P38a (re)build

[Escape]

Some more progress, but not as much as we hoped. We double checked the fuel pump, plugged some airleaks (crank ventilation for example), adjusted the throttle slightly open and after a few tries she finally came alive. She even kept running all by herself.

Unfortunately, the moment you touch the throttle, she dies. It feels like fueling rather than ignition. We played around with the cold start setting and acceleration enrichment, made sure it was open loop to take the O2 sensor out of it, but no avail. And we compared live data with @elbekko's running P38. Timing and injector pulse width are very similar, so the base settings seem to be right.

The mystery continues, but at least we're moving in the right direction. 

[Blanco]

This is where working by yourself gets tough, having Elbekko and the knowledge that's on here is such a boon, fingers are crossed on your behalf in this house, and I am sure many others. 🤞

[Bowie69]

I wouldn't be bothering with AE yet, richen up the fuel table a bit more at and around idle and that will help with the stumble. If you have a wideband then having that running will help to get a better idea of things. 

Check your AE setting, I was once setting it as 15% or something, only to find that is 85% cut

Are you running in a cam? All these starts will be no good for it, I'd get it up to 2K any way you can and get it broken in. 

[Escape]

I'll let @elbekko give the details about the fiddling we did, as he's (obviously) the one behind the keyboard, I'm just the greasemonkey turning the key. 😉

Yes, I do need to run in the cam and I know this is not the way to do it. But it's only been a couple of starts of a few seconds each, so not too bad hopefully. I did try to get it to 2000rpm (though annoyingly the tach isn't working yet, so need to look at the display) and that's how we discovered she cuts out when touching the throttle. Best we got was just over 1000rpm and that was with terrible hunting. 😞 Turned off immediately of course.

[elbekko]

8 hours ago, Bowie69 said:

Are you running in a cam? All these starts will be no good for it, I'd get it up to 2K any way you can and get it broken in. 

We're very well aware, and that's been our goal from day 1. The issue is it dies the moment you look at the throttle (as in, 4% throttle and it stalls). Overall I don't think it's ran for more than a minute combined, so it should still be ok.

We've now disabled the AE (I think) and the lambda correction (I think). We've moved the req fuel up and down with little effect (which should be the same effect as moving the whole map, right?).
The timing of #1 has been re-verified (ballpark, as we don't have exact marks, but should be close enough).

None of this is made easier with TunerStudio being quite possibly the worst piece of software I've laid eyes on since SAP, but I digress.

I've attached a log where the problem is very visible (feathering the throttle as best I could), as well as some of the MSQs we've been using. The "worse" is when we disabled as much as possible, and then it started hunting before dying. Also attached that log now.

Not sure where to go from here, part of me wants to somehow bench test this thing and re-verify that spark is happening correctly and such. A wideband might help, but with only a few seconds of idling I doubt it'll tell us much.

2024-03-03_18.24.54 playing with accel enrichment.msq 2024-03-03_20.34.56 worse.msq CurrentTune.msq 2024-03-03_20.15.22 dead on throttle.mlg 2024-03-03_20.29.41 hunting.mlg

[Ed Poore]

1 hour ago, elbekko said:

None of this is made easier with TunerStudio being quite possibly the worst piece of software I've laid eyes on since SAP, but I digress.

Having just opened TunerStudio I know that's not possible .

Before I left my previous job they brought in SAP - for almost a year and a half my boss had no clue how much budget any of his projects had. We used to have this beautifully simple internal web page called Project Browser where you could go, enter the project code and it had three tabs. Summary, Hours and Direct Expenses. On summary it said how much money the project was allocated, how much had been spent and how much was left. The other tabs broke down exactly what you expected. It was updated live as people charged to it. SAP - well after 3 years of being there when it was introduced I never successfully ran a query to know how much a project had spent.

Whilst I was farm sitting for my parents I had a visit from an electronics distributor rep - since he'd driven all the way down from Reading area gave him a cuppa and we sat and talked even though I wasn't involved heavily with the business at the time. They'd had SAP introduced a year before - one of the outcomes of that was that they were unable to ship anything for a month. Superb for a electronics component distributor...

[elbekko]

3 minutes ago, Ed Poore said:

Having just opened TunerStudio I know that's not possible .

Before I left my previous job they brought in SAP - for almost a year and a half my boss had no clue how much budget any of his projects had. We used to have this beautifully simple internal web page called Project Browser where you could go, enter the project code and it had three tabs. Summary, Hours and Direct Expenses. On summary it said how much money the project was allocated, how much had been spent and how much was left. The other tabs broke down exactly what you expected. It was updated live as people charged to it. SAP - well after 3 years of being there when it was introduced I never successfully ran a query to know how much a project had spent.

Whilst I was farm sitting for my parents I had a visit from an electronics distributor rep - since he'd driven all the way down from Reading area gave him a cuppa and we sat and talked even though I wasn't involved heavily with the business at the time. They'd had SAP introduced a year before - one of the outcomes of that was that they were unable to ship anything for a month. Superb for a electronics component distributor...

Hence *since* SAP It's better, but only marginally.

My main gripe is that it's all scattered around, full of abbreviations, some stuff saves automatically and some stuff you need to burn explicitly, ... It's just needlessly a pain in the butt. And the documentation sucks.

[Ed Poore]

5 minutes ago, elbekko said:

And the documentation sucks.

Sorry there was some?...

[elbekko]

Just now, Ed Poore said:

Sorry there was some?...

Yes, but you often come out more confused than when going in.

Not the best video, but a bit of motivation at least - this was the first proper start and running for more than a few revolutions yesterday   It all sounds pretty good at idle, so I don't think the issue is timing related, but on the other hand, maybe it is...

[landroversforever]

17 minutes ago, Ed Poore said:

Having just opened TunerStudio I know that's not possible .

Before I left my previous job they brought in SAP - for almost a year and a half my boss had no clue how much budget any of his projects had. We used to have this beautifully simple internal web page called Project Browser where you could go, enter the project code and it had three tabs. Summary, Hours and Direct Expenses. On summary it said how much money the project was allocated, how much had been spent and how much was left. The other tabs broke down exactly what you expected. It was updated live as people charged to it. SAP - well after 3 years of being there when it was introduced I never successfully ran a query to know how much a project had spent.

Whilst I was farm sitting for my parents I had a visit from an electronics distributor rep - since he'd driven all the way down from Reading area gave him a cuppa and we sat and talked even though I wasn't involved heavily with the business at the time. They'd had SAP introduced a year before - one of the outcomes of that was that they were unable to ship anything for a month. Superb for a electronics component distributor...

We went from SAP to Unit4, and from an end user only using it for timesheets really it's been a proper backwards step.

[Bowie69]

Only other think I can think of is a massive air leak?

That will lean things out and cause it to surge, and die on throttle, as it will be just too lean to recover.

or... is the TPS working? Assume so as basic stuff is to set top min and max readings.

[FridgeFreezer]

Changing REQ_FUEL does indeed have the effect of moving the entire map up & down - you can be quite bold in steps at this stage, moving it 20% would not be unreasonable.

With no AE you may well get a stumble - with any of these settings or corrections it's worth remembering how they work (which is usually quite simple although the documentation makes it sound complicated) - AE is what would be a throttle pump on a carb, and works either from rate-of-change of the throttle position sensor value OR from rate-of-change of the MAP sensor value, or a blend of both. Worth saying that a wobbly throttle pedal reading due to bumpy roads or whatever can result in AE being constantly triggered which will of course throw a lot of other stuff out.

At this stage it's not worth worrying about lambda sensors at all - use your nose, does it smell rich? TBH if it doesn't smell rich from cold, richen it up until it does to get it running, you can always tune it back down later.

[Escape]

TPS is working, we got it calibrated, even bothered to switch the wires around to MS giving an error that the closed value was higher than the open one (though that shouldn't matter with proper calibration as far as we understand).

I checked for air leaks, found one crankcase breather unattached and plugged the hole, but that made no difference.

It doesn't smell rich, my guess is too lean as well, so it can't cope with anything but the smallest throttle opening. On the other hand, adding 10% to REQ_FUEL didn't seem to make a difference either. Neither did lowering it by 10%. Guess we could try +20%.

[elbekko]

19 minutes ago, Bowie69 said:

Only other think I can think of is a massive air leak?

That will lean things out and cause it to surge, and die on throttle, as it will be just too lean to recover.

or... is the TPS working? Assume so as basic stuff is to set top min and max readings.

TPS is working and calibrated. The problem already occurs at ~4% TPS, barely feathering the throttle with half a toe.

17 minutes ago, FridgeFreezer said:

Changing REQ_FUEL does indeed have the effect of moving the entire map up & down - you can be quite bold in steps at this stage, moving it 20% would not be unreasonable.

With no AE you may well get a stumble - with any of these settings or corrections it's worth remembering how they work (which is usually quite simple although the documentation makes it sound complicated) - AE is what would be a throttle pump on a carb, and works either from rate-of-change of the throttle position sensor value OR from rate-of-change of the MAP sensor value, or a blend of both. Worth saying that a wobbly throttle pedal reading due to bumpy roads or whatever can result in AE being constantly triggered which will of course throw a lot of other stuff out.

At this stage it's not worth worrying about lambda sensors at all - use your nose, does it smell rich? TBH if it doesn't smell rich from cold, richen it up until it does to get it running, you can always tune it back down later.

Hence why we disabled all of that. Only effect is that now it's hunting a bit instead of idling smoothly.

We also tried fixing the spark angle, but that only made it a bit harder to start, no effect on the cutting out.

And to reiterate, this isn't a stumble. It doesn't recover. The moment you touch the throttle, it dies. We can't get it past idle at all, hence we can't do the cam run-in.
It felt to us like it was getting a face full of fuel, hence why we tried with disabling the AE.

We'll try throwing a bunch of extra fuel at it, but I have my doubts about that tbh. We had one plug out and it smelled a bit wet. But it also barely ran, so who knows. The exhaust is definitely not eye-watering.

[Bowie69]

3 minutes ago, elbekko said:

And to reiterate, this isn't a stumble. It doesn't recover. The moment you touch the throttle, it dies. We can't get it past idle at all, hence we can't do the cam run-in.
It felt to us like it was getting a face full of fuel, hence why we tried with disabling the AE.

I'd say it is a stumble it doesn't recover from -this is very common. 

I would really triple check for air leaks, if the exhaust doesn't smell rich then it's probably lean. What's the MAP value when you do get it to idle?

[elbekko]

MAP on idle is around 40-50. Goes high very quickly even with small throttle opening though.

Just tried doing a REQ_FUEL calculation. The Bosch injectors should be 137 g/min (so ~137 cc/min?), 5L engine, gives a REQ_FUEL of 30.6... which is very far off the ~18.0 we've been trying, which is the value we got from a friend's 4.6 with the same injectors 😕 

[Bowie69]

Fuel has a density of 0.77 that of water, so 137g/min = 177cc.

 

Edited March 4 by Bowie69

[elbekko]

3 minutes ago, Bowie69 said:

Fuel has a density of 0.77 that of water, so 137g/min = 177cc.

 

Ah! That's closer, 21.8 for a 4.6 and 23.6 for a 5L. Still not the same as the friend's 4.6, but worth a try anyway.

[FridgeFreezer]

Right now the simple version is it's a stone cold engine with all the enrichments disabled so until it SMELLS RICH it's NOT RICH ENOUGH for a cold start.

After-start & Warmup enrichment (ASE + WUE) add a surprising amount of fuel (25-50%), if those aren't enabled you're massively down on fuelling and it's like trying to start a carbed car from cold with no choke.

[elbekko]

Those are still on, only disabled acceleration enrichment because it was stalling when applying the throttle.

[FridgeFreezer]

OK, does it smell rich?

[Escape]

Not really, but hard to tell from just a few seconds running. And I don't wont to idle any longer because of the cam. We'll see if there's a difference if we increase fueling by a noticeable amount.

[Bowie69]

Can you wind out the throttle cable a bit like on an RRC? If so be worth trying to get it to start with the throttle open, by adjusting the req fuel/VE table.

I'm surprised you are struggling with this, normally if everything mechanical is right then the engine will at least fire and rev up even if not stay idling.

[Bowie69]

Also.... *IF* you have a plug lead in the wrong place then it can fire back through the inlet when you open the throttle, enough to stall an engine that is not quite happy already.

[elbekko]

10 minutes ago, Bowie69 said:

Can you wind out the throttle cable a bit like on an RRC? If so be worth trying to get it to start with the throttle open, by adjusting the req fuel/VE table.

I'm surprised you are struggling with this, normally if everything mechanical is right then the engine will at least fire and rev up even if not stay idling.

I know, it's like the anti-megasquirt-installation. Normally they don't want to idle, this one wants to do nothing but.

1 minute ago, Bowie69 said:

Also.... *IF* you have a plug lead in the wrong place then it can fire back through the inlet when you open the throttle, enough to stall an engine that is not quite happy already.

The only popping we've ever heard was at the first start attempt, so I doubt it's backfiring at all. But you never know, normally we should have all the ignition stuff figured out now, but yeah...