O2 sensor test and inverse duty cycle - PeachParts Mercedes-Benz Forum

LeaUK · #1 04-06-2005, 02:43 AM

I've read on Steve B site http://www.peachparts.com/Wikka/EngineControls that when measuring the O2 output with a scope (I hope to do this Saturday) the duty cycle is the ratio of low voltage to high voltage. But this is contradictory to the duty cycle definition in the electronics industry. Where it's always high to low.

ie Steve shows a nice piccy of a scope (Fig 1) showing 30%, but clearly this should be 70% in my book. Is this due to the electronics inverting the output before presenting it to the X11 connector?

I'm pretty sure that O2 sensors themselves o/p a voltage 0.1-0.9V (approx) with the output increasing with detected O2?

Could someone help and clear my confusion..

Cheers
Lea

A264172 · #2 04-06-2005, 10:26 AM

I ran across this recently when trouble shooting my emissions.
Kept getting 30% instead of 70% ignition on engine off
In reading thread after thread I found a reference to Mercedes being the reverse of the industry. To obtain the book value of duty cycle you have to reverse the leads (for me [86 300E] at the X11 diagnostic connector pin #3 the Lambda feed should be connected to the black or ground lead of the meter, and the red "com" lead is connected to pin #2, the ground pin).
This left me wondering for a long time if I had controller problems...
Not sure if the O2 sensor requires reverse leads or not.

Duke2.6 · #3 04-06-2005, 11:48 AM

The duty cycle is a synthesized signal that measures the relative amount of time that the lambda system spends adding a rich or lean bias to the basic mixture provided by the air flow sensor and mechanical section of the fuel distributor and is used to make any necessary mechanical mixture adjustments to keep the lambda system near the mid point of its control authority range.

It's essentially a measure of the mechanical mixture bias. If the duty cycle is 50 percent it means the mechanical mixture is spot on stoichiometric, but the mechanical mixture adjustment is not accurate enough to maintain an exact mixture, so the duty cycle will vary slightly as the lambda system corrects the mixture to stoichiometric with finer granularity and greater accuracy.

The other criterion is that you want the average duty cycle at idle to be within 10 percent of the average at 2000 RPM no load. I prefer a slightly less than 50 percent duty cycle at idle as this indicates a slightly rich bias of the basic mechanically derived mixture, which will aid cold starting and warmup driveability. When I checked my duty cycle it was in the range of 40-45 percent at idle and 50-55 percent at 2000, so I made no adjustments, and the cold start characteristics are excellent.

If you have a scope, you can eavesdrop on the actual O2 sensor output by pulling the O2 sensor connector out just enough to get a scope test lead attached with the other scope test lead to ground. The wave form should jump quickly between about 0.15 and 0.88 volt and back with a frequency of about 1-3 Hz at idle.

Duke

LeaUK · #4 04-06-2005, 03:05 PM

O2 sensor testing

Quote:

The duty cycle is a synthesised signal

Are you describing the signal measured at the diag connector or at the Lambda? If the diag signal represents a synthesised 'input' then can it be used to measure how much compensation the system is making? In a closed loop system, if only the output is measured, we cannot tell how much adjustment has been made.

Quote:

If you have a scope, you can eavesdrop on the actual O2 sensor output by pulling the O2 sensor connector out just enough to get a scope test lead attached with the other scope test lead to ground. The wave form should jump quickly between about 0.15 and 0.88 volt and back with a frequency of about 1-3 Hz at idle.

Looking at the graph for the O2 sensor, I assume this is due to the extremely high transfer ratio about the stoichiometric value. Is this true?

I have a scope and will be doing this check. But I'm not sure what I will achieve. If the closed loop system is running correctly I would expect a reading of 50%. I think my engine is a little lean and so as it's closed loop am I correct in thinking I'll still see 50%. I was hoping to find out how lean.

Which signal reveals more, the diag synth sig or the O2 sensor output direct. Whilst typing this, I'm starting to think the diag o/p....

Failing Cat?

Not only this, but my Cat is potentially failing CO emissions (at tailpipe), these are:

Heated Cat -
Co 0.2-0.3% (UK limit 0.3%)
HC 100ppm
Lambda 0.97-1.02

With such a good Lambda reading I'm guessing that the O2 sensor and control system are performing fine, but my 12 year old Cat is simply starting to give up. Hence I want to check the O2 sensor with a scope just to see what's really going on.

Can anyone recommend a test procedure by any chance?

Thanks
Lea

A264172 · #5 04-06-2005, 03:21 PM

How old are your cap/rotor/plugs?
and How far into your next oil change are you?

I achieved a 1.5% CO reduction and a drop of 100ppm HC with an oil change and a new cap & rotor.

LeaUK · #6 04-06-2005, 03:52 PM

Hi

An interesting thought, but the caps and rotors are only a year old although I've had to clean one recently due to tracking - these cars eat caps! Maybe worth a look and clean again?

More reading about my X11/4 and X11 connector (I have both) seem to suggest that the signal available is simply the O2 sensor output which I think I've convinced myself wont actually tell me much.

How does one measure the compensation value the closed loop system is using to keep the 14.7:1 ratio?

In the electronics industry it's easy, we simply measure the input.

A264172 · #7 04-06-2005, 04:40 PM

On my car, you measure it either at the EHA (with a harness) or at X11 (+ to pin#2 / - to pin#3) fully warmed up.
At X11:
You should get a fluctuating duty cycle (fixed is a fault).
50% fluctuating = stoichiometric
Lower %'s indicate leaning of the base mix
Higher %'s iindicate richening of the base mix

Was that your question?

James Redman · #8 08-12-2007, 03:24 PM

"If the duty cycle is 50 percent it means the mechanical mixture is spot on stoichiometric,...

Duke[/quote]

I think Stoichiometric gives a Lambda output voltage of about 0.45 to 0.48 volts (check this) on the black wire of a 4 wire O2 sensor.

I thought a 50% duty cycle meant the injectors are open 50% of the time....I am learning from this Web Site but have I missed something in this thread? At idle I would expect a duty cycle around 5 to 10% and WOT 85%.

In fact may I ask if anyone knows the idle duty cycle for a standard M119 engine?

Jim.

JimF · #9 08-12-2007, 04:06 PM

Menu#4 on my web page shows a 'graphical' Lambda of "1" (when it's at 0.45V) but, as you know, this voltage is sweeping from from top (0.8V) to bottom (0.05V), so it's continually changing.

If you measure Lambda on a 'd/c' meter, it should be 50%. . . at idle and at high rpm. The fact that it isn't could mean a lot of things. Missiing/fouled injector, bad plugs, contaminated O2 sensor, contaminated/failing MAF, and the list goes on.

James Redman · #10 08-21-2007, 03:31 AM

Jin F, Thanks for the steer to your menu 4 dealing with on-off ratio.

This may be silly question time but I am still confused. I think there is an Electronic Engineer Defintion of Duty Cycle and Fuel Injection Engineer Definition of Duty Cycle and both are floating around here.

On Jim's menu 4 I read that the Lambda = 50% is OK (but what is the on-off ratio?). I still cannot relate that to the nice and simple injector opening time definition of Duty Cycle.

Please can some one clarify what is going on for this simple mechanical engineer?

Regards,

JimF · #11 08-21-2007, 10:26 AM

Quote:

Originally Posted by James Redman

Jin F, Thanks for the steer to your menu 4 dealing with on-off ratio.

On Jim's menu 4 I read that the Lambda = 50% is OK (but what is the on-off ratio?). I still cannot relate that to the nice and simple injector opening time definition of Duty Cycle.

Regards,

Lambda is equal to "1". . . scale along the 'X' axis when the D/C = 50%.