame 7 and change the divider for Lambda 1 and Lambda 2 to '10'.
https://vdoubleyou-scott.tinytake.com/sf/MTA0MzIzNV80MjM4MTcw

Attached is the file you will need.
I would appreciate it if you can let me know if this works. If it doesn't I can try and make some changes.
Scott
Generic Dash stream file.lcs

PCLink Help gives an example of using a GP Limit table to control engine RPM based upon oil pressure. Have a read through of this and you'll see how something similar could be setup for fuel pressure.
PCLink Help > G4+ ECU Tuning Functions > Engine Protection > GP RPM Limit   The example is towards the bottom of the page.
Scott

Ok, here (quickly) are the steps you would normally take to set up the Knock detection system:
Make sure the ignition system is completely tuned. Set the knock gain to a value of 1. Select the correct frequency filter (we have covered this) Monitor the individual cylinder knock levels, adjust the gain until the max reading you get in the levels is approx 500. This is normally somewhere around high rpm and high engine load. Monitor the individual cylinder knock levels and adjust each cylinders gain until the cylinders all read evenly. Fill in each cell of the target table. Set the values to be 10% higher than the values you are seeing on the individual cylinder knock levels. Logging could be a helpful way to do this. Now whenever the knock level of a cylinder exceeds the current value from the knock target table, the ECU will try to retard the ignition timing for that cylinder to reduce the noise (knock) level. You can decide at what rate the ECU will retard the timing, and what the maximum level of retard can be. Benefits to having a quick retard and large maximum retard are engine protection. Benefits to having a slower retard and small maximum retard are the engine does not drop so much power.
Lets assume you have ended up with a setup like this:

At 50% throttle and 1500 RPM you are normally getting cylinder knock levels of 255, so you set the cell in the Knock Target table to 280. One day you lend the car to someone and they put a lower octane fuel into the tank. This causes the engine to knock lightly and produce cylinder knock levels of 320. The ECU sees this and retards the ignition timing until all affected cylinders (probably all of them) have a noise level less than 280 or until the Ign Retard Limit is reached.
Regardless of the cause of the higher cylinder knock levels the ECU will retard the timing in an attempt to reduce the levels to be below the level specified in the Knock Target Table.
One good way to view this is by logging on the Timeplot. Display each cylinders level and then then Knock target (threshold). This will allow you to easily see when a cylinder jumps up over the allowable level.
Scott

Hi Rob,
Have a look at the parameters for 'Inj Actual PW' and 'Inj Effective PW'. These are found in the left hand column of the fuel tab of the Runtime Values Window (F12 key to open).
The actual pulsewidth is the effective pulse-width plus the injector deadtime.
Have a look and see what value the pulse-widths are when the AFR is rich. Then monitor them as you reduce the master fuel number. You will not be able to go lower than the setting you have for Min Effective PW.
Scott

I havent studied your log in great detail but it looks like your main issue is the post start enrichment table is filled with -100% so that is taking all your fuel out after cranking enrichment ends.  Set post start enrichment to off for now.
free upload

Hi Hamish,
When a Link ECU sends out an ignition pulse it is a +5V pulse. Just about all ignition modules and coils with built in ignition modules use this type of signal.
I'm confident the 'Spark Edge' setting should be set to falling for your coils. What dwell values to use is more of a mystery, perhaps someone here has used these coils before and can advise.
Scott

Hi sat down tonight and re-wired with 2 ignitors, fired straight up, thanks a million.

You would use a timer to finish the gear shift control.
PCLink help has an example of this type of setup. See PCLink Help > G4+ ECU Tuning Functions > Motorsport > Gear Shift Control > Gear Shift Control Examples
Scott

the M50 does not have the same pinout , unfortunately 540 DBW come with 2 types of VDO module both have different pinout , the best practice is to open the throttle black cover and you can easily trace the pinout wiring , as you will have 2 pins for the motor and you can easily see the traces for the 2 TPS  with shared 5V and 0V , and dont worry if you swap the 5V over the 0V as you will eventually calibrate the throttle in the software , as for the motor to determine the 12V and 0V orentiation you will have to bench test them to see which way open the throttle as the wrong way will close it 
 
Best regards

Hi Vonny,
I've heard back from the engineers about this. The An Volt channels 13 to 16 are slower responding than the lower numbered AN Volt channels, so fast signals like TPS are not allowed. I will see if we can get this info added to the PCLink Help file.
Scott

Hi Niv,
I don't have a base-map. The PCLink help file contains some info for this engine, have a look at: 
PCLink Help > Engine Specific Information > Honda > Honda K20 PCLink Help > Wiring Information > Output Wiring > E-throttle > Specific E-Throttle Applications > Honda K20 - 60mm. Scott

It looks like they have most settings already to suitable choices for burning fuel in the exhaust system. You could try adjusting the values in the Launch Control Fuel tables to see if you can achieve better results. Be careful when doing this, excessive heat and pressure can cause damage to engine or exhaust system components. Make small changes and keep an eye on the system. Often you will be altering the balance between building boost/pops/bangs/flames and how long the exhaust/turbo/engine will last.
Scott.

The stock tacho on that engine is run from the factory computer not the ignition system.
So as long as the Atom is running the tacho you should be fine.
There is no need for an inductive clamp.  If the tacho doesn't work for some reason, at most you will need to purchase an MSD tach adaptor which will still be run from the Link Atom Aux out.
 
 

Looking at the schematic for the Link WRX V5-6 ECU the Fuel pump relay output from the ECU can be on pin B134-1 or pin B134-21. Both the pins are joined internally in the ECU. The ECU will apply a ground on this output when the fuel pump is required, the ground will then activate the fuel pump relay. I'm pretty confident the fuel pump relay will be supplying +12v to the pump.
Scott.

I dont know about your timing numbers but the 10 degree in the calibration menu is NOT added to timing. That is just what timing the ecu is locked at while you adjust trigger offset.
Just lock your timing with the calibration menu and check your timing/offset with an ignition lamp and see if it matches. Then you have your question/doubt answered

Just to update- with the new MPX5700AP in place, the MAP readings are now correct.
Looking back over logs I can see that the original sensor had drifted to +10 a couple of months earlier, and had reached +16 by the time I replaced it last week.
Thanks once again.

No they don't work, they all either charge or do data, but never at same time. Any that say otherwise are lying.
When you plug the ecu in the tablet usb port is running in host OTG host mode, so can supply power to the device, but you need power going the other way which requires getting the tablet to switch modes after plugging the device in, there's only one device that does that and it only works for certain brands of tablets https://www.amazon.com/dp/B00OVDE0GC/ref=cm_sw_su_dp mine is an acer so not on the supported list, it's also not something i would want to put into a car.
Dell also had a cable, but it had limited support for some of their tablets and can't be brought any more.
So realistically its in the far to hard basket and i would recommend anyone going down this route avoid single usb port tablets.
 
However there may be another way to use a similar tablet with a Kiwi3 via bluetooth, Just have to find some decent software for the gauges...