Lambda sensors?

[tvrolet]

Having been dragged kicking and screaming in to the 21st century with an ECU on the LS7 (as opposed to carbs and a distributor) there's obviously the option to do some fine tuning of the maps. I have 2 spare channels on the data logger (which read/record volts) and I was thinking about wiring them up to a couple of lambda sensors, one per bank.

They wouldn't be connected up to the ECU; just referenced when playing back the logs and the option of displaying live figures on the dash in volts. Having googled a bit some articles suggest sensors toggle back and forward many times a second? I don't plan to use any box of tricks to decode anything - ideally I want 3 wires maximum to each sensor (live, signal volts and earth - can I get then without an earth that just earth off the exhaust?). And I want a signal that's a straightforward steady voltage (range 0-12v).

Is this do-able to wire the sensors directly into a logger? And if so, is there any type/brand to go for or avoid?

[stevieturbo]

tvrolet said:

Having been dragged kicking and screaming in to the 21st century with an ECU on the LS7 (as opposed to carbs and a distributor) there's obviously the option to do some fine tuning of the maps. I have 2 spare channels on the data logger (which read/record volts) and I was thinking about wiring them up to a couple of lambda sensors, one per bank.

They wouldn't be connected up to the ECU; just referenced when playing back the logs and the option of displaying live figures on the dash in volts. Having googled a bit some articles suggest sensors toggle back and forward many times a second? I don't plan to use any box of tricks to decode anything - ideally I want 3 wires maximum to each sensor (live, signal volts and earth - can I get then without an earth that just earth off the exhaust?). And I want a signal that's a straightforward steady voltage (range 0-12v).

Is this do-able to wire the sensors directly into a logger? And if so, is there any type/brand to go for or avoid?

narrowband sensors, ie 3 wire are only accurate at lambda 1. ie 14.7:1
They are not ideal for what you want. They also output 0-1v.

And they only swing either side of stioch when operating in closed loop fuel control, which is the ecu causing the rich/lean swing, not the sensor. The sensor just tells the ecu and the ecu adjusts mixture accordingly.

The only thing that will be of any real use for what you propose is a proper wideband kit. If you dont need an actual mixture display, Haltech offer a dual channel wideband kit, which can output a 0-5v signal for each bank to your datalogger

most other setups will require 2 independent wideband kits, one for each bank.

[Richyvrlimited]

And the good ones have separate heater and signal grounds. They definitely don't ground through the sensor.

It's advantageous to have the ECU adjusting the fuelling based on the Lambda feedback too, (closed loop fuelling), can help save fuel when cruising and pass emissions at MOT time.

You need to be dragged kicking and screaming a little bit more I think

[stevieturbo]

of course it has to be asked. Your using an LS7 engine, so what ecu or method of fuel/spark control are you actually using ?

[tvrolet]

stevieturbo said:

of course it has to be asked. Your using an LS7 engine, so what ecu or method of fuel/spark control are you actually using ?

Its a Canems ECU that at the moment at least is on an open loop. I think it can be hooked up to O2 sensors, but we're still at the early stages of mapping (1 dyno session only). This is primarily a track car, so fuel economy, idling and smooth low throttle performance are nice-to-haves, but not top of the list.

I was more interested in monitoring the O2 through logger to pick up problems rather than going for the ultimate in great running...although that would be nice too.

[stevieturbo]

tvrolet said:

Its a Canems ECU that at the moment at least is on an open loop. I think it can be hooked up to O2 sensors, but we're still at the early stages of mapping (1 dyno session only). This is primarily a track car, so fuel economy, idling and smooth low throttle performance are nice-to-haves, but not top of the list.

I was more interested in monitoring the O2 through logger to pick up problems rather than going for the ultimate in great running...although that would be nice too.

You need widebands, anything else is a waste of time/money as they wont give you any usable information.
It seems your ecu can accept two inputs, so the Haltech dual wideband kit would be ideal and you can log everything with the ecu.

[spitfire4v8]

You have to make sure your 2 inputs can be calibrated against the 0-5v output of the sensors .. your sensor manual will have the afr:voltage curve that you need to input into the ecu. If the two inputs cannot be configured in this way then you can still usefully put an AFR gauge to them but you will have to reference the voltages back to an afr manually (ie the ecu won't display a converted-to-afr value immediately). If seeing the afr directly in the logs is what you really want (it's what I would want) and your ecu doesnt allow for this then the canems wasn't a good choice. I would be surprised if it can't do this though at least on one channel

[tvrolet]

spitfire4v8 said:

You have to make sure your 2 inputs can be calibrated against the 0-5v output of the sensors .. your sensor manual will have the afr:voltage curve that you need to input into the ecu. If the two inputs cannot be configured in this way then you can still usefully put an AFR gauge to them but you will have to reference the voltages back to an afr manually (ie the ecu won't display a converted-to-afr value immediately). If seeing the afr directly in the logs is what you really want (it's what I would want) and your ecu doesnt allow for this then the canems wasn't a good choice. I would be surprised if it can't do this though at least on one channel

The plan isn't to connect these to the ECU at all, but just to the data logger (Pi System 2). It already logs all pressures, temperatures, revs, speed, gear ratios, voltage, g-forces and does circuit maps independently of the ECU. I just thought that would be a good place to log O2 - allbeit in volts. Just means I could overlay O2/volts by bank over a chart of revs, load, temps etc to see how well the ECU was doing and spot any issues early or note areas for improvement.

[stevesingo]

http://www.14point7.com/Support/Software_Documenta...

just what you need for $105 each.

[stevieturbo]

stevesingo said:

http://www.14point7.com/Support/Software_Documenta...

just what you need for $105 each.

A few people over on Turbominis handed over money to that place....

And note, I dont mention them receiving any goods for that money.

IMO given your ecu can datalog, it would be silly not to input wideband data to the ecu for logging there.

[Krikkit]

It also leaves you open to using closed-loop fuelling later down the line with minimal work, assuming the ECU supports it.

[84Dave]

www.fuelairspark.com . They manufacture a dual wideband sensor system with a nice display that stores data for up to 44 minutes of time, & requires no computer.