$20 fuel injector tester

[Dino] is an auto mechanic and needed a way to test out fuel injectors. Commercially available tools start well over $100 and go up from there, but he built his own for about $20.

The injectors have a coil in them that needs to be tested. His design calls for a series of 0.008 millisecond pulses to test the coil. He started by setting up a 555 timer to output a one second pulse. This signal is fed into a second 555 chip that outputs the 0.008 pulses and in turn actuates a MOSFET to switch the coil on and off. To use it [Dino] connects to a 12V bench supply and to the injector, using a single button to start the test. See him explain the setup in the video after the break.

Comments

  1. TheBadWolf says:

    This is awesome beyond belief.
    Loves it

  2. andrew says:

    nice job

  3. GoJimi says:

    Excellent use of simplicity and cheap components to solve a complicated issue. Love it.

  4. mrasmus says:

    Excellent video, nice design; the clarity of the design is nice, and it serves as a nice demonstration of the mono- and a- stable modes of the 555. I also like his *exceedingly* clear process — from design all the way down to laying out the perfboard prototype, no steps are left unexplained, and all of them demonstrate a strong, clear understanding of the engineering behind them.

    I particularly appreciated his breakdown of each step of the signal — it’s a great demonstration of how compartmentalization in prototyping and design can prevent puzzling challenges; it’s a skill that so many people getting their starts in electronics lack, so it’s great to teach it along with breadboarding and basic soldering skills. I’ll be keeping this video link as a demo of 555 circuits for friends in the future, for sure.

  5. haxorflakes says:

    i like the self repair manifesto on the back wall from i-fixit.

  6. Peter says:

    The pulse width (~6us) seemed a little short for a mechanical device. I find it difficult to believe that a large mechanical relay (or fuel injector) can actually work with a pulse in the us range. The wikipedia page on fuel injectors (http://en.wikipedia.org/wiki/Fuel_injector) (about 3 quarters of the way down) suggests that the correct units would be ms (three orders of magnitude longer).

    Otherwise, nice hack/tutorial. Can’t agree more about the temp controlled soldering iron– bought a Hakko soldering station 6 months ago after owning a few pencil types over the years. Hands down the best $80 I have ever spent on any tool.

  7. Dino says:

    You’re right Peter. I did mention the times in the video in milliseconds… OOPS! :) It’s microseconds.

    Thanks for the comments guys.

  8. malikaii says:

    Dino, I believe Peter was suggesting it be the other way around. The time should be in milliseconds, not microseconds.

  9. Dino says:

    Right you are! My brain hurts! lol
    Thanks.
    It outputs 8 millisecond ON pulses followed by 7 milliseconds OFF for a duration of 1 second so that comes out to a total of 66.6 pulses.
    Right? :) 8+7=15 which is one cycle 1000/15=66.6

  10. jc says:

    May want to add a resistor to the power LED in the schematic (no doubt someone will build it exactly as it is shown). I suspect the LED was added afterwards as I didn’t see it in schematic in the video.

  11. jc says:

    Oh, and thanks for posting! Very cool circuit.

  12. Garreth says:

    I really enjoyed your video. Probably I’ll never need such fuel injector tester, however I love the way you described the process of designing and building it. It kept me really interested in the video. I hope you’ll get oscilloscope one day, which would let you make more interesting projects:)

  13. malikaii says:

    I never thought to use a PCB layout on a protoboard before. That’s genius. I usually end up with jumper wires all over the TOP of the board, instead of it looking like a PCB on the bottom. Great work.

  14. Cynyr says:

    http://www.dprg.org/tutorials/2005-11a/index.html A way to make an adjustable duty cycle PWM generator with a a-stable 555 timer and some bits. Works decently to control the speed of a DC fan.

  15. Simon says:

    I really appreciated the work and attention to detail that went into this project write-up. Thanks very much Dino!

    How come the green pulse LED doesn’t burn out when the injector closes? I see a spike of several hundred volts on mine.

  16. gcat122 says:

    I second all the positive comments above. Good process and results. I would caution that the relay and injector will both cause a huge positive voltage spike on the transistor when it shuts off. The 100 volts or more will try and fry the LED and FET. A diode will work but a series diode and resistor or series diode and zener(10volt or so) will allow the field to decay faster and keep the injection timing close to desired values (if critical). Do injectors have these built in?

    Good work.

  17. Daley says:

    Kudos Dino – not many mechanics have electronics knowledge, even really good mechanics. As stated, you do a great job of simplifying, explaining, and above all UNDERSTANDING.

    Whatever you’re making as a mechanic doesn’t do you justice.

    Hat’s off!

  18. gcat122 says:

    To clarify, point the diode to the plus supply. Anode to transistor,cathode to plus supply. The zener in series with the rectifier would point the other way. Another choice is a zener to ground from the top of the transistor. Use 15v or higher zener in this case (higher than the supply voltage), with cathode to the top of the transistor. Check that your transistor voltage rating is higher than the zener voltage plus the supply plus a safety margin.

  19. Dino says:

    @ JC, good catch on the LED resistor. I corrected that. You’re right, it was added last minute.

    @ gcat122, the MOSFET has diodes built into it as shown on the schematic. Makes them very easy to use.

    @ Garreth, After seeing the video, someone has offered to give me a Tektronix 7633 oscilloscope they no longer use! :) How cool is that?!

    @ Daley, I’m hoping that one day soon someone will see what I can do and offer me a job where I can really earn what I’m worth, and do what I love.

    Everyone else, thanks for the compliments. :)

  20. gcat122 says:

    I checked the schematic and noticed a gate protection zener and a non-zener body diode. The IRF510 (from radio shack) has a source to drain zener built in so that would work well. Even then the 100V pulse is potentially bad for the LED and wandering fingers. Borrow a scope or run a free visual simulator at falstad.com. Very enlightening.

  21. Fallen says:

    Well done Dino, Kudos.

  22. Dino says:

    Seems I have a bit of troubleshooting to do.
    I have an injector here I was testing out and it only made one short pulse. I checked the tester again with the relay and it worked as it should.
    The relay coil winding measures 82 ohms. The injector coil winding measures only 14 ohms. I think the lower ohms on the load side is causing the monostable timer to time out. I’ll have to get in there and check some things out tomorrow.

  23. Neolith100 says:

    Excellent job Dino, great explanation and concise and audible information! Keep it up.

  24. Eric says:

    Very interesting project here. There are low impedance and high impedance injectors, which it seems you have ran in to. Are you going to add a way to switch between them?

  25. fartface says:

    I just use a signal generator to do that. run the output into a simple transistor circuit to handle the power and all done.

    Also that is a GOOD way to clean injectors by pumping nothing but injector cleaner through them. You also can balance a set of injectors by running for a set time and measure the amount passed on each. then pic the set of injectors that pass close to the same amount. Voila; a super expensive set of matched injectors for peanuts.

  26. OpCode1300 says:

    nice build. just one note..

    There are two basic types of injectors. Low impedance(high current, 4-7A) injectors with a resistance of around 2.5-3 ohms, and high impedance(low current, 1A) injectors of around 12 ohms.

    worth a note.

  27. Jake says:

    Hmm. I see that he mentions that flow testing is the only way to properly test the injector (also, you need to observe the spray pattern), but I don’t see him testing any injectors in the video?

    A better title of this project might be “How to use a 555 timer”. Without a self-constructed flow testing apparatus, this is nothing more than a 555 timer circuit, which can be found in any 555 datasheet.

    I’d like to see a write-up on his build of the flow tester. I built one with a cheap aftermarket fuel pump, pressure regulator, and mechanical guages. It works well, and I just use an old HP function generator to actuate the injectors – I thought that was an easy way to give me the freedom to dial in whatever frequency, duty cycle and duration I wanted, and I got it for $40 on good ol’ eBay :)

  28. Dino says:

    @ Jake My intent here was to build a device to trigger the injector in the car with a fuel pressure gauge connected and static pressure applied. You then observe the pressure drop and repeat the process on the other injectors. If one has a low or high flow, the difference will be seen in the pressure drop.

    Good idea using a function generator for your setup. :)

    I may build a flow tester one day, but this is a tool that auto techs can use now, and they can use my plans and build it cheap if they are so inclined. Units like this one are commercially available and look very much like the one I made.

    I solved the problem I was running into with that 14 ohm injector. I’ve added a 47 mF capacitor between the power supply inputs. Now it pulses for a full second as it should. The schematic has been updated.

  29. you know what the next project is going to be,
    Recycled fuel injector based printer :-)

  30. SparkyGSX says:

    Zener diodes are way to slow for such applications, unless you bias them first with a resistor, which limits the zener diode to a value less than the supply voltage.

    The “build-in” diode of a MOSFET is usually also very slow, and besides that, it’s in the wrong place. The current in the inductor doesn’t change instantly, so a fast diode should be placed such the it will conduct this current.

    The way the circuit is now, the MOSFET probably goes into avalanche breakdown every time it switches off, which can be acceptable if it is properly avalanche rated, and you make sure you keep the avalanche energy wel below the limits.

    I’d add a little resistor or inductor in series with the injector, to protect the MOSFET in case of a shorted injector. A separate power supply (diode in serie with a buffer capacitor) would be good for the 555′s, along with power supply decoupling capacitors (about 100nF) to prevent erradicate behaviour long supply cables (stray inductance), or a supply that will sag a little when the MOSFET turns on.

    Also, an under-voltage protection on the reset pin of one of the 555 timers would keep the circuit from working if the supply voltage is to low, which could destroy the MOSFET because it’s not fully turned on by the lower gate-source voltage.

    Finally, a small resistor (in the range of 2-20 ohms) in series with the gate would help dampen the (very likely!) gate ringing, which is also bad for the reliability of the MOSFET.

    A (large) diode in series with the power supply will prevent damage in case the supply leads are reversed, which could destroy the 555 timers, and the MOSFET (and quite possibly the injector) because a large current would flow through the body diode and the injector.

    There are just a few small extra components, which should make the circuit nearly foolproof.

    I don’t mean to be negative, it’s a nice effort, but I think it just needs a little more work.

  31. Jake says:

    @Dino

    I wouldn’t use static pressure drop to indicate the condition of the injector. It’s nice as a basic troubleshooting test, but it’s not neccesarily a good indicator of injector condition. I’ve seen many an injector with an acceptable flow rate, but a *horrible* spray pattern. I always pull them out of the intake and test them that way.

  32. Dino says:

    @ Jake
    This is intended to be used as a basic troubleshooting tool, NOT to test the overall condition of the Injector. I agree. There’s nothing like a visual inspection of the spray pattern to tell you what’s really going on, and sometimes I pull them and do just that.

    @ SparkyGSX

    Thanks a lot for the tips. :)
    Part of the reason for sharing projects is to learn more. I’ll be working on putting your suggestions to work.

  33. bluewraith says:

    I used to throw wrenchs around too.. but none of that matters to me. What I LOVE about this project is the attention to detail on the perfboard. I would have never thought to use a PCB editor to lay out something like that. That is definitely going to get some thought in my next project.

    A clean breadboard is a happy breadboard, but sometimes ya just need a bit of wire-porn to make it through the day.

  34. rider says:

    Hello..
    I tried out your circuit.
    First i used the LED and it worked well.
    But i’m having trouble with the mosfet driver.
    It doesn’t seem to drive my 14-ohms injector. I dint hear any clicks or whatsoever.
    Any suggestions on the mosfets?

  35. zachiepb says:

    So I know this is a little old, but if my math is correct this moves the injector as if an engine is operating at 20000 RPM? I am not good with changing circuits, just copying others. is there any way to slow it down to a normal level?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s