Electronic timing

[Hey]

I am thinking to do a electronic controled timing(first version will only read timing at the injector) for idi diesels. But I need your opinion.

Do you think it would be necessary to know how deep the injector nozzel travel inside the injector or only when it does. I mean do you think a binary system on/off would be sufficient  or would it be important to know how deep it goes.

At first I was thinking that a binary system could work just fine but.. it is for my 1.9Td and it has 155/200bar injectors. As I don't fully understant how the dual stage injector work I was thinking you could help me find out.

So anybody that have any information about how to mesure dual stage injection, how to test it with a pop tester, how it works (images are welcome) or their opinion about a binary system... I am open.

If it works... I am going to make a DIY to explain you how I proceeded with programming codes, etc.

[Darkness_is_spreading]

First off the the injectors break at a certain pressure when the injector forces it too based on the timming of the injector pump. This functions like a gas injector except the process of injection is based on a pressurized fuel line and the injector opens up to a certain limit based on the comp.

The injectors are fully mechanical, there is no reall way to measure the tip of the nozzle, but you may be able to tap into the line before the injector from the pump to get a flow rate and thenfrom there calculate the fuel input.

As for  programming, you would receive a pattern of numbers that would have a higher or lower frequency based on the speed of the motor. The data would be useless because the pump that is on the motor is non electronic, other than the stop solenoid.  

I really can't think of a use for this, maybe for a pump pressure check ?

Also for the programming bit, you would need to have a device to store and run this program, these can be quite expensive.

 I deal with the axiom micro controllers, but you would have to have an LCD output to display any results, it also runs off a 5 v battery, and has very little power draw.

[jtanguay]

I think its easier said than done, like with most things.  For sure the IDI diesel can be transformed and made to mimmick electronic style engines.  You would first need some sensors to monitor inputs, outputs, etc.  Lots of data would need to be collected, and probably take a good few months just to get something stable.

If you were going to monitor the amount of fuel being used, I would recommend two sensors for each injector, which would have to be extremely precise.  One on the injector line, and one on the return line.  For the amount of fuel used, all you would need to calculate is the fuel going in, minus the fuel going out.  Match that with an rpm sensor and presto.  

If you want to get real complicated you could get an AIT sensor to monitor air temp as not to burn too much fuel if its real hot (smoke less)  and even an oxygen sensor, to know if the car can handle extra fuel... or less...  

How will you be controlling the timing electronically though?  some sort of actuator?  This would seem like it would cost a whole lot of dough

[RabbitJockey]

slap on a tdi pump?  i wonder how they would work, tdi pump on an idi, and yeah i know that there are a alot of differences than just the pump.  but i wonder how our cars would run with with an aaz head and a tdi pump and electronics.

[Hey]

The injectors are fully mechanical, there is no reall way to measure the tip of the nozzle, but you may be able to tap into the line before the injector from the pump to get a flow rate and thenfrom there calculate the fuel input

I know an injector is made.. it is the dual stage that I don't fully understant the consequences and the way it work inside

THERE IS a way... the injector are not pressurize at all. I mean the return lines are not pressurize as much as is the nozzel. So a piezo can be mounted inside the injectector with a little spring attached to the nozzel pin movement (to injector's spring or something like that). When the tip moves... it applies a force and I know it (binary system). I could aslo calibrate that force to have the travel of the nozzel=more complicated.


For the injection pump... NO need of a TDI pump.

Here is how I am going to do that:

Linear Hall effect detector 10$ attached to the flywheel at TDC... each time the flywheel pass to TDc, I know it. I also know the time that pass between each revolution, thus the RPM.

With RPM and time, a little of physics gives me the position of every piston at any time (no need of a tooted magnetic flywheel...it's for people that don't do physics). If another part gives me the exact time when the injector pin moves(piezo few dollars)... I get timing. It's electronic so it's precise to us which is more than enough.

For those who said it's hard. I only need 4 parameters. Frequency, T1(time passed to TDc), T2(injection begin), T3(injection stop). Nothing too complicated for a physician  :wink:  

Lots of data would need to be collected

  :roll:    


That would be a great tool for those you take a 1.6td pump with a 10-12mm head and put it back on a 1.9td. It is also a great tool to adjust the regulator axe (timing VS load) and timing in general.

Also for the programming bit, you would need to have a device to store and run this program, these can be quite expensive.

I will built it myself!!  I am working on it.... and I have 30% done. Put a video and a PC power supply apart and it is still 0$ right now!!  :lol:

[Hey]

If you were going to monitor the amount of fuel being used

I don't care how much fuel is injected... I can monitor it with my right foot  :lol:

What does matter is how long and when!!!!!!!!!!!

How will you be controlling the timing electronically though

That would be harder but I don't plan to do it. The pump goes well like that... but when you turn on the fuel screw... timing is disadjust and it may serve for greater performance adjustment and also diagnostic for adjusting home made rebuild pump.... which I do.

The best way to control electronically the timing would be to swap the cover in front of the pump with a controlable servo motor. But I would aslo need to monitor engine performance... which requires fast mesurement like EGT, opacity sensor maybe... I don't know... it is another projet. I would act more in a correctiv way than in a complete way... which would be much harder since the pump always tries to advance timing.

Begin simple right now

[Northern RD]

First off the the injectors break at a certain pressure when the injector forces it too based on the timming of the injector pump. This functions like a gas injector except the process of injection is based on a pressurized fuel line and the injector opens up to a certain limit based on the comp.

The injectors are fully mechanical, there is no reall way to measure the tip of the nozzle, but you may be able to tap into the line before the injector from the pump to get a flow rate and thenfrom there calculate the fuel input.

As for  programming, you would receive a pattern of numbers that would have a higher or lower frequency based on the speed of the motor. The data would be useless because the pump that is on the motor is non electronic, other than the stop solenoid.  

I really can't think of a use for this, maybe for a pump pressure check ?

Also for the programming bit, you would need to have a device to store and run this program, these can be quite expensive.

 I deal with the axiom micro controllers, but you would have to have an LCD output to display any results, it also runs off a 5 v battery, and has very little power draw.

You might want to check with the folks at Orbital Engineering: they`ve been trying to do exactly what you`re describing and apply it to 2cycle engines since `87. They haven`t had much success I might add so there`s likely more to deal with than what you`ve outlined here,...

[Northern RD]

http://www.orbeng.com.au/orbital/directinjection/technical.htm

[Hey]

I don't need more than what I described... I just need to know what is the incidence of dual stage injector on a diesel engine and if it would be a good idea to monitor it.

Don't tell me I can't do it... because I am going to do it anyway... don't waste your time.  

Also... the link you sent me.. it is for gazoline engine. As there is no such words as lean, mixture, explosion in a diesel engine, I don't need to monitor the injection quantity on my IDI. It is going to be a tool to monitor timing ...that is all.

[QuickTD]

The BMW 524 engine used an injector pintle lift sensor for closed loop timing control on california models. 49 state engines used the lift sensor to calculate injected quantity for use by the vehicle trip computer in fuel consumption calculations. The sensor armature rested on the end of the injector pintle and moved within a coil mounted in the upper section of the injector body. I can't recall if the armature was a permanent magnet or if it was "excited" by a small DC current in the coil.

 The 524's timing was controlled in the pump in the same manner as the TDI, the timing piston is driven to an excessively advanced position by injection pump housing pressure and a solenoid valve bleeds fuel past the timing piston to retard the timing back to the desired value.

 The 2 spring injectors are pretty simple. At low flow levels (idle and light load) the injector operates on the primary spring in the same manner as a conventional injector. As injected quantity increases, pintle lift increases and the pintle contacts a heavier spring (through a shim) that increases the pressure needed to further lift the pintle. As a result, at high loads and throttle settings the injection pressure increases.

 A peizo crystal clamped onto the injection line is used in most diesel timing lights to sense the shock wave in the line and detect the exact moment of injection. This sensing method might be simpler than the pintle lift sensor, at least for a "retrofit" system.

[Hey]

Thank you QuickTD.. exactly what I wanted to hear.

If I followed correctly.. it will only increase injection pressure as more fuel is injected and will tend to extend the injection time since greater injection requires more camplate lift...but will not change the start of injection. right?

So it wouldn't be of great importance to know the niddle lift since I don't care how much fuel is injected. I only need to know of long.

I didn't though of a piezo on the injector line... that would be soooo easy... it is just the noize on the signal that may be a problem but if you say it has been done, it is doable (is it a word?) :roll: .

I want to be able to easily compare car performance and be able to put back the same timing on another car(want to tweak the pump further), so it would greatly help not to have to change the injector

....one could also check if the injection pressure is ok without dismanteling anything

Thank you again.

[QuickTD]

If I followed correctly.. it will only increase injection pressure as more fuel is injected and will tend to extend the injection time since greater injection requires more camplate lift...but will not change the start of injection. right?

Correct, the start of injection will not change. The duration of injection is mostly dependant on the camplate profile so it won't change much either.

I didn't though of a piezo on the injector line... that would be soooo easy... it is just the noize on the signal that may be a problem but if you say it has been done, it is doable (is it a word?) :roll: .

I briefly tried to make my own setup for this once, and almost had it working. I used the piezo element from a peizo buzzer and wired it to a coil which I placed around the inductive pickup of a standard timing light. It worked intermittantly and was sensitive to the amount of pressure that was applied piezo/line junction. I'm sure it could be made to work given enough time. Getting just the right number of turns in the coil and the proper pressure on the line would be the tricky part. Perhaps a heavy spring could be used to get consistant pressure on the line/piezo element?

[Hey]

maybe adding a very low frequency filter could help as a cylinder inject at more than 8Hz at 1000RPM (500RPM at the pump and 500/60~8).

everything else... pumps cracks are max 2-3Hz. This may be done with a programmable piezo... I will check it.

[Darkness_is_spreading]

If thats what you want you device to accomplish then go for it, but you will be wasting your own time to do this.

[QuickTD]

I think "Hey's" primary goal is to build a diesel timing light, a useful device by any standard. Setting timing by plunger lift alone is, at best, not very accurate. Plunger leakage, housing pressure, fuel density, belt tension and stretch and many other factors play into dynamic timing, having a real time external reference would be an excellent tool for setting the timing in a repeatable fashion. The ability to observe dynamic timing would also be an excellent tuning tool. Not a waste of time at all, IMHO anyway.