Using a Cat With a Non-Cat ECU

Please add any links you have found that are useful to the group. Please submit why this link is helpful and a short description.

Moderators: GRNSHRK, ron, bfons

Post Reply
ScottAndrews

Using a Cat With a Non-Cat ECU

Post by ScottAndrews »

At long last, here is the circuit diagram for a box to hook up a Euro ECU to operate closed loop with a catalytic converter. This circuit has been termed the "Andrews Box" by some of the BGC folks. It functions differently from the Valentine and Johnson boxes, but it essentailly performes the same feat..allowing you to run closed loop with a catalytic converter and O2 sensor while using a non-cat equipped ECU (such as those on most Euro cars).

As many of you may be aware, to operate effectively, the cat needs exhaust that is produced from combusting a very precise air fuel ratio (AFR). I have posted thre scans from Bosch Fuel Injection and Engine Management by Probst (published by Bentley-if you want to understand your EFI system, buy this book) that shows the operation of the cat and the O2 sensor. As you can see, if the AFR is just slightly rich (relative to 14.72) the CO and HC rise dramatically, and if it is slightly lean the NOx rises even more dramatically.

The closed loop concept is rather simple. the O2 sensor has a very sharp voltage vs AFR characteristic (as can be seen in the picture). If the AFR is even slightly below 14.72 (which means more gas, hence, rich) the O2 sensor output is up around .8 volts; If it is slightly higher than 14.72, then the output is about .2 volts or less. In practice then the O2 sensor is like a switch that signals when the AFR crosses the magical 14.72 point. If you look at the CAT operation pictures, you can see that the cat needs to have exhaust from combustion right around 14.72 to then re-combust these products into less noxious gases.

The circuit I developed works as follows:

In the Bosch AFM circuit, the AFM temp sensor is a thermistor. It has higher resistance at lower temperatures. The ECU reads this higher resistance as colder (denser) air. Denser air needs more gas (since there is more of it per unit volume). So the ECU widens the injector pulse to add more gas. Higher temeratures imply fewer air molecules, and the loer resistance produces a lower voltage, so the ECU narrows the injector pulse width and the mixture leans out.

In the circuit, the O2 sensor output is used as a signal that operates a solid state switch. The circuit compares the O2 output to a voltage level (about .45 volts) obtained by a simple divider (A real purist would make a thermally stabilized voltage source, but I didn't feel that was all that necessary since the box is in the glove compartment, and the car interior temp is generally pretty stable). When the signal is above this voltage, the comparator changes state. The comparator output then drives the switch control. The switch is connected to switch resistance values in series with the AFM air temperature signal line. I switch in higher resistance when the O2 sensor output is low, which then drives the ECU to make the mixture richer, and lower resistance when the O2 sensor output is high, which drives the ECU to make the mixture leaner. I use potentiometers (variable resistors) so I can tune the system.

In operatoin the system oscillates between rich and lean. THisd is due to the transport delay in the engine. There is a finite amount of time from when the injector puls changes to when the exhaust at the O2 sensor reflects that change. THis means the system is always "hunting" back and forth across the 14.72 AFR point. It is this hunting that allows us to maintain a very stable and accurate AVERAGE AFR of 14.72.

Some of you will question adding resistance to the nominal operating point. In the lean state, the system will still have more resistance than the stock setup. This is true, but it is pretty well known that these cars tend to run lean, and many people find that adding in about 2K to 5K ohms resistance helps the engine a lot; so I took advantage of that observation, and add in some resistance at the low end, and simply add a lot more at the high end.

I included a temperature cut-out in the circuit, to basically shut off the closed loop operation when the car is cold, but I have found this is not really that important.

I use a Jaycar AFR display to show the AFR in real time. In operation the system takes about 1-2 minutes to warm up, at which time it oscillates back and forth from the low end of the display to the high end. At idle this is about a 1 second period, and it is a little ragged, as the engine speeds up the period speeds up (since there is a faster flow through the engine) and the oscillations become very regular. You can tune the threshold voltage to "center" the oscillations, and you can tune the high and low resistance values to adjust the extent of the oscillations. These three variables are inter-dependent, so in practice you have to tweak it a few times to get it balanced properly, after which you can forget about it.

Feel free to email me with any questions.

S
Attachments
ECU Mod3.gif
O2Sensor.jpg
NonCat_Cat_Exhaust.jpg
User avatar
Brucey
6 Series Guru
6 Series Guru
Posts: 10077
Joined: Fri Apr 09, 2004 7:17 am
Location: Cambridge, UK

Post by Brucey »

This seems like a really good tweak for the engine management system on l-jet cars. I wonder if it will work with open-loop motronic, too?

I wonder where the best place to mount the lambda sensor is?

cheers
~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
ScottAndrews

Post by ScottAndrews »

It is pretty cool. If I were not running a Cat (whihc needs really precise AFR control), I would use a "braodband" O2 sensor (it has a slightly softer transition), and then make a continuous control llo, so I could adjust the mix for optimum power (about 12:1).

S
Marquis Rex

Post by Marquis Rex »

If you're not running a Cat the correct thing to do is to run as lean as the engine is able to withstand- near the lean misfire limit- say a Cov of IMEP of 2.5 % - at part load operation and then run about 13.3:1 AFR at lower engine speeds going up to 12-12.5 :1 at higher engine speeds.


LBT or the leanest mixture for best torque isn't a fixed value but rather changes a little depending on engine speed.

I don't have a direct feel for how lean the M88 combustion system is able to run but lambda 1.2 or about 17:1 seems likely. Don't forget that the M88 engine with port throttles is likely to be running alot less in cylinder residuals than conventional engines and this should allow leaner running before the onset of misfire or unstable combustion.
Cogeniac
Posts: 79
Joined: Mon Nov 27, 2023 4:25 am
Location: Petaluma,CA
Contact:

Re: Using a Cat With a Non-Cat ECU

Post by Cogeniac »

I see the links are now all dead.
I have recently updated this circuit,and will publish a revised FAQ soon. I am planning to fab this on a PCB, so those who are interested can contact me for the details.

Scott
Scott Andrews
Petaluma, CA
1980 Euro 635, "Rocinante", the Truth.
1972 3.0 CS
Image
plip1953
Posts: 400
Joined: Mon Aug 13, 2018 4:15 pm
Location: Leeds

Re: Using a Cat With a Non-Cat ECU

Post by plip1953 »

That was a really interesting read.

Probably a dumb question, but what would be the point of retro-fitting fitting a cat? Is it because it is a legal requirement to do so in some US States or would be mostly be because it's a better thing for the overall well being of the planet?
Phil
__________________________

1988 Highline 635 CSi
Jubilados
Posts: 161
Joined: Sun Mar 08, 2020 4:48 am
Location: So Cal

Re: Using a Cat With a Non-Cat ECU

Post by Jubilados »

Cogeniac wrote: Thu Jan 11, 2024 5:37 am I see the links are now all dead.
I have recently updated this circuit,and will publish a revised FAQ soon. I am planning to fab this on a PCB, so those who are interested can contact me for the details.

Scott
Scott, glad to see you’re back on this!
I’ve sent you an email, I’m still interested in this.
Cogeniac
Posts: 79
Joined: Mon Nov 27, 2023 4:25 am
Location: Petaluma,CA
Contact:

Re: Using a Cat With a Non-Cat ECU

Post by Cogeniac »

plip1953 wrote: Thu Jan 11, 2024 7:19 am That was a really interesting read.

Probably a dumb question, but what would be the point of retro-fitting fitting a cat? Is it because it is a legal requirement to do so in some US States or would be mostly be because it's a better thing for the overall well being of the planet?
It is a requirement for many US states.
In CA any car built after 1976 must pass an emissions test every 2 years. Sort of like the UK MOT test, but specific to emissions.
Scott Andrews
Petaluma, CA
1980 Euro 635, "Rocinante", the Truth.
1972 3.0 CS
Image
plip1953
Posts: 400
Joined: Mon Aug 13, 2018 4:15 pm
Location: Leeds

Re: Using a Cat With a Non-Cat ECU

Post by plip1953 »

Cogeniac wrote: Sun Jan 14, 2024 4:54 am
plip1953 wrote: Thu Jan 11, 2024 7:19 am That was a really interesting read.

Probably a dumb question, but what would be the point of retro-fitting fitting a cat? Is it because it is a legal requirement to do so in some US States or would be mostly be because it's a better thing for the overall well being of the planet?
It is a requirement for many US states.
In CA any car built after 1976 must pass an emissions test every 2 years. Sort of like the UK MOT test, but specific to emissions.
Understood. In the UK there are also emissions tests as part of the annual MOT test, but no obligation to retrofit a cat on cars first registered before around 1992. And I imagine the emission requirements for the test for the non-cat cars is therefore less stringent.
Phil
__________________________

1988 Highline 635 CSi
Cogeniac
Posts: 79
Joined: Mon Nov 27, 2023 4:25 am
Location: Petaluma,CA
Contact:

Re: Using a Cat With a Non-Cat ECU

Post by Cogeniac »

plip1953 wrote: Sun Jan 14, 2024 7:35 am
Understood. In the UK there are also emissions tests as part of the annual MOT test, but no obligation to retrofit a cat on cars first registered before around 1992. And I imagine the emission requirements for the test for the non-cat cars is therefore less stringent.
In the US the Federal US DOT regulates new cars through the manufacturers (emissions, safety, etc). The states manage all other motor vehicle realted stuff (licensing, emissions and safety checks, etc. ). So, what regulations apply depends lot on what state you register the car in.

CA requires an emissions test every 2 years for any car built after 1976, and over 5 years old (presumably new cars will pass the test, so they don't require testing until the car is some number of years old - 5 , I think)

CA used to have a rolling smog date, so, any car older than 25 years or so was exempt, but then around 2000 they changed it to the fixed 1976 date. Seems pretty silly because any car older than 25 years is, at this point probably either in a junkyard, or is in the hands of a collector. Collector cars generally do not get driven very far each year, so even if they are not emissions equipped, the total output of emissions is probably less than a modern vehicle driven 50K miles per year. If you figure a collector car gets driven, say 500 miles per year (that is my insurance limit), then to equal the emissions of a modern car driven 50K miles it woudl need to emit 100X as much..ANy reasonably well running engine probably isn't that bad.
Scott Andrews
Petaluma, CA
1980 Euro 635, "Rocinante", the Truth.
1972 3.0 CS
Image
User avatar
86_6series
Platinum Member
Platinum Member
Posts: 1069
Joined: Mon Apr 02, 2007 10:08 pm
Location: West Nyack, NY

Re: Using a Cat With a Non-Cat ECU

Post by 86_6series »

Cogeniac,

What collector car insurance only gives you 500 miles per year?

Bob V
86-635CSi
96-MB C220 SOLD
98-740i
71-Continental MK3
bpoliakoff
Platinum Member
Platinum Member
Posts: 3312
Joined: Thu Sep 28, 2006 9:28 pm
Location: scottsdale, arizona
Contact:

Re: Using a Cat With a Non-Cat ECU

Post by bpoliakoff »

86_6series wrote: Fri Feb 16, 2024 11:22 am Cogeniac, What collector car insurance only gives you 500 miles per year?

Bob V
[/quote

Grundy unlimited mileage? But IIRC at least 10K but for pleasure or events only, no shopping Full coverage on my 6 including an agreed upon value of $35K is about $275 annually Grundy has been around for decades and the policy is underwritten by a large S&L in Philly
81 Euro 635 M90 with motronic. Currently under a complete nut and bolt restoration. Pictures at
flickr.com/photos/bertsphotos.
Post Reply