Accurate instrumentation possible on a diesel?
 

In another thread, rmay635703 said:

I'd like to ask diesel vehicle owners what kind of instrumentation they're using, given this. My old 2007 Jetta currently has an UltraGauge and I'm a big fan. It would be a shame if I couldn't use it on my new 1999 F250.

I'm using MPGuino on Berta, but it has FE signal comming out of ECU, so it was easy to make it work.

Electronics made the things easier regarding fuel-efficiency instrumentation in Diesels, but with an old-school one it's still a challenge...

True. Flow meters are expensive, and those cheap ones sometimes are just not good enough to measure FE accurately.

The 2014 Cruze Diesel has the same MPG display as its gasoline counterpart...and it just counts the number (and thus volume) of injection squirts (the diesel uses more than one) to estimate diesel consumption.

1999 isn't that old, is it? The truck does have an OBDII port.

The promotional material for the ScanGauge II indicates that it supports diesels. Is that just wrong? Or is there something about my truck that's unsupported. Still confused here...

I have a ScanGauge II on my 2000 Jetta TDI and it does a reasonably good job of estimating the fuel consumption. Not perfect, but good enough to know how well I did on a commute or an errand. I normally run B20, and when I change to pure dino-diesel, the SG will get confused and will be off by a several MPGs. You can check out my fuel log; I've started to include the SG results with each fillup.

If it's compliant to OBD-2 your truck might be supported.

In a Mercedes OM 603 engine, there is a EDS control unit that is supposed to control the amount of diesel to be inyected by injection pump.

I have EDS scheme diagram. One should take advantage of it and construct a math model that calculates diesel fuel used, and together with speed, calculate mileage.

So why don't you share the diagram? ;-)

If it was compliant with OBD-2, might work with MPGuino and other similar devices :thumbup:

I use a scangauge II. It becomes way off (from 2 to 15 mpg wrong over a tank) if I implement my EGR delete, go on a road trip, change biodiesel concentration, or make any significant change in driving technique.

Fortunately, I can calibrate it for B100, no EGR, and with my driving techniques fairly settled it is a useful tool for monitoring my daily commute over a tank, including some non-commuting driving over the weekends. I can make seasonable adjustments to maintain reasonable accuracy from spring to fall with B100 fuel.

Yeah, it'd be nice to have real fuel flow measurement.

Mercedes diesel Injection Diagram OM 603
 

Hello guys,

The Mercedes 300D turbo of 1993, on a W124 chassis, has the follwing Electronic Diesel Injection Diagram:
http://www.flickr.com/photos/oldbeaver/8982381950/

As you can see, there are six electric / electronic incoming signals to the injection control unit.

Can we take advantage of them to make a real time mileage monitor?

OldBeaver

Can't see the diagram :(

Can´t see the Diesel Injection Diagram
 

I couldn´t see it also, at first.

I found the way to allow it to be seen: see it below:

OldBeaver

Intercepting the signals in an old Mercedes Diesel
 

AndrzejM, OldBeaver (myself) and others are long time seeking how to use MPGuino in an old diesel, like a Mercedes from 1987 to 1995.

One of the signals to be used is VSS. One step forward would be to find an easy point in the car where one can intercept this signal in the cleanest and safest posible way.

As we talked with AndrzejM, the other signals has to be taken from the injection system. The question is how to get them easily, to conduct them to the cabin to be processed.

Maybe the test terminals may be useful for this purpose. At least, rpm can be taken there.

Does anyone know where to intercept the effector signals from the ECU, for example? I am talking about the physical spot in the car and the practical electronic way to do it. Not just the place in the EDS diagram.

OldBeaver

I think the best way would be to hook up to the back side of ECU plug or plug connected to the injection pump. It will be easier to find the right cable connected to the second place. :) Fewer connections, that's for sure :)

EDS signals spot
 

I agree. The ECU harness is more complicated. And I would prefer not to touch it.

EDS Inputs to this system are:

Voltage supply via OVP relay (K1) - overload protection relay I suppose
Engine speed via speed sensor (L3) - rpm?
Coolant temperature via sensor (B11/4)
IP control rod travel via sensor (L7) - rack position?
Air quantity via AFM (B2/1)
Manifold (boost) pressure via OVP relay (K1) - useful for real time fuel estimation?
Barometric pressure via sensor (B18)
Base circuit resistance via reference resistor (R18/2)
Resistance trim via ELR knob (R18/1) - regulable fix resistance I think

Outputs are:
Variable control voltage to the ELR actuator (Y22)
Voltage to VAC transducer (Y31) and on to the ARF valve (Y31/1)
Connection to test connector diagnostic block

Basicaly, the fuel injected is governed vía ELR actuator in the back side of the injection pump, where more voltage means more fuel and less voltage means less fuel. Very useful information.

To me is very clear that ELR controls fuel injected. You want to combine it with the rack position sensor signal. Question: Is it L7?

Second, I suppose this voltage controls VAC transducer (Y31) which is the effector for .... what? The turbo somehow?

I will try lo locate all this elements in the engine compartment next to the injection pump after I recover my car, hopefully tomorrow.

Comments, enlightments?

Oldbeaver

Your pump is controlled by two factors:
1. gas pedal linkage - it's setting up rack position - thus main fuel dosage
2. ELR signal - it's changing by some percentage fuel dose to meet value counted by ECU, based on all the sensors

So you have to know rack position adt its offset by ELR signal, then you have the exact fuel dose. Combining that with rpm signal you have the exact moments when you have to measure rack position and offset. Lets call that a trigger signal. And finally combining that with VSS signal will give you FE.

And the other question about Y31 - I presume that's a valve for EGR

Processing signals for accurate diesel instrumentation
 

Ok, very good. Lets try to get the signals.

Rack position: there is a plug on the back of the governor of the injection pump with three wires. One of these should have the signal we want.

I took a picture of it but forgot how to insert here.

ELR signal: I am searching for it but won´t be difficult.

RPM signal is available in test socket. In this same socket there is an "diesel injection" signal, which I don´t know what is measuring, maybe just that the EDS is working ok.

You said to use RPM as the trigger signal as for when measuring the others, isn´t it?

Wouldn't it be possible to do the following, as an alternative:

Measure the rack position modulated by the ELR signal (surely some of the three wires has it already), and multiply it by the rpm value, for having a measure of how much fuel is being injected?

It would be a analog signal, I think.

I think it as follows: rack position tells how wide open is the gate that governs the passing of fuel (liters/stroke), while RPM tells at how much strokes per second it working the injection pump (strokes/sec).

Multipliying the first times the second should give liters/sec of injection.

Of course there are some constants to calibrate, as the pump may rotate at twice the speed of the engine, and we don´t know how much fuel per stroke each injector injects. But I think I have that info somewhere.

However, the liters/stroke each injector injects in the cylinder is not even at every rpm. It has a range that depends, again, on the RPM.

So, lets try to model that :

Iq = ((a + b*Rack * rpm)) + c*rpm*rpm

where:

Iq = inyected quantity (lts/sec)
a = callibration constant to be corrected with the full tank mileage
b = max injected volumen per stroke specific for the BOSCH pump
c = small value, probable negative, that indicates a limit vs RPM
Rack = rack position already modulated

How do you like it?

My intuition tells me it maybe less accurate, but simpler in the sense I think I can do it. While I don´t see how to materialize yr method. Maybe if you explain a Little more?


Best,

OldBeaver


VSS signal: I will see in which place I will get it.

Jose,

You can't measure constantly any values here so you need a trigger to measure fuel dose at the injection pump. That's why you need rpm signal. It's a trigger for the moment that you should look into other signals from the IP.

http://ecomodder.com/forum/attachmen...1&d=1373874666
That's an example of fuel dose. Rack position is combined with an ELR offset and that gives you exact fuel dosage. Measuring those two values at the moment of trigger signal (rpm) gives you the amount of fuel per one rpm. Changing this to time of injection pulse gives you proper signal for MPGuino.

So you don't need any pump specific volumes. You just need to calibrate it properly and voila!

RPM as a trigger on diesel fuel measure
 

Dear AndrzejM,

I try hard to understand and follow yr thoughts, but I cannot do it fully.

RPM is also a signal that is present at every moment. When shall we consider that it is triggering the measure of Rack + ELR? Are you talking of sampling signal values?

Let me think: we have one injection per cylinder every two turns of the engine, right? Equals 0,5 injections per turn per cylinder. If so, as we have 6 cylinders, then we have: 0,5 * 6 = 3 injections per turn of the engine.

So you want to sample (measure) rack position 3 times per revolution, and just in the moment injection is produced? Is that the idea? And how do you do that?

I am very slow today but I perceive you have a very clear idea in mind, but not how it Works. But my interest increase any minute.

Thank you for the graphic and the explanation, but I still don´t follow you.
I lack of knowledge for it. Please. try to express it even simpler, if posible.

Looking forward for it,

Old Beaver

Hi Jose,

Sorry for late response (I've seen your e-mail), but I was on a holidays with no internet access. Anyway I'll take a look at your document and get back to you ASAP.

I was thinking about this issue during my free time and I think I have some answers.

Cheers

Instrumentation for a diesel.
 

Thank you! I look forward to it.

Scangauge supports some diesels.
 

Yes, Scangauge is supporting some diesels. But they are only OBDII compliant diesels. Don´t know if ANY OBDII compliant diesels.

We are trying to develop instrumentation for old diesels.

Oldbeaver

Wait a minute. Isn't your Skoda OBD2-compliant?

Isn´t yr Skoda Diesel OBDII compliant?
 

Ha, ha, ha.... I surely said it not very clear. Of course Skoda 2009 is OBDII compliant and it has a Factory installed very powerful trip computer and diag socket. There is even a special diag software for all VAG group.

The fact is that I am working for my 1993 MBZ turbo diesel, 6L, engine OM603 on a W124 chassis. And that old car isn´t OBDII compliant.

There are many as it, many other diesel cars and jeeps and trucks, in the same situation. If they had a trip computer, they would have been able to test any mod and improvement in real time mpg, which is the best overall indicator . So developing one would be very interesting.

Best,

OldBeaver

Duino for processing signals before MPGuino.
 

Dear AndrzejM,

On our email conversations, we come to the fact that it is possible to get the necessary signals from a 1993 MBZ diesel turbo, non OBDII compliant, for feeding a MPGuino trip computer.

In brief:

VSS signal is available in Mercedes and possibly in any old diesel. Or can be taken from a GPS.

Fuel injection signal will be obteined from Governor + ECU and rpm.

Sampling will be triggered by revolution sensor.

To make a signal that is suitable for MPGuino, we need to process it a little bit, before input it to that board.

Y have a old MPguino, very well assembled, with English units which carries a ATMega168 processor. I tried to load to it the Canadian version of the software but it has not enough memory for that.

Due to that, I bought a ATMega 368 processor thinking in swapping it in the MPGuino, having this way a lot more memory, which we eventually may use for the additional code needed.

I also have a programmer which carries a ATMega368 also.

And I have a FreeDuino board, however, with no parts, empty. All bought in the US by mail a couple of years ago.

Bellow I attached small pictures of all these parts.

What is the best way to accomplish our objectives, from the hardware point of view?

Oldbeaver

Jose,
For the beginning you don't have to change MPGuino to get metric code working or to upgrade any code. First we need to get it work just for a proof of concept.
For conversion unit I think the best would be to use ATtiny. It's small, energy efficient and it can be fitted anywhere.
Then you can try to rebuild your MPGuino to get any code you want working and with conversion unit attached.

Using MPGuino for diesel instrumentation and metric conversion
 

Dear Andrzej,

Ok to all. I realized that it is just a constant to be changed, so no need to modify code. Of course!

Clear.

I sent images of the MPguino I have, and of a kind of a TinyGuino I already have. Did you see them?

I need to remember from where in the US I got it to get the parts. It came with some parts included, maybe it use only these.

Do you refer to this Arduido specimen or to another one?

My car is at the garage by now, for a Little stretching of chassis.

Best,

Oldbeaver

Yes I saw those pics. You can use that TinyGuino for a conversion unit, of course. But ATTiny is a single chip that can do the same.

ATtiny chip for simple conversions
 

Do you mean something like this?:

https://encrypted-tbn2.gstatic.com/i...wKcSZ8mZH2rlK3

When I searched for ATtiny microchip, there showed several versions, small and bigger. For what I can see, you can solder it in any exp board and, yes, put it everywhere.

Two questions:

First: normally Arduino boards and these small chips use 5VDC or even less. How do we get that voltage from the 12 to 14 VDC of the car? I used a voltage divisor or such, with two resistors.

Second: how do we take the analog signals of the car (normally 0-12V) to input them in MPguino and ATmini without a) burn the microprocessor, b) alter the car parameters?

I am weak in electronic knowledge, just able to solder kits and circuits made by others and that is it.

Oldbeaver

Jose,
Take a look here: MPGuino - EcoModder

You'll find answer to both of your questions. First answer is 7805 voltage stabilizer, second answer is 50k resistor and 5,1V Zener diode. But that's for measuring pulses. For analog signals you need a voltage divider made of two resistors.

And regarding ATTiny there are many versions of the chip and we'll find the right one for the task.

Customizing instrumentation for a diesel .
 

Ok, so we will customize an instrumentation solution for a 1993 Mercedes Diesel, with the help of MPGuino and some versión of ATTiny. If we succeed, at least many Mercedes diesel cars may benefit of it, starting with mine. Excellent! Thank you for sharing yr knowledge.

OldBeaver

Mercedes Diesel ready to start the project
 

Dear AndrzejM and all Forum guys,

At last, I have my car back home from the garaje, after some painting, radiator repair, change of cooling fan, some body work, passenger seat command repair and several other parts replacements, such as thermostat, wáter pump, and many others.

Step forward:

First, I think the first to do is to identifiy the necessary signals spots to get them, and have them available for use in the Project. Any ideas on a good place and holder of the signal wires ends, hardware to use, and such? I can deliver the signals to inside the cabin or leave them under the Bonnet.

Second, construct a small board where I can solder the zener, the voltaje regulator, the voltaje dividers, and so on. Maybe some led maybe useful to see if there are signals coming. Leave space for the ATtiny. I think it may be soldered here.

Third, have some kind of signal sensor to check before input signals to MPGuino.

Fourth: Install MPGuino, power it and get signals to input it. Callibrate in the car.

I am almost sure I saw some small circuits as to test signals. Not necessary an Oscilloscope.

Andrzej: please give some advice about the hardware to buy, and the circuit scheme, to start building.

Attached are some pictures of the sensors where I plan to get the signals for the Project. There is a picture of the car also.

Looking forward to yr comments and advice,

OldBeaver

Jose,

I think that the best way is to get all signals from ECU harness.
Regarding hardware that you need to buy, first of all you need zener diodes, resistors and working MPGuino. And it doesn't matter which version of the code you have.

Definitely scope would be handy to see what's going on between ECU and injection pump.

Connecting to diesel injection signals.
 

Ok, I will locate the ECU, make the connections and borrow a scope from someone to start testing.

First, I need to locate diesel ECU,

I think I have a working MPGuino which I may test connecting VSS signal and rpm signal, besides the power and zener diodes. Anyway, the MPGuino I have has its own zeners, and wants 12V input, from what I recall. This will be ok for testing.

I need to prepare a cable and plug to bring signals to MPGuino.

Will take pictures to share with the Forum.

OldBeaver

You're right... You don't need zeners and resistors for MPGuino... I forgot, because I have it build on my own from Arduino Uno. As soon as you locate all signals we need we'll think of converting signals to pulses compatible with MPGuino.

Getting signals
 

I think I will not touch the ECU harness at all. Instead, I will get the signals from several harnesses that are almost all identified.

Now I am trying to identify the precise wire based on the color. However, the color code abbreviations I have do not match the ones in the circuits. The ones I have are:

BK black

BR Brown

BU Blue

IV Ivory

GN Green

GY Gray

TR Neutral transparent

OR Orange

PK Pink

RD Red

VI Violet

WT White

YL Yellow

Please see the circuit of the Mercedes diesel injection systems I am sending by email due to its size, and you will realize they are using different color abbreviations.

If you can help me with that it will be nice.

Last minute: these are the German abbreviations for collors, as was pointed to me by AndrzejM. American Mercedes Manuals use a three letter abbreviations, well known for anyone.

Oldbeaver