K20A2 Trigger Scope Check Please

[atlex]

Hi Folks,

I've switched from Kpro to G4X on my K-swapped ZZW30 MR2. It's a stock A2 motor.

I'm working on the tune at the moment and I want to be sure my triggers look good.

Is the timing OK or will I need to mess with offsets ?

It's a COPS car obviously so I may need to hack something up for running a timing gun.

Alex

TriggerScopeLog.llgx

[dx4picco]

Signal looks very noisy to me? 

What is the wiring based on? 

[atlex]

Wiring is mostly stock K20A2 from the EP3 Type R - it's a CTR engine loom with some links into the original toyota harness where necessary.

I could attend to/clean up some of the earths and take another gander at the scope.

The car runs. Does anyone have an example for me to compare with ?

[Adamw]

I dont think the noise is anything to worry about at this stage, I have seen the G4X scope exaggerate noise on hall effect triggers before, something like an artifact between resolution, sample rate, etc. 

But one problem I do see is the low level voltage is only coming down to about 0.9V, it should be close to 0.0V.  This would suggest a ground offset somewhere between the crank sensor ground and ecu ground.  The switching threshold is 1.0V so you are only a small margin away from having a problem. 

 

Also there is something fishy with your MAP sensor signal.

 

[atlex]

9 hours ago, Adamw said:

But one problem I do see is the low level voltage is only coming down to about 0.9V, it should be close to 0.0V.  This would suggest a ground offset somewhere between the crank sensor ground and ecu ground.  The switching threshold is 1.0V so you are only a small margin away from having a problem. 

Thank you. If you have any pointers to tips on identifying/rectifying this I'd appreciate it. I've got a multimeter, guessing to look for voltage between the signal ground and the sensor ground plugs ?

I'll try and address this before looking at the map signal. I could also switch to the onboard map sensor but I'd rather not.

[atlex]

That trigger graph was while the engine was idling and iacv running so we'd _maybe_ expect to see some change in the MAP signal, I think, although I may be reading it wrong.

 

Battery negative to chassis ground _appear_ to be good - I see 0-1ohm resistance between far chassis points and the battery negative and when there's current, i.e. the engine is running, this goes up to around 10 ohms. [This seemed better than another car I have]

I've been reviewing some guides on this - https://dsportmag.com/the-tech/education/how-to-avoid-and-fix-automotive-electrical-issues/ and the wiring diagrams for the EP3. I'm going to have a go with the testing of these. Will see how disconnecting sensors may clean up the trigger scope.

I see the the grounds for the Inlet, Exhaust and Crank position sensor are a bit odd in that they appear to be to the A23 LG2 A24 LG1 logic grounds which are also hooked up to a Chassis ground (I don't know why?) - will start testing the sensors and wiring for these.

 

[Adamw]

It is not uncommon in OEM stuff to ground hall effect sensors to chassis or engine.  In that case you want the main ECU grounds going to the same point.  

[atlex]

I've done the quickest thing first since it's hot out there. I disabled fueling and started pulling connectors and taking scopes.

inlet cam disconnected was normal bad as before, 0.9v ground ref on trigger 2 ex cam

exhaust cam sensor disconnected was a mostly 3.89v on trigger 2 ex cam (!) with a few very brief drops to 3.66v

both inlet and exhaust cam sensor produced random spikey noise between 3.66v and 3.89v on trigger 2 ex cam (!)

in both cases crank sensor signal trigger 1 was mostly fine but I will try disconnecting that just to see what that produces too.

would expect to see a flat line 0v with the exhaust cam sensor disconnected and as for the noise with both disconnected.. I want to try disconnect the crank sensor too.

Onward to the wiring....

Both Cam Sensors Disconnected TriggerScopeLog.llgx Exhaust Cam Sensor Disconnected TriggerScopeLog.llgx Inlet Cam Sensor Disconnected TriggerScopeLog.llgx

[Adamw]

With sensors disconnected the trigger inputs should sit at about 3.6-3.8V as they have pull-up resistors pulling them up.  

If you get a paperclip or similar and with the sensors unplugged, short the ground pin to the signal pin then you should see close to 0V in the scope.  If you still see 0.9V, then that would mean the ground wire at the sensor plug is 0.9V higher than the sensor ground pin at the ecu.  

[atlex]

Tested with the paperclip shorting sig to gnd - 0.9v confirmed in the trigger scope.

 

[dx4picco]

So now, find where you sensor ground is and bring it to the ecu ground connection

[atlex]

So I've found the point at which LG1/2 and PG1/2 go and it's the same ground connector, the one called G101.

LG1/2 are brown/yellow and PG1/2 are pure black and those wires are all here.

I'm going to try clean it up a bit more but I'm stumped a little now. I will check if that's dirt on the connector or it could be that the ground was fried. If you look at the shade of the brown/yellow cable as it comes to the metal it's darker than it should be frankly.

I will head to a local marine chandler and buy some suitable wire to make grounding cables and redo what there is.

Time to redo the battery to chassis grounds also.

[atlex]

Cleaned up the battery-to-chassis and added more chassis-to-block and chassis-to-head grounds.

Where it looked dark on the 'G101' picture above was just dirt and that cleaned up. This indicates the LG and PG are to the same location.

Also tried pulling the crank sensor connector.

No change to the ground voltage offset.

I ran another test with the map sensor plug disconnected and saw 0.07 volts signal in the trigger scope for  map signal voltage.

Riddle me that!

I've also ordered a spare used k20a2 loom since I need a way to take that out of the equation.

I could add one more ground, but I'll need to get the car up on a lift to do that. I doubt this will help.

[Adamw]

1 hour ago, atlex said:

I ran another test with the map sensor plug disconnected and saw 0.07 volts signal in the trigger scope for  map signal voltage.

You would need to short the MAP signal pin to the MAP ground pin to confirm if the MAP ground was different.  With just disconnecting the signal wire it would be close to ECU ground as it has a weak internal pulldown.  

The quickest test on your crank sensor would be to pull the ground wire out of the connector and run a temporary one direct to the ecu.  

[atlex]

So thank you for putting up with me so far.

I haven't hit the map sensor with the paperclip check again or done the crank sensor direct ground yet.

I've instead removed the grounding between LG 1/2 and the Motor. Per the G101 photo above it was the same place as PG1/2, I just snipped off the brown/yellow wires at a suitable point where I could reattach them with crimp connectors should I want them back.

I then proved (in as much as I can) with a multimeter that there's no other source of continuity to the LG1/LG2 wiring other than the point at which they connect to the ECU. Alas this isn't strictly true of continuities to ground itself, there's continuity to ground on the brake switch input and the vtec oil pressure switch inputs.

This made the ECU the ground for these 'logic' sensors - the two cam sensors and the crank sensor. And this made no difference to the 0.91v ground voltage offset in the trigger scope. And the signal looks no different to what it was before. Even with the sensors disconnected we saw the same offset. The paperclip grounding to signal proved that. 

I'm left scratching my head as to the source, as I've literally gone against the advice and should have made it worse.

Tomorrow I'll try grounding the PG1/2 to another point but there's only 0-1 ohms between it and the battery negative.

Is the problem even real ? The new fresh loom hasn't arrived yet and I will probably get a friends garage to take on that work.

I'm going to try independently verifying the offset voltage with an actual oscilloscope. Seems like the right thing to do.

[DerekAE86]

What type of G4X is this? A wire-in model?

I would be aiming to make things better than OEM rather than working around what the OEM might have done
(in relation to sharing Sensor Grounds and Chassis Grounds)

Make sure the Sensor GNDs don't have continuity to the Chassis/Engine at all. Make sure the ECU Power Ground is good to both the Chassis and Engine.

[atlex]

12 hours ago, DerekAE86 said:

What type of G4X is this? A wire-in model?

I would be aiming to make things better than OEM rather than working around what the OEM might have done
(in relation to sharing Sensor Grounds and Chassis Grounds)

Make sure the Sensor GNDs don't have continuity to the Chassis/Engine at all. Make sure the ECU Power Ground is good to both the Chassis and Engine.

This is a plug in. I'm using the stock-ish K20A2 loom.

This loom has separate Power, Logic and Sensor grounds..

map sensor ground SG1 A11 direct
tps sensor / IAT sensor / ECT sensor ground SG2 A10 shared
fuel tank pressure sensor (unused and unwired) ground SG3 E4

CMP A (sensor VVT inlet) - signal A25
CKP (crank position) - signal A7
CMP B (sensor b for tdc aka exhaust) - signal A26
all three share the same LG1/LG2 ground 

(which would also be to the G101 location  along with PG1/2 but even if the only ground path (I cutLG1/2 off G101), there remains 0.91v on the signal.

ECU Power Grounds are to PG1/PG2 which are A4/A5 respectively and these go to the head of the motor (G101) but I've tried different grounding points, made no difference.

I'm thinking to disconnect all the sensors using SG1/2 grounds and see what happens.

My oscilloscope is nearly ready.

[atlex]

So the update so far:

I have a working oscilloscope!

when I ran it I saw a relative offset ground voltage (0.3v) and signal voltage (1.5v) on both the g4x and the kpro having backprobed the inlet plug while running the motor.

interestingly on the kpro I saw offset voltage via multimeter between the logic/signal ground for the cam sensors and the battery/chassis negative - 0.8v - and if I pulled the fm2/ign2 fuses this was halved to 0.4v. but on the g4x this doesn't appear and I get 0v readings.

but the oscilloscope reading was the same.

if I attach my multimeter to the inlet cam sensor (only used for vvt control - it's on DI1) between that sensor plugs ground and the battery ground I get 0v even when the engine is running.

I also went back to test the map sensor since that was noted before. 0.2v signal if the plug is disconnected, 0v if signal and signal ground are shorted - grabbed a trigger scope of this and it made no difference to the trigger signal, just saw a 0v map signal...

I've a few more tests I want to run next week when I get access to a car lift.

[atlex]

did a quick check of the exhaust cam sensor plug (sensor disconnected)

 

I see 4.1v between the signal wire and the chassis ground/logic ground - seems wrong to me?!

I've never seen the 0.9v between the logic ground and chassis ground with my multimeter... even when cranking.

[Adamw]

On 8/11/2023 at 10:15 AM, Adamw said:

The quickest test on your crank sensor would be to pull the ground wire out of the connector and run a temporary one direct to the ecu.  

Did you do this yet?

[atlex]

On 8/11/2023 at 12:15 AM, Adamw said:

The quickest test on your crank sensor would be to pull the ground wire out of the connector and run a temporary one direct to the ecu.  

I did this to the exhaust cam sensor (trigger2) since it was easier to reach and suffering the same ground offset.

I disconnected the logic grounds in the original loom and ran a new pin to either LG1 or LG2 (the other one left disconnected)

 

Original signal and power to the sensors remained.

Turned off ignition and injection... results..

0.91v

 

trigger2 retains the same old 0.91v ground offset with the new dedicated ground. is this an ECU at fault ?

can't explain the trigger1 result though, if there's no earth wired up to it... is it earthing via the signal, or something else ?

New-Dedicated-Direct-Trigger2-LogicGround.llgx

[atlex]

I'm working to find another car to test this ecu in.

On 8/18/2023 at 6:19 PM, atlex said:

when I ran it I saw a relative offset ground voltage (0.3v) and signal voltage (1.5v) on both the g4x and the kpro having backprobed the inlet plug while running the motor.

The oscilloscope was uncalibrated (derp) so I will repeat that and get screenshots of the signal.

[Adamw]

Can you short pin A5 to A7 with a paperclip or piece of wire and do a scope capture like that.  

[atlex]

16 hours ago, Adamw said:

Can you short pin A5 to A7 with a paperclip or piece of wire and do a scope capture like that.  

I put some paperclip in precisely as described. Sat at 0.91v without the engine turning, never went below 0.91v when it was turning.

Ignition and fueling were disabled for these tests.

The screenshot below is of me turning the motor over without the LG being connected for the CKP, but with the a5a7 short in.

 

TriggerScopeLog-ckpgrounddisconnected-a5a7shorted-startermotorturning.llgx TriggerScopeLog-groundsconnected-a5a7shorted-notstarted.llgx TriggerScopeLog-groundsconnected-startermotorturning.llgx

[atlex]

My oscilloscope still isn't calibrated as I don't have a regulated power supply (...yet.. it's on order!) but I have managed to get a decent signal out of the sensors.

This was with ignition and fuel - the motor was just idling.

Here's the run I got out out of the ex cam sensor.

It's showing an offset of -180mv at the bottom (calibration needed..)

p2p is probably more important here and that's close to 3.8v

in the calibration mode it shows a -0.23v with the probe shorted to its own ground - so if you wish, you could read the -180mv as being a 50mv ground voltage signal. a far cry from 0.91v :/