need help with CIRCUIT ECU controlled O2 sensor enhancer

Thread Starter

Mrusten

Joined Dec 23, 2015
17
Everything I've seen about the wideband AF ratio O2 sensor says it's characteristics are opposite of the old standard sensor.

Have you induced a rich and lean conditions and observed the outputs? Under normal conditions the wideband ideal is around 3.x V and the older O2 sensors is about 0.45V.

Which sensor (model number) are you using and do you have the output specifications ? It appears you want a level translation. If you want to go from 2 to 4 volts (wideband A/F O2 sensors) to 0.1 to 0.8 volts, it can be done.

There are plenty of youtube videos out there describing the sensors, what their outputs look like, and how to test them. By testing them, I'm not saying yours is good or bad, but how their output responds to varying stimuli. The display on your oscilloscope didn't look like any that were on the youtube videos.



Put yourself in the shoes of those here that might be willing to help you ... for no compensation. What do you think the first words in my mind were when I read it?
My scope picture is the load pwm where hz is measured and length of pulce, ideling at the pickture and trottle at the measure line

i dont know what the end valued need to be to get the ecu doing a propper job thats why i ask for a circuit that is adjusable at load pwm and idle pwm, between this points it has to be linear.

Adjustable according to pwm need to be an offset to the orginal o2 sensor voltage (voltage follower with an adjustable offset)

Then if the orginale o2 sensor signal is corect leav the pots at siro and no changes is made to the signal

This is not a forum to find errors and fixes for cars its about circuits
im a educated car mechanic with big interest for electronicks and automation and dont ask for help to find any errors.

I ask for help to make a circuit to offset the o2 sensor signal baced on the load pwm signal
 

JoeJester

Joined Apr 26, 2005
4,193
You want a circuit ... Ok ... what is the input voltage levels from the sensor, and what do you want them to be when they leave this "circuit"?

That first circuit could do what you want. What I find humorous about that circuit is they want you to input a 1V dc and adjust the output for 1V dc. Of course they amplify it twice and you get the same results you started with ...

Ok, your the auto mechanic. Why doesn't your scope presentation look like other O2 sensor outputs? It resembles neither the older O2 sensors or the newer A/F O2 sensors.
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
JoeJester said:
You want a circuit ... Ok ... what is the input voltage levels from the sensor, and what do you want them to be when they leave this "circuit"?

That first circuit could do what you want. What I find humorous about that circuit is they want you to input a 1V dc and adjust the output for 1V dc. Of course they amplify it twice and you get the same results you started with ...

Ok, your the auto mechanic. Why doesn't your scope presentation look like other O2 sensor outputs? It resembles neither the older O2 sensors or the newer A/F O2 sensors.
Im sorry im not trying to be ruude here,
My scope picture is from the load signal going to the ignition (ezk)
Has nothing to do with the o2 signal
about operation of sensor here http://www.autotap.com/techlibrary/understanding_oxygen_sensors.asp

this values do dot actual matter cos i need only offset by +-500mv following the orginal o2 signal

i mean adjustable by pot +-500mv at idle pwm (loadsignal from scope)

And another pot to adjust +-500mv at trottle pwm.

Between those adjusted values should be linear.
Example:
0% duty cycle +100mv offset adjusted by pot
100% duty cycle +500mv offset adjusted by the other pot
it makes 50% linear and at 200mv bechause its 400mv between 0% and 100% in this example
50% duty cycle

im electric amatuer so im sorry for terms
by offset i mean example 0% +100 mv is added to the orginal o2 sensor signal

it may be a good idea to convert the square wawe signal to a triangle wawe by a opamp then input triangle wawe to another opamp to convert it back to square wawe but with a pulsewith adjustment (pot at the noninverting terninal)

Something like this to have controll of the input to adjust the whole circuit.

then maybe the signal could be converted to dc values depended of the load again this could be used as a referense voltage for the opamps for the o2 signal

am i on the right way or am i lost ?
 

tcmtech

Joined Nov 4, 2013
2,868
Definitely lost from my view of what you say you want to do and how you appear to think that the fuel injection control systems work.

I'm guessing english is a second language for you as well?
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
Definitely lost from my view of what you say you want to do and how you appear to think that the fuel injection control systems work.
why lost?
This is how bosch lh2.4 works.
anyway there is lots of different systems out there

im just looking for a circuit to add or take away adjustable +-500mv to a signal
controlled by pwm.

as you see this way im able to try out all possibiliys to see whats answers its possible to get from the engine

I'm guessing english is a second language for you as well?
hehe you guest it. And no editor :p
 
Is the MAF (Mass Air Flow) sensor good?

You can enrich the engine easily with propane. I used the "propane enrichment" method to set carbs way back.
Back in the old days, I put a gate valve on a propane torch head. Removed the head and adapted a gate valve and a vacuum line to it.

The technique involved increasing the propane until there was a peak in the RPM. Too much propane would reduce the RPM. The idea was to peak the RPM and to turn off the propane and the RPM should not vary more than 50. This made sure the carb was in the idle system. Worked every time.

Never tried it on an EFI engine.
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
Screenshot_2015-12-28-08-22-58.png
Screenshot_2015-12-28-08-22-38.png
Screenshot_2015-12-28-08-22-10.png

If the circuit works like this its no limits for testing

only 0% and 100% is adjustable the rest is linear as you see
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
Is the MAF (Mass Air Flow) sensor good?

You can enrich the engine easily with propane. I used the "propane enrichment" method to set carbs way back.
Back in the old days, I put a gate valve on a propane torch head. Removed the head and adapted a gate valve and a vacuum line to it.

The technique involved increasing the propane until there was a peak in the RPM. Too much propane would reduce the RPM. The idea was to peak the RPM and to turn off the propane and the RPM should not vary more than 50. This made sure the carb was in the idle system. Worked every time.

Never tried it on an EFI engine.

All sensors are new
im looking for a circuit for testing please look earlyer in the tread
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
What test equipment did you use to get the raw data for those drawings?
Are you here to help or to make fun for your self?

if you read more attentevly you wil understand that this is examples for how the circuit should work to be able to do tests.

anyway that was maybe a bad way to try to explain how the adjustable 0% and 100% dutycycle should work with all dutycycles between those 2 setpoints.
What i ment with linear
 

JoeJester

Joined Apr 26, 2005
4,193
Are you here to help or to make fun for your self?
I re-evaluated all you wrote and decided not to participate anymore. Have fun.

I'll leave you with this. If there is a piece of test equipment that will do what you want, tell people about the model number. At least they will get a better idea of your specifications.

need help with CIRCUIT ECU controlled O2 sensor enhancer
sounded like you wanted to modify your O2 sensor's signal. You never mentioned the type of sensor, but mentioned the output indirectly. Then you changed and showed what you called a PWM that you wanted to "modify".

Now, it's varying the PWM signal's DC level from negative 0.5 volts to positive 0.5V and modifying the PWM signal from 0 to 100%.

Settle on what you really want to do.

I'm done.
 
Last edited:

tcmtech

Joined Nov 4, 2013
2,868
why lost?
This is how bosch lh2.4 works.
anyway there is lots of different systems out there

im just looking for a circuit to add or take away adjustable +-500mv to a signal
controlled by pwm.

as you see this way im able to try out all possibiliys to see whats answers its possible to get from the engine
From what I am getting from your posts you feel that the type of O2 sensor should have a constant and stable output voltage when the engine is running correctly. Unfortunately that's not how any O2 sensor and fuel injection systems work in normal closed loop operation. Rather the system constantly swings from one end of the sensors limits to the other only briefly crossing the actually wanted ~.450 - .5 volt stoichiometric A/F ratio the system wants 2 - 4 times a second trying to average out the upper and lower swing limits to achieve the averaged A/F ratio it is programed to try and maintain.

Given the O2 sensors lower output voltage is ~.1 volts at a ~16:1 A/F ratio and near 1 volt at a rich ~12:1 A/F ratio that's all the voltage range the computer system will accept as being a valid O2 sensor signal. Anything it sees below the ~.1 volts or above the ~1 volt levels it will simply see as being out of range which it will likely log as a sensor error condition and ignore or set a trouble code for.

There's what you have to work with. A .1 to 1 volt range that is constantly swinging from one end to the other only averaging out to the mid point A/F ratio.
If you want to bias the average signal to make the engine run steadily in a leaner or richer A/F ratio you only have a at best about 3 tenths of a volt bias above or below the midpoint to work with for either adding or subtracting voltage from the actual O2 sensor's signal for biasing while still staying within the .1 - 1 volt limits the computer will accept as being a valid signal.

As for the PWM concept you are talking about I don't see where it comes into play when dealing with the O2 sensors actual signal being it's a analog signal not a PWM one.
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
From what I am getting from your posts you feel that the type of O2 sensor should have a constant and stable output voltage when the engine is running correctly. Unfortunately that's not how any O2 sensor and fuel injection systems work in normal closed loop operation. Rather the system constantly swings from one end of the sensors limits to the other only briefly crossing the actually wanted ~.450 - .5 volt stoichiometric A/F ratio the system wants 2 - 4 times a second trying to average out the upper and lower swing limits to achieve the averaged A/F ratio it is programed to try and maintain.
offcorse i ment average voltage, my bad

Given the O2 sensors lower output voltage is ~.1 volts at a ~16:1 A/F ratio and near 1 volt at a rich ~12:1 A/F ratio that's all the voltage range the computer system will accept as being a valid O2 sensor signal. Anything it sees below the ~.1 volts or above the ~1 volt levels it will simply see as being out of range which it will likely log as a sensor error condition and ignore or set a trouble code for.
i agree but i dont actually think the LH2.4 set any trouble code at the O2 sensor
cos that code dosent exist, closest one is the code #to rich, to lean#

anyway a zener diode or some how to cut the voltage outside of range should fix that, I don't think this old ecu competable to read the signal that attentevly to understand that the sinewawe become dc for a short while when its out of range, but it might understand that the voltage of the sinewawe goes out of range


There's what you have to work with. A .1 to 1 volt range that is constantly swinging from one end to the other only averaging out to the mid point A/F ratio.

If you want to bias the average signal to make the engine run steadily in a leaner or richer A/F ratio you only have a at best about 3 tenths of a volt bias above or below the midpoint to work with for either adding or subtracting voltage from the actual O2 sensor's signal for biasing while still staying within the .1 - 1 volt limits the computer will accept as being a valid signal.

As for the PWM concept you are talking about I don't see where it comes into play when dealing with the O2 sensors actual signal being it's a analog signal not a PWM one.
as I said above I belive the voltages outside the range of the o2 sensor could be removed.


##On the other hand this problem will newer happen because when the ecu read the bias signal it will do what ewer it can to change the AFR back to its programmed value## it means the value it can read after passing the bias circuit


the reason of 500mv was just to make a number for another user here, but 500mv is reasonable and in this case I will not be handicapped under testing.

I will try to explain again about the PWM, if you look at the circuit in the first post, the o2 sensor enhancer circuit, its controlled by a POT, so here is why I want controlled by pwm from the ECU

my goal here is not to adjust a fixed bias to the o2 sensor signal but to change the bias to the signal after what is required by the load of the engine.

example orginal the ecu work like this (just example may not be 100% correct numbers!)
5% load- closed loop- afr 14,7
10% load- closed loop- afr 14,7
20% load- open loop- afr 14,5
30% load- open loop- afr 14
40% load- open loop- afr 13,8
50% load- open loop- afr 13,6
100% load- open loop- afr 12

the reason for open loop is for the engine to be able to run more rich and work good without pinging and self distraction.
open loop = ECU works from fixed values in the maps no feedback from the O2 sensor.

the narrow band o2 sensor can not detect this values cos its out of range.
The ECU don't read any other signals eater and there is no way to change the programmed afr ratio in closed loop.

anyway the threshold for open and closed loop is adjustable, so I want to be in a closed loop longer and in this stages the ecu should think that it is still at the programmed afr value

after modification example
5% load- closed loop- afr 14,7 bias example 0
10% load- closed loop- afr 14,7 bias example 0,05
20% load- closed loop - afr 14,5 bias example 0,08
30% load- closed loop - afr 14 bias example 0,10
40% load- closed loop - afr 13,8 bias example 0,11
50% load- closed loop - afr 13,6 bias example 0,20
100% load- open loop- afr 12

So in this case I got a wideband with a narrowband output, or I could actually use the wideband voltages too if the signal can be biased by 5 volt, the signal is always biased so that the ECU think that it is in the programmed spot of 14,7

When I have the PWM output from the ECU is can be used to do the adjustments of the signal bias circuit.
The ECU is programmed to keep the input of O2 sensor at 14,7 at ANY cost, what is written in the main fuel map don’t matter at all so if the wideband meter show correct afr the bias circuit can be adjusted so the ECU got the 14,7 input and the ECU will be happy


Thanks for your time at this project


2 weekends ago I modified the mainfuelmap and deactivated the closed loop option
I notest emiditly less fuel consumption ann the car uses 10 times the distance to get to engine work temperature compared to before

Yesterday I had a test finally,
1981 Volvo 240
Average outside temperature: -1,5C
Distance 205 km
7,9 L/100km
Average speed 82,10km/h
Not at highway so speed was down and up a lot
Last 20 min was at highway and constant speed of 110kmh (gps)
5 speed gearbox
Rear axle 3,54

This is werry good as I come down from 15++/100km
I belive when its tuned more to perfection not by just random it will use eaven less
But to do this it need to be in closed loop
Too much fuel can still be smelled from the exhaust at times


The wideband is not fited yet cos its no reason to before it’s a way to bias the signal
 

pbr333

Joined Jan 22, 2012
2
Mrusten the old hot-wire injection on that volvo shouldn't be affected by todays fuel unless your burning E-% at the pump. The feed back
circuit works to maintain 14.7:1 stoik....by using a reference voltage of .5 v and anything above/below drives it rich/lean but it's just trying to
tweak a fuel metering system on little 4cyl heavy car that will never get good fuel mileage. Of course that is 50/50 my memory/ my opinion!
but I didn't know Bosch had removable or programable chips in those early systems either, that's interesting. I need to read more on this, but
what got my attention was you stated it like this
"ECU stay at what is thinks is AFR 14.7 that is 1 volt from the LAMBDA (O2 sensor)"
and I always understood it as I stated at the beginning of my post. I am a little short on time but I will be back to read more.

pbr333
 

Thread Starter

Mrusten

Joined Dec 23, 2015
17
Mrusten the old hot-wire injection on that volvo shouldn't be affected by todays fuel unless your burning E-% at the pump. The feed back
circuit works to maintain 14.7:1 stoik....by using a reference voltage of .5 v and anything above/below drives it rich/lean but it's just trying to
tweak a fuel metering system on little 4cyl heavy car that will never get good fuel mileage. Of course that is 50/50 my memory/ my opinion!
but I didn't know Bosch had removable or programable chips in those early systems either, that's interesting. I need to read more on this, but
what got my attention was you stated it like this
"ECU stay at what is thinks is AFR 14.7 that is 1 volt from the LAMBDA (O2 sensor)"
and I always understood it as I stated at the beginning of my post. I am a little short on time but I will be back to read more.

pbr333
Im sorry offcorse its wrong, should write 0,5v. Anyway the actual goal of the ecu dont matter, what mattet is the real afr measured by the wideband and the manipulated signal goes to the ecu.

about bosch LH 2.4 there is chipable ecu and ezk's (ignition)
A qick google let you find the turbobrics forum where its loads at this
 

cork_ie

Joined Oct 8, 2011
428
hello i got an interesting project but have no experience in designing a circuit

so basic story is that i got an old volvo that get too much fuel with the new generation fuel, i think its because new fuel contanes more oxygen burn cleaner and leave more oxygen in the exhaust witch the O2 sensor pick up and the ECU think its running lean then compensate with more fuel.
hopefully someone can help
Save yourself a lot of bother first with some quick checks
Fuel , Petrol aka Gas in the USA does not contain oxygen that is why it is called a hydrocarbon !! If it were oxygenated it would be an alcohol or aldehyde.
Check your emissions reading (especially Lambda) on a gas analyser and compare it with the O2 sensor reading.
Check the sensor reading is fluctuating in a range of 0.2V to about 0.8V once per second when engine is hot If there is a discrepancy it is likely that you have
1) A poorly performing O2 sensor
2) A crack in the exhaust or leaking manifold gasket. Believe me, the exhaust gases will pull air in through any little crack or leak as they flow through the exhaust system. The O2 sensor will detect this extra air as a weak mixture and signal the ECU to increase the fuel quantity.

You also need to check the control pressure on the KH Jet unit.
I did a lot of work on these systems 20 years ago - they were also fitted on the mercedes and gave plenty trouble
 

cork_ie

Joined Oct 8, 2011
428
Having read through all the posts there appears to be some terrible confusion .
The O2 sensor AKA Lambda sensor in this car is not a wideband sensor I am 100% sure of that .
It is a standard narrowband Zirconia sensor and as far as I can remember had a slight voltage shift so that you had to buy a new one from Bosch or Volvo and could not fit a universal type.
Wideband Zirconia sensors are relatively new and generally have 6 wires instead of 4.
The only wideband sensor of that era were the Nippon Denso Sensors fitted on Toyotas which worked on a completely different principal.

You can do anything you like but the Bosch LH ECU was the very first made with adaptive learning and it is going to continually adapt against you no matter what you do, if their is an underlying problem.
Have a read of this and you will understand exactly how the Lambda map is controlled in this system.
http://ipdown.net/jetronic.info/tiki-index.php?page=LH+2.4+Lambda
 
Top