TS 'miss'

[GTVeloce]

Ok, chasing some thoughts before I spend on a new lambda sensor as I suspect that is the cause of my problems.

When I run my TS on 98 RON open loop it makes great power and delivery from low in the rev range, however, it doesn't like idling very well and has a tendency to hunt and stall, especially with the AC on.

When I run it on 95 RON closed loop it starts and idles beautifully (even with AC), but, accelerating (except at WOT) has a miss until I am getting up around 4k and up. Fuel economy is good (getting 6.9l/100k on the open road). It doesn't sound like it is pinging.

Any ideas? I am leaning towards lambda sensor because AFAIK it is the only changed sensor by switching between open and closed loop. Is any other sensor interrogated differently?

Thanks

[ALF750]

I wouldn't change the lambda sensor yet.   It would be good to borrow another or use a temporary O2 sensor (Innovate or similar) to check what is happening.    My understanding is that all fuel quality plug programs will be in closed loop at low power settings and go open loop during acceleration or at high power.  Even 98/no cat would operate closed loop at low power I believe, 98/w.cat would just cycle between rich and lean to keep the cat at the right temp.   Festy or some more knowledgeable people here would know.  Your fuel economy is good, maybe too good - suspect this is a 'no cat' setting and ECU trims fuel to lean of stoich all the time without cycling to rich?    I'd check all the plugs after a period of rough idling (98 setting) and see what that tells you- do same during acceleration with 95 setting.   I'd also inspect and clean the distributor caps and rotors.   I ran my TS on a 'no cat' setting for a while, despite having a cat fitted, I thought it went off tune at low revs after a bit and I'm not sure why.   Another thing is that the ECU 'learns' when in closed loop mode and applies a fuel trim to the closed loop map, I'm not sure if this 'trim' continues to be applied if you change maps (by changing fuel quality plug setting).   It may take a while for the ECU to re-learn between changes, so running is degraded for a while.   Hope this helps.

[festy]

The 98/no cat and 95/cat settings use completely different tables (maps) in the TS ECU. There's a few that are common between them (AFM transfer, AE/DE thresholds, dwell, injector voltage etc) but pretty much every other table influencing either fuel or ignition is unique to a fuel setting.
And apart from the difference in tables, the actual operation is different - it's like a different program, not just different settings.

The most visible difference between those two fuel settings is that the 95/cat uses an 02 sensor whereas the 98/no cat does not.
The O2 sensor is used for closed loop control to operate the catalytic converter, and is only invoked for very low load conditions like idle or light throttle cruising. It's also not used until the engine has warmed up, so cold idle is always open loop.

The 98/no cat fuel setting obviously has no closed loop mode (i.e. no feedback from an O2 sensor) so it uses a fuel trim adjustment to enable fine tuning for emissions. That's done via the "CO pot" on the AFM - it's usually sealed to prevent adjustment. The 75TS I recently worked on had a pretty rough idle and was bogging down on throttle crack. I backed off the CO pot a little until the idle smoothed out, and it made quite a difference.

The learning capability of these ECUs is almost nil. They have just enough spare RAM to remember the last 5 fault codes and not much else.
When you change the fuel quality plug setting, the change is picked up the next time the engine speed changes by +/-40RPM which in practice is pretty much immediately.
 
Checking the fault codes is always a good place to start diagnosis, all you need at a minimum is a small 12v globe and a couple bits of wires for a makeshift "check engine" lamp, the ECU will use it to flash the codes.

My first guess would be that your O2 sensor is working fine, but maybe your AFM has a dirty wiper track (or someone has messed with the flap spring).
My reasoning is that it idles fine in closed loop but not so good in open loop because the O2 sensor feedback is allowing the ECU to correct the slightly-messed-up fuel mixture, but as soon as you start accelerating it drops out of closed loop and the effect of the dirty AFM contact isn't being cancelled out anymore.
At higher RPM the AFM flap is swinging a long way, onto a rarely used (and not worn out) part of the wiper contact so the problem goes away.
As I said it's just a guess, but based on your description that's where I would look first.

[LukeC]

I can't help with the problem at here, but I am just installing AC in my TS-GTV and may be relevant. I'm getting it gassed tomorrow, and not too soon.

There is a grey/yellow wire at the RHS rear of the engine bay connector where the Motronic loom meets the vehicles main loom. This is 12V going to pin 32 (I think) on the ECU. I would have thought that power to this may lift the idle speed, but it does not appear to do anything. This is not downstream of the A/C pressure switch.

Any ideas what it actually does? I can't imagine there would be an input that does not do anything.

[GTVeloce]

All good info - thanks. Should have also mentioned it is a GTV-TS with an O2 sensor but no cat on a 2in open exhaust.

I put the excellent fuel economy down to the lighter weight and slipperier shape of the GTV.

I will start by cleaning the AFM track and maybe messing with the trimpot as both of those are at a price point I like 

Luke - There is a wire (I don't have the diagram in front of me ATM) that sends a signal from the AC compressor to the ECU that I assume 'tells' the ECU that AC is on and therefore raises the idle slightly. I have quite got around to hooking it up but plan to do so. Conversely, when I turn the AC on at idle it drops a little and then catches itself. Of course, this could just be related to my other problem...

While I should have been working on fixing my miss over the weekend I got stuck doing this instead:

[LukeC]

And this has been keeping me off the streets at night. Back together, hope its worth it...

[festy]

There are a couple of inputs (and outputs) that don't do anything - there's even more on a Potenziata ECU though.

Pin 29 and 32 are the A/C request inputs. They're active low - i.e. ground them to enable.
My memory is a little sketchy as it's been quite a while since I looked at that part of the code but IIRC whenever the ECU thinks the engine is idling (low RPM, low load, idle switch closed) it looks up the desired idle speed from the relevant maps (based on engine temp, battery voltage etc) then checks to see if either of those two A/C pins are grounded.
If so, it looks up a "minimum A/C idle speed" setting and if the previously determined desired idle speed is lower than this, it uses the A/C idle speed instead.
     
I recall the minimum A/C idle speed being something like 840 RPM, so you'd only ever see the idle speed increase if it was lower than that without the A/C switched on.

The TS I recently worked on preferred to idle a fair bit higher than that, so switching A/C on and off had no effect - but I recall seeing idle speeds around 800 RPM in some of the maps so under some conditions it could bump the idle speed a bit.

[ALF750]

well, I got that all wrong, my bad!  But it provoked some good discussion at least.

Festy - so to confirm, the O2 sensor is just along for the ride when in a 'no cat' fuel plug setting?   (do UK cars without cats have no O2 sensor?)  And our ECUs can not learn or trim fuel at all above low power when we make exhaust or intake or any engine mod to improve breathing?  I have read your 'what it can and cannot do' thread, but don't remember this bit.  Thanks for your huge work on the topic, blows my mind.

[festy]

Festy - so to confirm, the O2 sensor is just along for the ride when in a 'no cat' fuel plug setting?   (do UK cars without cats have no O2 sensor?)

That is my understanding - non-cat cars didn't have an O2 sensor. The O2 sensor is only there so the catalytic converter can operate, there's no other reason for the engine to be run at 14.7:1 air:fuel ratio. That's too lean for decent power, and too rich for lean cruising, but just right for the cat.
With the Alfa 33, Australian and UK cars got a different ECU to the rest of the world.
Because there was no need for a catalytic converter here, no O2 sensor was fitted. They also saved a few cents and didn't fit the sensor driver chip to our ECUs.

And our ECUs can not learn or trim fuel at all above low power when we make exhaust or intake or any engine mod to improve breathing?  I have read your 'what it can and cannot do' thread, but don't remember this bit.  Thanks for your huge work on the topic, blows my mind.

By design, the AFM accommodates minor intake/exhaust improvements without re-tuning. If your engine is breathing better, the AFM sees the increased air flow and so the ECU adds fuel to match.
Whereas with a MAP based ECU, changes like that usually require messing with the ECU's VE table to make the most of breathing improvements because they don't directly measure the air consumption.

The learning in these old Motronics is really limited, they do a bit of averaging but not really learning.
In some ways these Motronics were pretty advanced compared to their contemporaries, but by the late 80s they were pretty much dinosaurs.

[GTVeloce]

Ok, had a go at cleaning the track but it doesn't seem to have had much of an effect. So far all I did was spray contact cleaner on to it and operate the flap. Is there a better way?

Will adjusting the CO pot affect idle, throttle under load or both? Currently the car idles nicely and cruising at speed is fine, it's just under load that it 'misses'.

[ALF750]

thanks Festy.

[GTVeloce]

Thought I would refresh this thread. I managed to pick up another ECU and AFM in good condition. Plugged the ECU in and power definitely picked up but the stalling issue was still there(running on open loop). Swapped the AFM and hey presto, it has never stalled again and smooth power delivery. However, now idle is not quite as smooth and seems quite low.

Next step is to play around with the CO trimpot and see whether that smooths out the idle like it did for festy. I will report.

In the mean time, really enjoying the new Bridgestone Re003's!