So you use 2 sensors ?
Do you know the thread size in the header? The bosch sensor thread is M18x1.5, so will probably need some kind of adaptor?
No, I use three. Placed in the stainless headers, one for each carb. I’d check the thread in the manifolds and go from there.
ok. I have the factory EFI. So using one sensor in the header will probably be inaccurate?
EFI distributes more evenly than old carbs, if your plugs look similar you should be fine.
That sounds promising, see how hard the plugs are to get out.
Did I read Frank’s comment correctly, that the sensor needs to be disconnected while making mixture adjustments?
That makes no sense to me, but my experience is only wideband sensors and carbs.
I’ll ask Frank to elaborate. Thanks for your help, very useful.
Hi Frank, I need some clarification on what you’ve said as I don’t fully understand the concept. Are you saying that mixture adjustments need to be done with the sensor disconnected, is that because the ECU is disabling the Lambda circuit? during throttle changes? I don’t understand that at all, can you elaborate/simplify that for me? This s3 has factory EFI and ECU but no oxygen sensor, so is what your saying still applicable?
Here is a link to the Lambda gauge, it connects to a gauge that can display AFR or Lambda;
https://www.bosch-motorsport.com/content/downloads/Raceparts/Resources/pdf/Data%20sheet_69034379_Lambda_Sensor_LSU_4.9.pdf
**
No, Zenit - you do not have the Lambda
The point of the Lambda system is that it strives to maintain Lambda=1 - which means that if you alter the mixture screws the Lambda will counteract the adjustment. Ie, the AFR readings will not change. The point of the mixture adjustments is to get the mixture as close to spec as possible - the Lambda is just a refinement; the engine should be set up to run perfectly without the Lambda…
Also, all mixture adjustments should be made with steady throttle - moving the throttle; the ECU (or carbs) deliberately fattens the mixture to avoid power ‘flat spots’ during acceleration. The mixture may deliberate differ with rpms, and while adjustment specs usually use idle for adjustment - other engine speed may be specified, market depending…
Frank
xj6 85 Sov Europe (UK/NZ)
Is there also no OEM O2 sensor?
No OEM O2 sensor. Not sure what info goes to the ECU in this configuration but must be relatively simple, CTS is about all I know that it looks at but must be more than that???
Sorry - I see you mentioned this twice already…
I think reference to disconnecting O2 sensors relates to cases where there is a system with an OEM narrow-band O2 in place. Then if a tune-up calls for a mixture adjustment under conditions where the engine would normally be running in closed loop (C/L, the ECU continuously varying fuel delivery to drive continuous switching of the O2 sensor output to confirm stoic), such as a fully warmed-up, steady rpm, low load state, with the sensor connected, the ECU would defeat any mixture adjustments as you do them as Frank pointed out. An engine running in C/L can only run at stoic.
But in your case, there is no sensor to disconnect to make a mixture adjustment.
I believe wide-band sensors are not supposed to be disconnected on a running engine (can’t remember where I saw this yet), so you can make mixture adjustments whenever you like, just as the ECU does.
Thanks Frank and Colonial for explaining the Lambda circuit.
I’ve got the AFR meter installed. Just waiting for a leaning ring for the LPG mixer.
The gauge only display AFR for petrol, for other fuels with different stoich points the manufacturer recommends changing to the lambda display.
What lambda values would you recommend for various loads?
Obviously, 1 at idle. I wonder where it ideally should be when cruising up a long hill but at low rpm, around 1800rpm.
On the flat, light load cruise I guess just under 1?
Just looking for a starting point for lambda at various load conditions.
Cheers,
Where did you end up fitting the sensor?
There was only one small section of exhaust pipe where it wasn’t split in two. That was near the gearbox, sensor runs up the side of the gearbox tunnel, leads run up to center console where the ashtrays are.
So the thread on the manifold was no good?
**
Lambda=1 at idle is not(!) correct, Zenit; the idle set-up must be slightly fat; 02 sensor (if fitted) disconnected. Using CO is the only general specs given, but varies (incredibly) with markets between 0,5 to 1,75% for petrol. However, the AFR/Lambda equivalent of something between 0,5 and 1% should be OK for petrol - but LPG data is anybody’s guess…
The reason for the initial fat set-up is that ‘our’ EFI is not precise enough to maintain constant mixture throughout rev and load range. So to avoid lean mixture, which would cause hesitation or flat spots; fat set-up is called for…
I got a reasonably priced offer to weld in the bunk so I went with that. As I wasn’t sure how good the readings would using exhaust bank, and the sensor thread is metric so I probably would of spent money get that adapted anyway.
Got the LPG shim today, so will get that in tomorrow and see if I can get a good balance with the two fuels, obviously compromises have to be made for dual fuel.
Thanks for the advice, will keep it on the fat side.
I think O2 sensor threads have been consistent since they came into existence, but getting the plug out of the manifold exhaust might have proved too hard anyway.
In your shoes I’d try setting of idle mixture at maximum manifold vacuum on LPG. This is usually richer than stoic on petrol cars. It will be interesting to see what lambda you get on LPG if you try it.
I think the main compromise with dual fuel is in ignition timing, as LPG calls for a different advance curve. Electronic devices used to be available to solve this, but it looks like these have disappeared from the market. At least the curve for petrol is “safe” (slower than optimal) for LPG.