Help to Design EGR Valve Simulator- Cheater

[joseico90]

Hi,

I'm removing EGR from my car's engine and the only remaining thing is the valve itself, it's connected to the ECU.

I need to keep the ECU happy with a "cheater" circuit and I need of help design it.


Please see page 3 on this doc for more details on this valve. http://www.ms-motor-service.com/ximages/PG_SI_0085_EN_WEB.pdf

Correct me if I'm wrong..

As the valve opens and closes with power to it's coil from the ECU, the pot sends a valve position signal back to ECU on pin 6. The more the valve opens the higher it is.

I think I'm dealing with two basic circuits at the valve.

1-Coil (pins 1, 5) 8 Ohms, operated with 5v max I think.

2-Potentiometer (pins 1, 6, 4 ) 6 being the wiper.
Total resistance at pins 2, 4 should be 3k7 Ohms. 5v

THE key is that the ECU needs the pin 6 feedback voltage ranging from under 1v volt to 3v or more (the pot is fed with 5v)


I will be sooo grateful for any help designing this!

Feel free to have a look at my car mod's thread here,
http://www.renaultforums.co.uk/showthread.php?p=817333&posted=1#post817333

 

[Resqueline]

Deducting from the available data and using some "intuition" I'd say that the solenoid is PWM driven from 12V - with an averaged voltage of 0 to 5V.
I'm sure we can concoct a circuit good enough to fool the ECU. Is it a Turbo Diesel? Have you had any electronic suggestions at all so far?
The question is how high the PWM frequency is, and if the ECU senses the current draw or flyback effect from the solenoid coil.

Summing up the connections:

Solenoid:
Pin 1: Controlled Ground (0-5V less than B+)
Pin 5: Power Supply (on board voltage) / Battery Voltage (+12V)

Potentiometer:
Pin 2: Power Supply Voltage (+5V)
Pin 4: Ground
Pin 6: Wiper signal (1-3V)

 

[joseico90]

I appreciate you taking the time to help me, thank you

Apparently some people has had good results on other valves using these, don't know if it would be any good for my needs though.

http://www.z22se.co.uk/forum/viewtopic.php?f=86&t=5891&start=100

Please bear with me as my electronics knowledge is limited.

1- Yes it is turbo diesel.
2- No suggestions so far..
2-Yes I think it's PWM driven as I can feel the rate when I touch the valve.
3-Don't know the freq but it seems on the low side.
4- I don't know if the ECU senses the current draw or flyback effect from the solenoid coil.

I do know of people that managed to get the dash warning light off by just plugging a resistor between the coil pins. But I want to do a "proper" job

I've got OBD diagnostics equipment that allows me to see what's going on at ECU, & among other things, EGR valve data. See pics.

It will be a good help for testing the prototype circuit.

Thanks for your input,

I'm excited at this challenge!


Tony.

 

[markm6164]

Hi the actuator is usually driven with 12v pwm and has a lift sensor to measure the valve lift. I have removed many erg valves and DPF filters. The best way is to modify the ecu map. Not sure if I can post this but you will find your answer on the digital kaos forum in the garage section.

If mentioning another forum is not allowed the please edit my post.

 

[markm6164]

Also I can tell you how it works if you like?? The actuator has a gnd and 12 v pwm. The sensor part has a gnd and 5v reference and a signal return to the ecu. When ignition is off on pin 4 signal return you will get 0v. With ignition on you will get 0.6v which is valve closed. With valve fully open you will get 4.2v. Average opening at idle is a 65% duty cycle with around 3v on the signal return.

 

[joseico90]

Hi the actuator is usually driven with 12v pwm and has a lift sensor to measure the valve lift. I have removed many erg valves and DPF filters. The best way is to modify the ecu map. Not sure if I can post this but you will find your answer on the digital kaos forum in the garage section.

If mentioning another forum is not allowed the please edit my post.

It makes sense to modify the ECU map, I've read stuff about it, MAINLY on petrol engines, not so much on Diesels. It's obvious that most people that blank-remove EGR don't bother any further in dealing with the ECU side of things, maybe it gets complicated & expensive...and the car runs anyway..but I'm a perfectionist

I've also read on guys who have fitted a cheater and build a corrected base fuel table.
http://www.verboom.net/info/speedster/20070729/index.html
http://www.verboom.net/blog/index.html?single=20100426.0

Either way I need to FIND and consider my options,

 

[Resqueline]

Ok, I think a 1-transistor circuit should suffice.
The theory is:
At idle there's 0V across the coil (= 1k resistor), hence 12V at Pin 1; the transistor conducts and delivers 0V on Pin 6.
At 65% duty (is that voltage or effective-wise?) the transistor conducts only 35% of the time, averaging 3.25V on Pin 6.
Guessing the PWM works around 10Hz, I gave the output filter a 47ms time constant (or a 3.4Hz cutoff frequency).

It's perfectly ok to link to other forums when it's relevant to the issue at hand btw..

 

[joseico90]

Ok, I think a 1-transistor circuit should suffice.
The theory is:
At idle there's 0V across the coil (= 1k resistor), hence 12V at Pin 1; the transistor conducts and delivers 0V on Pin 6.
At 65% duty (is that voltage or effective-wise?) the transistor conducts only 35% of the time, averaging 3.25V on Pin 6.
Guessing the PWM works around 10Hz, I gave the output filter a 47ms time constant (or a 3.4Hz cutoff frequency).

It's perfectly ok to link to other forums when it's relevant to the issue at hand btw..

Looking great! hey thanks I really appreciate your effort

The ECU has something called "EGR adaptives" I think it has to do with it "adapting " to some degree to particular small differences in the valve, things like ageing, carbon bulid up etc, this valves can get messy and still won't trigger a dash warning.

Just saying this as the ECU might tolerate some variations, this "EGR adaptives" is in memory and can be reset, say when replacing the valve. or in this case when fitting the cheater.

I think I'm gonna order the components and get soldering ..

 

[markm6164]

Here is a typical EGR circuit

 

[joseico90]

BTW What transistor do I need?

Also looking at the data collected this morning on my way to work (see graph above)

We can see that the valve it's open most of the time, from idle up to 2000 RPM or when I accelerate sharply. When I coil measured in real time at idle and accelerating I got 4.5v and a bit more when accelerating, (I noticed the polarity showed as reversed on my dig multimeter, on both, coil and pin 6, why?)

Look at the the 3 top graphs, on the second down, we can see pin 6 voltage variations while driving, idle etc. It never drops below 1.1v and peaks at 1.88v it gives an indication on how much the valve is opening.

I can actually see this at the valve, real time with the bonnet open while stationary. The valve opens even at idle, maybe it has to do with current climate low temperatures?

Thanks

 

[joseico90]

Thanks Markm

Yep, I posted a link to this valve's detailed service info & specs, above on the first post.

At this point I don't have the skills to do a remap..I think..

It can be a risky thing to play with ..right?

don't wanna end up with a non runner..

could always save couple of hundred quid to get it remapped..

but I also like fiddling..

 

[markm6164]

Usually the ECU takes no reference from the actuator (pins 1 and 5) this is only an output from the ecu. When starting from cold you should have a closed valve from about 1-2mins then a 50% valve then a 75% open the around a 98% open valve then settle back down to around 65% when engine at normal operating temperature. During driving the valve will fluctuate between 50% and 90% depending on driving and engine load etc. When not accelerating and car is rolling valve will close. When clutch dipped the valve will open again. Anything above 2000rpm you don't want the egr valve open because the turbo kicks in you you need total clean air and fuel for power. The engine is contantly measuring pin 4 as its signal return. It uses the 5 volts as a reference and as the resistance changes in the lift sensor so the voltage back to the ecu changes. You cannot put 5v on the signal return because the ecu see's this as a fault. Probably around 4.8v is the max. Also you can't put 0v back as the ecu thinks you have an open circuit, hense the 0.6v when valve is fully closed.

As a side note here is how the egr valve works....

Most people think the egr valve just reburns exhaust gases to lower emissions although this is true its not how it works. The egr valve cools the engine!!! Supprised?? Well its true,, by returning exhaust gases into the cylinder reduces the amount of oxygen that can fill the cylinder. Less oxygen equals a cooler burn and a better burn. More of the mixture is burnt so the emissions are lowered. BUT anything above 200rpm is when we want power!!! More power equals and bigger explosion in the cylinder. How can we achieve this?? Add more oxygen to the mixture burns hotter so we don't want egr so we close the valve.

Knowing all this you have to produce a circuit that can stay off from cold for 1-2mins then gradually increase in voltage. Then above 2000rpm it has to produce 0.6 volts. Anything other than that then the ecu sill see a problem and the engine management light will be on and/or even go into limp home mode.

By rewriting the software in the ECU we take away the ECU's ability to look for the egr.

Hope all that helps.

 

[joseico90]

Hi again Mark,

Yep, I've read loads on EGR, all started when I had to do some maintenance on my engine and discovered the mess EGR had done in the inlet manifold. I decided to remove EGR from my engine there and then!


You say..

1- "Less oxygen equals a cooler burn and a better burn"

Well..maybe cooler.. but definitely not "better"


-EGR is typically not employed at high loads because it would reduce peak power output. This is because it reduces the intake charge density."

-For that reson EGR goes off when we turn on the air conditioning.

-EGR is also omitted at idle (low-speed, zero load) because it would cause unstable combustion, resulting in rough idle."

Since the EGR system recirculates a portion of exhaust gases, over time the valve can become clogged with carbon deposits that prevent it from operating properly. Clogged EGR valves can sometimes be cleaned, but replacement is necessary if the valve is faulty.

Personally I don't think EGR is nothing but poor engineering feat to meet emissions regulations.

To me it seems absurd to have our engines suck all those particles and nasties, messing oil acidity, potencially shortening bearing life, rings, liners, valves.
For goodness sake.. look at the mess they make on our inlet manifolds, especially on diesels.

Have you seen any off road, competition, etc vehicle with EGR?, nope, they take e'm off straight away if the engine had it to begin with.


oh well...my next car is gonna be electric anyway

you said..

" a circuit that can stay off from cold for 1-2mins then gradually increase in voltage"

I don't understand why you say that as the valve, which is what we're trying to emulate, doesn't behave in such way.
So the valve simulator wouldn't need to either. What you say is done by the ECU.

Cheers.

 

[jackorocko]

Well getting the circuit to wait 1-2 mins before activating is not very hard either. I wouldn't let that slow you down.

 

[joseico90]

Well getting the circuit to wait 1-2 mins before activating is not very hard either. I wouldn't let that slow you down.

Don't think there is any need as the valve doesn't do that. The simulator wouldn't need to either.

The valve simply opens-closes controlled by the ECU and he is the one making the calculations to do so.

At same time the ECU receives feedback from the valve's position sensor (pot) , simple.

 

[joseico90]

Ok, I think a 1-transistor circuit should suffice.
The theory is:
At idle there's 0V across the coil (= 1k resistor), hence 12V at Pin 1; the transistor conducts and delivers 0V on Pin 6.
At 65% duty (is that voltage or effective-wise?) the transistor conducts only 35% of the time, averaging 3.25V on Pin 6.
Guessing the PWM works around 10Hz, I gave the output filter a 47ms time constant (or a 3.4Hz cutoff frequency).

It's perfectly ok to link to other forums when it's relevant to the issue at hand btw..

What transistor do I need to build this?

-I get around 4.5v at coil even at idle, measured yesterday, I need to find out why that is. The valve opens even at idle. Maybe it's a dodgy valve but it's pretty new.

Maybe the ECU has to force it to open to reach the wiper 1.1v min expected at ECU?

That being the case its no biggy, the simulator will hopefully behave the right way.

And, again, if that's the case, it goes to show how the ECU adapts to the situation.


Thanks.

 

[Resqueline]

If the ECU interprets both 0V & 5V as a fault then you'll have to include a couple of diodes to narrow the voltage swing by 0.6V either way. Use 1N4148's.
The transistor can be any (universal/ general purpose) small-signal NPN; BC546-BC550, 2N2222, 2N3904, 2N4401, 2SC945, etc.
Get a 47k resistor and a 1µF capacitor too while you're at it, just in case better output filtering is needed.

When you measure 4.5V at the solenoid, is that across the coil, or at pin 6 with reference to engine ground? What was the thing about reverse polarity?
I'd expect the mechanical travel to be substantially narrower than the electrical travel of the potentiometer. TPS pot's behave this way too.
The ECU surely tracks the min & max values encountered and calculates the % valve opening between these two extremes.
I notice a discrepancy between the EGR document you first linked to (P1 = +12V, P5 = Sw. GND) and the one you posted last (P1 = 12V PWM, P5 = GND).
You may want to measure the voltages on those pins (relative to engine ground) to verify what it really is like.

My reference to idle in my theory was only loosely meant, call it "EGR at rest" if you want.

I wouldn't worry about needing to remap a Diesel since they are not throttled. Gas engines might've needed both fuel & spark remaps though.

My V8 has spark retard applied during warmup to make the combustion process more inefficient - to ensure a faster warmup.
Maybe EGR applied to a (cold) Diesel accomplishes the same thing. EGR was "invented" to reduce NOx (smog).
I'm thinking of using my (inoperative) EGR system to control a water injection system instead, to reduce both carbon buildup, NOx & pinging.

 

[markm6164]

I'm very interested to see how this works out. Keep us updated

 

[joseico90]

Thank you.

If the ECU interprets both 0V & 5V as a fault then you'll have to include a couple of diodes to narrow the voltage swing by 0.6V either way. Use 1N4148's.
The transistor can be any (universal/ general purpose) small-signal NPN; BC546-BC550, 2N2222, 2N3904, 2N4401, 2SC945, etc.
Get a 47k resistor and a 1µF capacitor too while you're at it, just in case better output filtering is needed.

When you measure 4.5V at the solenoid, is that across the coil, or at pin 6 with reference to engine ground? What was the thing about reverse polarity?
I'd expect the mechanical travel to be substantially narrower than the electrical travel of the potentiometer. TPS pot's behave this way too.
The ECU surely tracks the min & max values encountered and calculates the % valve opening between these two extremes.
I notice a discrepancy between the EGR document you first linked to (P1 = +12V, P5 = Sw. GND) and the one you posted last (P1 = 12V PWM, P5 = GND).
You may want to measure the voltages on those pins (relative to engine ground) to verify what it really is like.

My reference to idle in my theory was only loosely meant, call it "EGR at rest" if you want.

I wouldn't worry about needing to remap a Diesel since they are not throttled. Gas engines might've needed both fuel & spark remaps though.

My V8 has spark retard applied during warmup to make the combustion process more inefficient - to ensure a faster warmup.
Maybe EGR applied to a (cold) Diesel accomplishes the same thing. EGR was "invented" to reduce NOx (smog).
I'm thinking of using my (inoperative) EGR system to control a water injection system instead, to reduce both carbon buildup, NOx & pinging.

1-Got some 1N4001 are they any good?

2-" and the one you posted last (P1 = 12V PWM, P5 = GND)"

I think that was posted markm, not me

3- Yep, across the coil, i noticed that my multimeter showed the - minus sign (Inverted) polarity while measuring coil and pin 6 (if reversing probes then it would dissapear, tried with a 1.5 dc battery to see if the tester was wrong, but worked normal, puzzled.

4-Not ruling out a remap or chip tune in future, but mainly to gain BHP's and a better engine response. I had id done in the past to another car of mine..best thing I ever did

After all my engine is the 80 ish BHP version, apparently the only difference beteewn mine and the 100+BPH version of these 1.5dci's is ECU map.

5-Water injection system? elaborate please..

Again, thank you so much..

I'm a bit unwell at the moment but will be starting the project as sooon as,

 

[joseico90]

I'm very interested to see how this works out. Keep us updated

Will do