Measuring injector dead time

Over the past few months I’ve been putting a lot of focus into fuel injector dynamics. I’ve been curious to develop cost effective method of being able to analyze fuel injectors. The goal is to be able to present testing methods to find injector flow rate, injector opening time, and pulse width battery compensation to the avid electronic fuel injection hobbyist. Injector dead time, injector latency, and injector opening time are all ways to describe the time it takes for the injector to start delivering fuel after the injector coil has been energized.

A quick search on the internet will reveal that one method of measuring injector dead time is using an oscilloscope to display the current ramp as the injector opens. The idea being that the current draw of the coil differs if the injector pintle is moving or not moving. I’ve used this method in the past and it seems to get you in the ballpark. This method, however, is not very accurate.

Two considerations must be made: The injector opening time is going to vary with fuel pressure. Higher fuel pressures will make the injector open slower due to pressure of the fuel acting against the pintle. Using the oscilloscope method with a “dry” injector will show injector opening times generally quicker than if the injector was loaded in the application. Also, injector opening time represents the time it takes for the pintle to unseat and start fuel delivery. Observing current draw to determine injector opening time will show you when the pintle has stopped moving, when it is fully open. Most fuel injectors will begin delivering fuel before the pintle is fully open.

A simple method of measuring injector opening time is to simulate the injectors operating environment and physically observe when the injector begins to deliver fuel by precisely controlling the injector pulse width. To simulate the injector operating environment I assembled an apparatus with parts around the shop that will emulate a vehicles fuel system. This consisted of a fuel pump, fuel pressure regulator, and a fuel injector to be tested. Instead of using mineral spirits to test the injector I used the same type of gasoline that cars run on to make the test as accurate as possible. The injector is controlled by a megasquirt ECU I had on my test bench and two power supplies are used to power both the injector and fuel pump. The goal of this experiment is to measure the injector opening time of a few different types of fuel injectors I had laying around  and be able to determine the slope of the opening time as battery voltage varies.

The following parts were used to perform this experement:
-1988 Mazda RX-7 Turbo II in tank fuel pump
-Adjustable fuel pressure regulator
-Device to contain fuel
-13.8 volt power supply
-8.67 volt power supply
-Bosch 280150734 fuel injector (200cc 1988 Volvo 240 LH2.2) High Impedance
-Denso B240H fuel injector (240cc NA 2.0 DSM 90-94) High Impedance
-Denso 195500-1370 fuel injector (550cc RX-7 TurboII S4) Low Impedance
-Denso B450L fuel injector (Turbo DSM) Low Impedance

Test apparatus. A generator fuel tank was used to contain the fuel and fuel pump. The regulator, and fuel injector.

Test apparatus. A generator fuel tank was used to contain the fuel and fuel pump. The regulator, and fuel injector.

13.8v and 8.66v power supplies, and megasquirt ECU

13.8v and 8.66v power supplies, and megasquirt ECU

Injectors used for test. From left to right: Bosch 280150734, Denso B450L, Denso B240H, Denso 195500-1370

Injectors used for test. From left to right: Bosch 280150734, Denso B450L, Denso B240H, Denso 195500-1370

The setup of the apparatus was pretty simple. I have a generator that has a large enough opening in the fuel tank to fit my fuel pump and return line from the regulator. The generator’s fuel tank acted as my fuel reservoir.  A hose was clamped to the outlet of the fuel pump and connected to a brass tee where my fuel injector would be connected on one end and the other end feeds into the fuel pressure regulator. The return of the fuel pressure regulator fed back into the fuel reservoir. The regulator was set at 42psi which is the general standard for testing fuel injectors. It would be wise to adjust the pressure to match the operating pressure in the application being tested. Since this is just an experiment, I’m using 42psi. It goes without saying that we’re dealing with pressurized gasoline so this experiment is extremely dangerous. If you choose to do this yourself take every precaution to ensure that every hose connection is tight. Only use as much gasoline needed to pressurize the injector . Keep a fire extinguisher on hand at all times. Use common sense when performing this experiment.

Now I have pressurized fuel being fed to my injector. Time to energize the injector with precise pulses to determine the opening time. I used a megasquirt ecu to control the fuel injector. The megasquirt firmware has an output test mode that allows me to define both pulse width and interval between pulses.  I start off with a number higher than the opening time to make sure the injector is pulsing as it should. I started out with 10ms pulse width every 100ms. Each injector pulsed as it should. I kept the interval at 100ms but brought the pulse width down to 1ms. From there I adjusted the pulse width in .1ms increments until the injector stopped delivering fuel then increased the pulse width another .1ms. Since injector opening time is time until start of delivery, we’re looking for just a dribble of fuel coming out of the injector. Instead of waiting for an atomized stream of fuel, adjust pulse width until the fuel just begins to squirt out of the injector. This is the start of delivery and the pulse width required to start the delivery is the injector dead time. I tested all of the injectors at 13.8 volts then switched to a 8.66 volt power supply and did the test again. Current limiting was not used on the low impedance injectors because it is normally not used until the injector is fully open. In addition to measuring injector dead time, I measured the flow of each of the injectors I tested using the method documented on the Megasquirt Manual website.

Output test mode in megasquirt used to energize the injector

Output test mode in megasquirt used to energize the injector

Below are the results I observed with the injectors that I tested. The low impedance injectors were tested in two modes: Peak current, and saturated by adding a 7.5 ohm inline resistor.

Bosch 280150734:

  • 1.1ms @ 13.8v
  • 2.1ms @ 8.66v
  • Battery voltage correction: .19ms/v
  • Flow rate: 210cc/min @ 42psi

Denso B240H:

  • .97ms @ 13.8v
  • 1.7ms @ 8.66v
  • Battery voltage correction: .14ms/v
  • Flow rate: 249cc/min @ 42psi

Denso B450L:

  • .68ms @ 13.8v
  • 1.0ms @ 8.66v
  • Battery voltage correction: .062ms/v
  • .93ms @ 13.8v with 7.5 ohm resistor
  • 1.4ms @ 8.66v with 7.5 ohm resistor
  • Battery voltage correction with resistor: .09ms/v
  • Flow rate: 450cc/min @ 42psi

Denso 195500-1370:

  • .75ms @ 13.8v
  • 1.2ms @ 8.66v
  • Battery voltage correction: .085ms/v
  • 1.2ms @ 13.8v with 7.5 ohm resistor
  • 2.0ms @ 8.66v with 7.5 ohm resistor
  • Battery voltage correction with resistor: .15ms/v
  • Flow rate: 584cc/min @ 42psi

Additional injectors that have been tested using this method:

Bosch 0280155203:
High impedance, Mercedes Benz 5.0l V8

  • .85ms @ 13.8v
  • 1.5ms @ 8.66v
  • Battery voltage correction: .13ms/v
  • Flow rate: 203cc/min @ 42psi

Denso MDL560P
Low Impedance, Mitsubishi Evolution VIII

  • .62ms @ 13.8v @ 38psi
  • 1.03ms @ 8.66v @ 38psi
  • Battery voltage correction: .08ms/v
  • .76ms @ 13.8v @ 38psi w/DSM injector resistor (6 ohms)
  • 1.28ms @ 8.66v @ 38psi w/DSM injector resistor (6 ohms)
  • Battery voltage correction: .10ms/v
  • Flow rate: 520cc @ 38psi

Tests on the Evo VIII injector were performed at 38psi and with a DSM injector ballast resistor to accurately represent the injector operating environment when swapped into the DSM platform.

RC Engineering 1000cc
Low Impedance

  • .68ms @ 13.8v @ 38psi
  • 1.06ms @ 8.66v @ 38psi
  • Battery voltage correction: .07ms/v
  • .83ms @ 13.8v @ 38psi w/DSM injector resistor (6 ohms)
  • 1.48ms @ 8.66v @ 38psi w/DSM injector resistor (6 ohms)
  • Battery voltage correction:  .13ms/v
  • Flow rate: 981cc @ 38psi

Tests on the RC 1000cc injector were performed at 38psi and with a DSM injector ballast resistor to accurately represent the injector operating environment when swapped into the DSM platform.

Bosch 280150804
Low Impedance

  • 0.76ms @ 13.2v @ 42psi
  • 1.18ms @ 8.66v @ 42psi
  • Battery voltage correction: 0.93ms/v
  • 1.0ms @ 13.2v @ 42psi w/6 ohm resistor
  • 1.82ms @ 8.66v @ 42psi w/6 ohm resistor
  • Battery voltage correction: 0.18ms/v
  • Flow rate: 340cc @ 42psi

Bosch 280150357
Low Impedance

  • 0.9ms @ 13.2v @ 42psi
  • 1.38ms @ 8.66v @ 42psi
  • Battery voltage correction:  0.11ms/v
  • 1.16ms @ 13.2v @ 42psi w/6 ohm resistor
  • 1.96ms @ 8.66v @ 42psi w/6 ohm resistor
  • Battery voltage correction:  0.18ms/v
  • Flow rate: 330cc @ 42psi

Bosch 280150762
High Impedance

  • 1.07ms @ 13.2v @ 42psi
  • 1.8ms @ 8.66v @ 42psi
  • Battery voltage correction: 0.16ms/v
  • Flow rate:  218cc @ 42psi

About nsfab

Nick Salyer (nick@nsfabrication.com)
This entry was posted in Technical. Bookmark the permalink.

Comments are closed.