Hi there, I'm basically working in a project where we are building a dual fuel injection system. For prototyping we are tapping the injection pulse of a fuel injector of a running engine and reading it via optocoupler and actuating another injector by pulse read by a microcontroller. We used Arduino for first stage of prototype, where the Arduino gets digital input from the optocoupler and adjusts the pulse width and gives out as output in a digital pin, which is connected to a MOSFET's gate pin. The drain and source pin of the MOSFET are connected between the injector's control line.

But we need a more reliable approach, since the injector we are using is peak and hold diesel injector. Currently we referred few sources and built a circuit, which resembles something like in the image. But is there any better way to read the signal and controlling the injector pulses using a microcontroller?


circuit.png
Google Docs

 

The Texas Instruments LM1949 IC was specifically designed for firing and holding low impedance injectors. Not sure if it's obsolete, certainly still available, not necessary with newer high impedance injectors.

 

Note: This thread is also running on the ETO site. The schematic shown has numerous errors.

 

The Texas Instruments LM1949 IC was specifically designed for firing and holding low impedance injectors. Not sure if it's obsolete, certainly still available, not necessary with newer high impedance injectors.

Yes I have similar experience,
I have an injector setup which uses the diesel injector, currently Im using MOSFET to control the low side of the injector. The high side is connected to constant supply 12v SMPS. Now I'm able to control the injector with this setup. But I faced 2 issues,
1) I lost 2 injectors due to this approach, not in a single use. But after prolonged testing.
2) in idling of engine, by tapping the engine's injector pulses and mirroring it to the MOSFET which controls the injector, I got around 90ml of diesel for 60seconds. Which seems very unrealistic.

 

Note: This thread is also running on the ETO site. The schematic shown has numerous errors.

Yes, I'll fix the circuit. But is the approach of the circuit correct? Apart from the connecting issues.

 

Hi there, I'm basically working in a project where we are building a dual fuel injection system. For prototyping we are tapping the injection pulse of a fuel injector of a running engine and reading it via optocoupler and actuating another injector by pulse read by a microcontroller. We used Arduino for first stage of prototype, where the Arduino gets digital input from the optocoupler and adjusts the pulse width and gives out as output in a digital pin, which is connected to a MOSFET's gate pin. The drain and source pin of the MOSFET are connected between the injector's control line.

But we need a more reliable approach, since the injector we are using is peak and hold diesel injector. Currently we referred few sources and built a circuit, which resembles something like in the image. But is there any better way to read the signal and controlling the injector pulses using a microcontroller?


circuit.png
Google Docs


View attachment 370857

I have an injector setup which uses the diesel injector, currently Im using MOSFET to control the low side of the injector. The high side is connected to constant supply 12v SMPS. Now I'm able to control the injector with this setup. But I faced 2 issues,
1) I lost 2 injectors due to this approach, not in a single use. But after prolonged testing.
2) in idling of engine, by tapping the engine's injector pulses and mirroring it to the MOSFET which controls the injector, I got around 90ml of diesel for 60seconds. Which seems very unrealistic.

The circuit is a rough idea, but please give some of your approach in working with this project.

 

Note: This thread is also running on the ETO site. The schematic shown has numerous errors.

I see a total of TEN schematics in post #1. So which one is being used?? OR, is it a different one that I don't see??

 

At all times, fuel injector timing and duty cycle are very important!! Check the manufacturer's specifications. DRY operation is not recommended!!!
As the TS has discovered, it is easy to overheat and damage injectors, which are cooled by fuel flow in normal application. The BETTER QUALITY injector test stands are able to protect the items being tested against overheating damage.

 

Yes I have similar experience,
I have an injector setup which uses the diesel injector, currently Im using MOSFET to control the low side of the injector. The high side is connected to constant supply 12v SMPS. Now I'm able to control the injector with this setup. But I faced 2 issues,
1) I lost 2 injectors due to this approach, not in a single use. But after prolonged testing.
2) in idling of engine, by tapping the engine's injector pulses and mirroring it to the MOSFET which controls the injector, I got around 90ml of diesel for 60seconds. Which seems very unrealistic.

90 ml/Min would translate to 5.4 l/h. What kind of motor is this ? No of cylinders, displacement, power, rpms in idle, what kind of injection system ?
For a truck or a smaller ship this might be ok, for a passenger car certainly not.

Did you MEASURE these 90 ml/Min or did you just calculate this value via the pulse width ? Some injection systems (resp. the injectors) take an itsy-bitsy ONE ms (some even more) from the start of the pulse to the start of the injection. (1 ms sounds much, but considering that for a 4-stroke 4-cylinder engine the "cylinder-frequency" for 6000 rpms (crankshaft) is a meager 50 Hz, you have 20 ms for each cylinder to position and execute your injection. And the fuel delivery rate for high rpms can be astonishing high

 

My last experience testing injectors was with some used in tank engines, at Delphi Diesel div. I gave the client the results, telling them that I did not believe the aparent flow rate. They responded that the flow rate for that pressure and pulse rate was correct. So evidently there are some big engines that use a lot of fuel, and run quite fast. No additional information was provided when I asked.

 

The question about how you obtained 90 mL/min is worth answering first. Was that fuel actually collected from the injector nozzle, measured in a supply/return line, or calculated from the electrical pulse width? Those measurements describe different things. Although 90 mL/min equals 5.4 L/h, we need the injector type, fuel pressure and injection frequency to judge whether it makes sense.

If you later automate the quantity measurements, low-flow fuel meters describes positive-displacement meters with pulse outputs and ranges expressed in mL/min or L/h. It’s useful background for selecting measurement equipment, but you would need confirmation of suitability for the actual pressure, pulsating flow and allowable pressure drop. An average flow reading also won’t establish whether individual injections are timed correctly.

For the damaged injectors, I’d prioritise obtaining the specified peak/hold current profile and comparing it with your measured coil-current waveform. The LM1949 datasheet explains the current-control principle, but suitability for your particular diesel injector still needs checking. Could you share its exact part number and operating pressure?

 

Really, you need a triggered pulse system to limit the pulse "on time", AND you must have fuel flowing at the rated pressure to provide the correct mechanical valve load. They work differently with no fuel and no pressure!
Injectors need to be protected, as the TS has discovered.