Author Topic: fuel injector timing (PWM)  (Read 30268 times)

0 Members and 2 Guests are viewing this topic.

Offline old greggTopic starter

  • Regular Contributor
  • *
  • Posts: 130
  • Country: 00
fuel injector timing (PWM)
« on: June 27, 2021, 11:37:05 am »
HI,

Does anyone have data on fuel injector timing ?

I'm working on a fuel injector tester based on a microcontroller. The idea is to test them in correlation with each other (to find out leaks). There's a 3 presets on the circuit. But I can't find the proper timing for signal to send to the injector. I'd like to have three different one : idle(800rev/min), médium rev (1500rev/min) and high rev(3500rev/min).

Doesn't have to be extremly precise but still something realistic.
thanks !
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3306
  • Country: gb
Re: fuel injector timing (PWM)
« Reply #1 on: June 27, 2021, 12:23:41 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

The period that the injector is open for will vary enormously depending on airflow which is controlled by throttle opening and engine RPM.  Typically the maximum duty cycle (i.e. at WOT, peak power RPM) will not exceed around 80%.  At idle the "electrical" open time may only be around 1ms, the actual open times will obviously be less as there is a dead time to consider for opening and closing that vary with injector type, voltage and even fuel pressure.

 
The following users thanked this post: bborisov567

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3288
  • Country: de
Re: fuel injector timing (PWM)
« Reply #2 on: June 27, 2021, 03:56:59 pm »
Hmm,

what type of injector?

port fuel injector
direct injection injector (GDI)
(I guess it is no diesel injector).

with best regards

Andreas
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6167
  • Country: de
Re: fuel injector timing (PWM)
« Reply #3 on: June 27, 2021, 04:01:42 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

 

Offline Miyuki

  • Frequent Contributor
  • **
  • Posts: 907
  • Country: cz
    • Me on youtube
Re: fuel injector timing (PWM)
« Reply #4 on: June 27, 2021, 04:38:56 pm »
Also, some modern injection system using multi injection - inject in multiple short pulses during intake (then it can be really short pulses)
Especially at light load when an injection is very short, can be just a few % compared to WOT
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6167
  • Country: de
Re: fuel injector timing (PWM)
« Reply #5 on: June 27, 2021, 04:41:11 pm »
Also, some modern injection system using multi injection - inject in multiple short pulses during intake (then it can be really short pulses)

Change "some" to "most, if not all", then I'm with you. Applies only to direct injection, though, but which engines have indirect injection these days?

 

Offline Miyuki

  • Frequent Contributor
  • **
  • Posts: 907
  • Country: cz
    • Me on youtube
Re: fuel injector timing (PWM)
« Reply #6 on: June 27, 2021, 04:53:10 pm »
Also, some modern injection system using multi injection - inject in multiple short pulses during intake (then it can be really short pulses)

Change "some" to "most, if not all", then I'm with you. Applies only to direct injection, though, but which engines have indirect injection these days?
In my collection, none of them is direct injection gasoline. Plenty are carburetor still (motorcycles) and unit injector diesel truck
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6167
  • Country: de
Re: fuel injector timing (PWM)
« Reply #7 on: June 27, 2021, 05:13:12 pm »
Also, some modern injection system using multi injection - inject in multiple short pulses during intake (then it can be really short pulses)

Change "some" to "most, if not all", then I'm with you. Applies only to direct injection, though, but which engines have indirect injection these days?
In my collection, none of them is direct injection gasoline. Plenty are carburetor still (motorcycles) and unit injector diesel truck

We're not talking Skoda 105 here.
Today: FSI, TFSI, TSI etc. Almost all modern cars with petrol engines are direct injection. Trucks as well (Diesel are always direct injection, otherwise it doesn't work). Motorcycles: who cares?

 

Offline Miyuki

  • Frequent Contributor
  • **
  • Posts: 907
  • Country: cz
    • Me on youtube
Re: fuel injector timing (PWM)
« Reply #8 on: June 27, 2021, 05:35:22 pm »
Diesel are always direct injection, otherwise it doesn't work.
Yes, but common rail with fast piezo injectors is a recent thing with arguable reliability. So not all engines use it.
Motorcycles: who cares?
Look at Austria with their ban on loud ones ::)
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6167
  • Country: de
Re: fuel injector timing (PWM)
« Reply #9 on: June 27, 2021, 06:20:55 pm »
@old gregg:
Why don't you just put a scope on one of the injector control signals on a real car? That should give you an idea. There shouldn't be any issue, it's just 12 V pulses coming from the ECU.

There's no big difference between cars, the signals are pretty much the same.

 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11762
  • Country: us
Re: fuel injector timing (PWM)
« Reply #10 on: June 27, 2021, 06:21:27 pm »
HI,

Does anyone have data on fuel injector timing ?

I'm working on a fuel injector tester based on a microcontroller. The idea is to test them in correlation with each other (to find out leaks). There's a 3 presets on the circuit. But I can't find the proper timing for signal to send to the injector. I'd like to have three different one : idle(800rev/min), médium rev (1500rev/min) and high rev(3500rev/min).

Doesn't have to be extremly precise but still something realistic.
thanks !

Maybe just a signal generator and driver.   They are fairly inexpensive now and would allow you to create pretty much any waveform you want.   

I made something a bit more complex.   Here you can see it being used to evaluate an automotive DMM. 
https://youtu.be/q_89qoFMivg?list=PLZSS2ajxhiQBvWvqMVLdRQMjGofKpQUJr&t=2193

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4652
  • Country: dk
Re: fuel injector timing (PWM)
« Reply #11 on: June 27, 2021, 06:26:49 pm »
Also, some modern injection system using multi injection - inject in multiple short pulses during intake (then it can be really short pulses)

Change "some" to "most, if not all", then I'm with you. Applies only to direct injection, though, but which engines have indirect injection these days?
In my collection, none of them is direct injection gasoline. Plenty are carburetor still (motorcycles) and unit injector diesel truck

We're not talking Skoda 105 here.
Today: FSI, TFSI, TSI etc. Almost all modern cars with petrol engines are direct injection. Trucks as well (Diesel are always direct injection, otherwise it doesn't work). Motorcycles: who cares?

and many of them also have port injection, it works better at some loads and has the added benefit of keeping the intake clean


 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: fuel injector timing (PWM)
« Reply #12 on: June 27, 2021, 06:27:55 pm »
I don't know if this is still the case, but with the 80s-90s stuff I'm familiar with there are two basic types of injectors, low impedance and high impedance. The low impedance injectors require an external ballast resistor so you don't burn them out.
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4652
  • Country: dk
Re: fuel injector timing (PWM)
« Reply #13 on: June 27, 2021, 06:31:36 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

why? some might inject on a closed intake valves to evaporate the fuel on the back of the valve, but it high loads there really isn't any choice because the injector is on almost all the time


 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: fuel injector timing (PWM)
« Reply #14 on: June 27, 2021, 06:53:12 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

why? some might inject on a closed intake valves to evaporate the fuel on the back of the valve, but it high loads there really isn't any choice because the injector is on almost all the time

Maybe he's confusing direct injection and port injection? With port injection it's impossible to send unburned fuel straight out the exhaust, the atomized fuel/air mixture can only enter the cylinder during the intake stroke. Even if the intake valve is fully open, the injector is not shooting a stream of fuel down into the cylinder, it's spraying out a mist of fuel in the intake port which is picked up by the air flowing down into the cylinder. The injector timing does not have to be precise in relation to crank angle and injecting while the intake valve is closed poses no problem at all. My cars are batch fire, all four injectors open and close at the same time, the ECU knows the RPM but has no knowledge of the valve positions.
 

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11762
  • Country: us
Re: fuel injector timing (PWM)
« Reply #15 on: June 27, 2021, 06:57:15 pm »
I've heard of injecting at odd times to heat the cat for emissions. 

About 25 years or so, I was reading an article about Honda bringing a modified 2-stroke bike to the USA to run the Baja.  And HCCI was born...

https://en.wikipedia.org/wiki/Honda_EXP-2

Sounds like OP just wants to sort out if injectors by comparing flow rates and such.   Using the signal generator and some custom drivers for the various injector types would seem easy enough.   Harder part may be building some sort of generic flow bench.    Interesting project.   Post some pictures as you progress. 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4652
  • Country: dk
Re: fuel injector timing (PWM)
« Reply #16 on: June 27, 2021, 07:20:44 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

why? some might inject on a closed intake valves to evaporate the fuel on the back of the valve, but it high loads there really isn't any choice because the injector is on almost all the time

Maybe he's confusing direct injection and port injection? With port injection it's impossible to send unburned fuel straight out the exhaust, the atomized fuel/air mixture can only enter the cylinder during the intake stroke. Even if the intake valve is fully open, the injector is not shooting a stream of fuel down into the cylinder, it's spraying out a mist of fuel in the intake port which is picked up by the air flowing down into the cylinder. The injector timing does not have to be precise in relation to crank angle and injecting while the intake valve is closed poses no problem at all. My cars are batch fire, all four injectors open and close at the same time, the ECU knows the RPM but has no knowledge of the valve positions.

probably fired once per revolution just to make sure all cylinders get the same injection timing regardless of the engine phase

 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6167
  • Country: de
Re: fuel injector timing (PWM)
« Reply #17 on: June 27, 2021, 08:00:54 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

why? some might inject on a closed intake valves to evaporate the fuel on the back of the valve, but it high loads there really isn't any choice because the injector is on almost all the time

Maybe he's confusing direct injection and port injection? With port injection it's impossible to send unburned fuel straight out the exhaust, the atomized fuel/air mixture can only enter the cylinder during the intake stroke. Even if the intake valve is fully open, the injector is not shooting a stream of fuel down into the cylinder, it's spraying out a mist of fuel in the intake port which is picked up by the air flowing down into the cylinder. The injector timing does not have to be precise in relation to crank angle and injecting while the intake valve is closed poses no problem at all. My cars are batch fire, all four injectors open and close at the same time, the ECU knows the RPM but has no knowledge of the valve positions.

probably fired once per revolution just to make sure all cylinders get the same injection timing regardless of the engine phase

A couple of users here need to attend "Otto-Engine 101".

The basic fuel injection (around 50 years ago) was the single point injection. This was a carburetor replacement that injected fuel high up in the intake manifold (before it branches out). It was inefficient and not successful.
The real breakthrough came with injection into the single manifold branches in the 70s/80s (Jetronic, anyone?). And no, these did not inject all the time. They injected when the associated intake valve was open and intake air streamed by. The purpose of a carburetor or injection is to evaporate the gasoline, making the air/fuel mixture into a stoichiometric and combustible gas.

The idea of having a "puddle" of gasoline waiting above the intake valve is ridiculous.

Direct fuel injection was the next step in this development and used in 80% of today's petrol cars. At least in the western hemisphere.

« Last Edit: June 27, 2021, 08:02:37 pm by Benta »
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4652
  • Country: dk
Re: fuel injector timing (PWM)
« Reply #18 on: June 27, 2021, 08:37:54 pm »
A typical modern sequential port injection system will fire the injector once for every two crank revolutions since it's timed to the opening of the inlet valve e.g. 6000RPM would give 50Hz frequency or 20ms period.  The older batch injection systems (all injectors fire together, or all injectors on one bank of a V engine fire together) may have one injection event per crank revolution e.g. 6000RPM would give 100Hz frequency or 10ms period.

I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

why? some might inject on a closed intake valves to evaporate the fuel on the back of the valve, but it high loads there really isn't any choice because the injector is on almost all the time

Maybe he's confusing direct injection and port injection? With port injection it's impossible to send unburned fuel straight out the exhaust, the atomized fuel/air mixture can only enter the cylinder during the intake stroke. Even if the intake valve is fully open, the injector is not shooting a stream of fuel down into the cylinder, it's spraying out a mist of fuel in the intake port which is picked up by the air flowing down into the cylinder. The injector timing does not have to be precise in relation to crank angle and injecting while the intake valve is closed poses no problem at all. My cars are batch fire, all four injectors open and close at the same time, the ECU knows the RPM but has no knowledge of the valve positions.

probably fired once per revolution just to make sure all cylinders get the same injection timing regardless of the engine phase

A couple of users here need to attend "Otto-Engine 101".

The basic fuel injection (around 50 years ago) was the single point injection. This was a carburetor replacement that injected fuel high up in the intake manifold (before it branches out). It was inefficient and not successful.
The real breakthrough came with injection into the single manifold branches in the 70s/80s (Jetronic, anyone?). And no, these did not inject all the time. They injected when the associated intake valve was open and intake air streamed by. The purpose of a carburetor or injection is to evaporate the gasoline, making the air/fuel mixture into a stoichiometric and combustible gas.

The idea of having a "puddle" of gasoline waiting above the intake valve is ridiculous.

the valve is hot it doesn't just puddle, and no matter what you do once you use more that ~30% of the fuel injector some of it is going to be on a closed intake vale

 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: fuel injector timing (PWM)
« Reply #19 on: June 28, 2021, 01:35:13 am »
A couple of users here need to attend "Otto-Engine 101".

The basic fuel injection (around 50 years ago) was the single point injection. This was a carburetor replacement that injected fuel high up in the intake manifold (before it branches out). It was inefficient and not successful.
The real breakthrough came with injection into the single manifold branches in the 70s/80s (Jetronic, anyone?). And no, these did not inject all the time. They injected when the associated intake valve was open and intake air streamed by. The purpose of a carburetor or injection is to evaporate the gasoline, making the air/fuel mixture into a stoichiometric and combustible gas.

The idea of having a "puddle" of gasoline waiting above the intake valve is ridiculous.

Direct fuel injection was the next step in this development and used in 80% of today's petrol cars. At least in the western hemisphere.

I think it is you who needs to read up on how this stuff works because most of what you said is flat out wrong. I have been tinkering, maintaining, repairing and building engines for decades and have owned numerous cars with Bosch injection systems and worked on many others. I still have a car that I retrofitted with Megasquirt EFI that I built, wired and tuned myself so I have quite intimate knowledge of how that system works.

Throttle body injection not a success? It was extremely successful, used on many millions of cars and trucks built right up into the late 90s-mid 2000's. Emissions and fuel economy requirements eventually resulted in port injection displacing it but to say it was not a success is absurd.

The Bosch mechanical CIS injection used on many different European cars up into the early 80s absolutely does inject 100% of the time, hence the CIS name Volvo used for it, Continuous Injection System. Officially it was called K-jetronic, I suspect the K stands for "konstant". The injectors are pressure actuated and fuel metering is controlled by an airflow sensor plate that mechanically moves a piston in the fuel distributor. The simplest of these systems had no electronics at all except for the fuel pump and a cold start injector.

The later Bosch Jetronic systems used more modern style electrically actuated injectors but these are batch fire, all four (or however many) injectors fire at the same time, at least half of them are guaranteed to inject while the intake valve is closed. Fuel doesn't puddle,  it just doesn't happen that way, despite what you might imagine. The injectors do not piss a stream of fuel into the cylinder, as I said before, they spray an atomized mist into the intake port on the head, it mixes with the air and is drawn into the cylinder on the intake stroke. Trust me, it is not timed to the valve actuation, and yet it works well. There exist sequential injection systems that do try to inject while the valve is open but these have historically been uncommon. I don't really work on anything newer than early 2000's so I don't know what the latest cars are doing, mostly direct injection I assume.

I found a manual for the old L-Jetronic system, this particular system is not one I've dealt with but it looks similar to the LH-Jetronic which I am very familiar with, the main difference is the LH system utilizing a hot wire air mass meter instead of a mechanical flapper.

I've also attached an excerpt from this manual produced by Bosch, describing exactly the operation you are saying is ridiculous.

Edit: Just realized I forgot to include the link to the manual I mentioned. http://www.cardiagnostics.be/-now/Educational_sites_bestanden/BOSCH%20L-Jetronic%20Injection%20Manual.pdf
« Last Edit: June 28, 2021, 07:36:21 pm by james_s »
 

Offline John B

  • Frequent Contributor
  • **
  • Posts: 816
  • Country: au
Re: fuel injector timing (PWM)
« Reply #20 on: June 28, 2021, 02:06:31 am »
Port injectors do indeed spray onto closed valves. Since most standard injectors are fairly small in flow rate, the inj on time is significant in relation to the mechanical sequence of stuff going on in the engine. By the time the intake valves are opening, the injector pulse should be finished or very close to.

You might think it's a good idea to have the injector spraying while the intake is open, but the velocity of the intake air would drop as the piston reaches the bottom of the intake stroke.

Also for a fully sequential injection system, you want only one injector pulse per intake because you want to minimise the effects of injector dead time.

These effects are most apparent at low loads and low RPM, as the engine speed and load increases, injector timing matters little as you should expect the injector duty cycle to reach over 80% anyways.

There's no set way though, as engine geometry and head flow come into play. Since a standard production engine would have no overlap at idle and low RPM (for emission reasons), the air/fuel charge partially shooting through the closing exhaust valve when the intake is opening wont happen. Things change when high overlap and lift cams are thrown in, but these would be modified engines.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7997
  • Country: us
Re: fuel injector timing (PWM)
« Reply #21 on: June 28, 2021, 04:06:58 am »
A couple of users here need to attend "Otto-Engine 101".

I'll leave it to others to pillory you for your statements.  I'll just point out a few things to ponder...

When it comes to injection timing, if the engine and it's valve timing will work with a carburetor, it will work with any injection timing, including continuous injection--without any 'shoot-through' of fuel.  The rest is fine-tuning--and that is a whole art and science of its own.

There's almost no fuel system you can think of that hasn't already been tried, and most things have been tried more than a hundred years ago.  As someone that has seen, worked on and used almost all of the ones that made it to production, I can tell you that the implementation matters more than the type when it comes to results.  Early designs sometimes lacked the underlying technology to make them work as well as we expect from modern designs, but were still well-regarded in their day. Others were simply tried too soon.  Also, when you contemplate 'Otto' engines, there's a lot of technological history in aviation, marine and stationary engines.

'Single point' (actually sometimes double-point) fuel injection mainly came in two basic varieties, one had a mechanical control mechanism in the intake manifold and injected fuel into a plenum.  The one I'm familiar with were the Bendix systems and AFAIK they were used only in aviation engines.  Its advantages were that it worked well with pressurized manifolds (supercharging) and was not sensitive to G-force so that it worked inverted. I've flown planes with it and while it is perfectly workable, it's operational quirks would not be tolerated by the automotive driving public--things like a mixture control on the dash wouldn't go over well.  The only car engine I know that had anything remotely comparable was the Mitsubishi Starion/Chrysler Conquest that had an electronic system that kinda sorta was similar.  These systems were developed from the late 1800s to the 1950s, the ones that worked well enough to survive were developed in the late 1920s and 1930s.  They are still made today.

The other flavor of single (or double) point fuel injection was Throttle Body Injection (TBI) which was essentially a bolt-on replacement for a Rochester (GM) or Holley (Chrysler) carburetor.  These arrived in the late 70's and with the exception of a few early duds, were reliable and efficient (albeit with fairly low power density engines) for two decades.  These injected the fuel outside the manifold, at atmospheric pressure, onto the throttle plates, thus the name.

Electronic port fuel injection and direct injection (GDI) both had their mainstream automotive debuts in the 1950s, with limited success (Bosch GDI) and failure (Chrysler Electrojector).  Chrysler introduced two different EFI systems, in the 1958 300D and the 1980 Imperial, that had to be recalled and replaced with carburetors.  Bosch finally introduced the analog D-Jetronic a decade or so later, the first widely produced reliable system.

You seem to imply that all modern cars have, or should have, GDI as if it were inherently superior.  That's just not true, IMO.  GDI has become more popular over the past decade, but like port fuel injection over TBI, the main issues are power density and emissions, not reliability or efficiency.  Both of my non-GDI cars are reliable and very, very clean (SULEV and PZEV, respectively) and have lifespans of 200k miles plus, but their power density isn't as high as the insane levels people seem to demand these days.  GDI has had some serious issues since it appeared 25 years ago, including carbon build up on valves and particulate emissions.  Maybe they have it working well enough now,
« Last Edit: June 28, 2021, 04:13:06 am by bdunham7 »
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline mck1117

  • Contributor
  • Posts: 36
  • Country: us
Re: fuel injector timing (PWM)
« Reply #22 on: June 28, 2021, 04:10:02 am »
I think you're confusing injection with ignition. The scheme you describe would send 50% uncombusted fuel through the exhaust.

No, it won't.  The injector fires at the closed intake valve (or open intake valve during the intake stroke), so it can't shoot much fuel directly out the exhaust.
 

Offline mck1117

  • Contributor
  • Posts: 36
  • Country: us
Re: fuel injector timing (PWM)
« Reply #23 on: June 28, 2021, 04:13:04 am »
Port injectors do indeed spray onto closed valves. Since most standard injectors are fairly small in flow rate, the inj on time is significant in relation to the mechanical sequence of stuff going on in the engine. By the time the intake valves are opening, the injector pulse should be finished or very close to.

You might think it's a good idea to have the injector spraying while the intake is open, but the velocity of the intake air would drop as the piston reaches the bottom of the intake stroke.

Also for a fully sequential injection system, you want only one injector pulse per intake because you want to minimise the effects of injector dead time.

These effects are most apparent at low loads and low RPM, as the engine speed and load increases, injector timing matters little as you should expect the injector duty cycle to reach over 80% anyways.

There's no set way though, as engine geometry and head flow come into play. Since a standard production engine would have no overlap at idle and low RPM (for emission reasons), the air/fuel charge partially shooting through the closing exhaust valve when the intake is opening wont happen. Things change when high overlap and lift cams are thrown in, but these would be modified engines.

Correct - the biggest concerns for road vehicles at this point are emissions and fuel economy (usually in that order), and injector timing (specifically firing at the back of the closed intake valve) gives big gains for emissions.  You get all the fuel fully vaporized before the valve opens, such that you have no chance of liquid droplets hitting the piston and ending up as HC emissions (or soot).  In fact the major problem with GDI is soot production from fuel impinging on the piston crown, hence the introduction of GDI+PFI dual injector setups that are now common on GDI engines.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3288
  • Country: de
Re: fuel injector timing (PWM)
« Reply #24 on: June 28, 2021, 04:52:29 am »
There shouldn't be any issue, it's just 12 V pulses coming from the ECU.

Sorry but 12V pulses for activation (usually on the low side) are true only for port fuel injectors.
The free wheeling is always into around 50-70V here for a fast closing of the valve.

On GDI systems we usually have booster voltages >= 50V for fast opening of the valves.
(The time for possible activation is shorter so we have to hurry up).
The actuation is current controlled.
And caution: none of the both pins has relation to ground or UBat since both pins are controlled by the power stage.

The free wheeling goes back to the booster voltage when the valve is closed.

with best regards

Andreas

edit:
a typical GDI power stage can be found in the PT2000 datasheet of NXP (see Figure 35 on page 139)

https://www.nxp.com/docs/en/data-sheet/MC33PT2000.pdf?pspll=1
« Last Edit: June 28, 2021, 06:39:46 am by Andreas »
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 7125
  • Country: hr
Re: fuel injector timing (PWM)
« Reply #25 on: June 28, 2021, 06:37:06 am »
I guess OP wants to do something like this...

As Andreas said, for many years (before GDI) injectors are driven with constant current, and cleverly using injector own inductance to create high voltage for increased voltage compliance for current drive... But before that they used simply 12V and open collector type outputs...  Injector type needs to be known beforehand and proper drive type used.
 

Offline old greggTopic starter

  • Regular Contributor
  • *
  • Posts: 130
  • Country: 00
Re: fuel injector timing (PWM)
« Reply #26 on: June 28, 2021, 01:09:12 pm »
Quote
Why don't you just put a scope on one of the injector control signals on a real car? That should give you an idea. There shouldn't be any issue, it's just 12 V pulses coming from the ECU.

There's no big difference between cars, the signals are pretty much the same.

I don't have power in the garage to put the oscilloscope on the injectors. Would have been to easy, but that's the ultimate solution if I can't find the proper data. 

actually the project is alive and running, just need to fix two issues : the timing of the injection and a software related issue (stm32).

Quote
I guess OP wants to do something like this...

As Andreas said, for many years (before GDI) injectors are driven with constant current, and cleverly using injector own inductance to create high voltage for increased voltage compliance for current drive... But before that they used simply 12V and open collector type outputs...  Injector type needs to be known beforehand and proper drive type used.

indeed, looks a lot like it ! I've to read it. thanks !



 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7997
  • Country: us
Re: fuel injector timing (PWM)
« Reply #27 on: June 28, 2021, 01:14:39 pm »
I guess the OP's questions deserve an answer!

The first reply post from mikerj was the basic correct answer for that type of fuel injection.  However, since I have some actual experience with fuel injector testers, I can tell you that unless you have or can reverse engineer the specs for the particular, exact injector system you are wanting to test, you won't be able to do very much that is useful.  In other words, you have to know exactly how the injector functions in its normal environment and then reproduce that as best you can.

The variety of systems used over the years covers almost every way of doing things that you could imagine.  Here's a few tips on testing injectors.

Electrically, as james_s referred to, there are injectors that function at 12 volts without any current limiting and those that are lower impedance and need current control.  Ballast resistors were a sort of early Datsun 280-ish solution, the way it is typically implemented is to use full current for the turn on (say 1ms) and then use PWM current control after that.  If you can think of a different way to do it, someone has probably done it that way.  High pressure injectors such as modern GDI units have different drive systems and use higher voltage and power.  If you don't closely replicate the original control system, you will get incorrect test results.

Mechanically, injectors can fail in a variety of ways in service and it can be hard to spot some of these on a test bench.  For example, a typical port injector will have a pintle blocking the nozzle that is pulled back by the coil, allowing fuel to flow.  The pintle may vibrate and this helps in atomizing the fuel into a mist.  This pintle vibration may be delayed in a worn or contaminated injector, resulting in a non-atomized stream for the first few ms of operation.  On a test bench with continuous flow the pattern looks OK, but with a 2ms pulse, it isn't. 

I don't think generalized solutions are practical.  If you want advice on the subject, you'll have to be very specific as to what system or systems you want to test.
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4652
  • Country: dk
Re: fuel injector timing (PWM)
« Reply #28 on: June 28, 2021, 02:18:58 pm »
Quote
Why don't you just put a scope on one of the injector control signals on a real car? That should give you an idea. There shouldn't be any issue, it's just 12 V pulses coming from the ECU.

There's no big difference between cars, the signals are pretty much the same.

I don't have power in the garage to put the oscilloscope on the injectors. Would have been to easy, but that's the ultimate solution if I can't find the proper data. 

actually the project is alive and running, just need to fix two issues : the timing of the injection and a software related issue (stm32).

Quote
I guess OP wants to do something like this...

As Andreas said, for many years (before GDI) injectors are driven with constant current, and cleverly using injector own inductance to create high voltage for increased voltage compliance for current drive... But before that they used simply 12V and open collector type outputs...  Injector type needs to be known beforehand and proper drive type used.

indeed, looks a lot like it ! I've to read it. thanks !

I see a flyback diode, that'll give you a very slow turn off. You need a fet that clamps or is rated for avalance at ~40V




 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: fuel injector timing (PWM)
« Reply #29 on: June 28, 2021, 07:39:21 pm »
I don't have power in the garage to put the oscilloscope on the injectors. Would have been to easy, but that's the ultimate solution if I can't find the proper data. 

Inverters are cheap, and even a small one will power most scopes. You can power it from the car with the engine you're wanting to test.
 

Offline m3vuv

  • Super Contributor
  • ***
  • !
  • Posts: 1738
  • Country: gb
Re: fuel injector timing (PWM)
« Reply #30 on: October 18, 2021, 02:41:26 pm »
i suppose it depends how wild the cam timing is ,ie valve overlap.!!
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: fuel injector timing (PWM)
« Reply #31 on: October 18, 2021, 05:57:40 pm »
No it doesn't. The injector timing has no relation to the camshaft profile, the ECU doesn't know what kind of camshaft is installed, though the cam does affect the optimal fuel mapping.
 

Offline robzy

  • Regular Contributor
  • *
  • Posts: 142
  • Country: au
Re: fuel injector timing (PWM)
« Reply #32 on: October 19, 2021, 03:31:01 am »
If you're looking for prior art, check out the Speeduino project:

https://speeduino.com/home/

I used Speeduino as a reference when I was designing my own ECU.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf