Simple Fuel Pressure Alarm Averts Diesel Disaster

If you could spend a couple of bucks on a simple project that might prevent a $2000 repair bill on your vehicle, you’d probably build it, right? That’s the idea behind this simple low-pressure alarm for a diesel fuel system, and it’s so simple it makes you wonder why the OEM didn’t do it.

We normally see [Bob Johnson] coming up with nifty projects (like this claw or this camera slider) that more often than not combine woodworking and electronics. But no tree carcasses were harmed in the making of this project. [Bob]’s goal is just to sound a warning and flash a light if the output of a pressure switch goes to ground. That indicates the lift pump in his Dodge Ram’s fuel tank has failed, which could lead to the sudden failure of the downstream injector pump for lack of lubrication by the fuel itself. His simple ATtiny85 circuit lives on a small perfboard in a 3D printed case and taps into a $30 fuel pressure switch. The microcontroller code enables a short delay to prevent nuisance alarms, and if the pressure drops below 5 PSI, [Bob] gets a chance to shut down the engine and disappoint his mechanic to the tune of $2000.

Maybe it’s planned obsolescence on the OEM’s part, or maybe it’s not. But kudos to [Bob] for a simple hack that averts a potentially expensive problem.

42 thoughts on “Simple Fuel Pressure Alarm Averts Diesel Disaster

  1. The problem is it only really solves one possible fault condition.
    What if something else in the car went wrong, like perhaps a wire snapped or one of the pistons broke free or the oil suddenly turned into cheese.
    How quickly can you reasonably expect to pull over and kill the engine considering once you notice the dial dropping its already failed.

    1. Huh? I think you missed the point. This is specifically to warn of low fuel pressure from the lift pump. The lift pump supplies diesel to the injector pump. If the injector pump runs dry you can ruin it.

      This just tells him off a low fuel pressure issue so he can shut the engine down before the high pressure pump gets damaged.

      1. This presumes that lift pump failures are common enough tor require a warning system all their own. This exposes a HUGE design flaw that should be fixed at the source.

        It is convenient and cheap to use the fuel to lubricate the injector pump, but if simply running out of fuel kills your vehicle (running the tank dry would have the same effect as a fuel pump failure) you have failed engineering 101.

        1. welcome to the automotive world. where if your timing belt breaks and you have an interference engine the pistons will happily crash right into the valves. or even when you have a timing chain, losing oil pressure in some engines means you lose pressure to the hydraulic tensioner and the chain will jump teeth, sometimes leading to pistons smacking into valves.

          1. Yes, think of it as not “designed to fail” but “designed to profit” from repairs.
            Failure of a Kevlar timing belt lead to a $5000 repair bill in an intrusive valve T6 Volvo engine kept on the manufacturer’s maintenance schedule.

          2. @andres Absurd comparison. You’re talking about mechanical failure caused by wear. We’re talking about mechanical failure caused by operator error. The former is unavoidable for a perfect case. The latter is utterly unacceptable.

          3. Also, don’t forget that hydraulic chain tensioners work only when the engine is running, so putting the car in gear instead of using the handbrake (in winter some like to freeze up and get stuck, real pain to get it unstuck, with the engine off it’s fairly safe to do this on small slopes), so doing this in said cars with a manual transmissions is a reliable way of needing a new engine ;-)

          4. @garbz, this problem isn’t just one of operator error. there are plenty of diesel trucks out there with inadequate lift pumps that fail very often. It only takes a very small amount of time without fuel to damage the injector pump. lifter pump failure is definitely a failure method that falls outside of operator error.

        2. Running out of fuel with a diesel vehicle is a significant problem, not like a gasoline vehicle. When you run a diesel vehicle out of fuel, you have to bleed the injector lines before the engine will run again. This is a messy PITA to do. Trashing an expensive injector pump is still an excessively bad consequence though.
          I know one person who owns a Dodge diesel truck and he has had at least one injector pump failure in the last 6 years.

        3. There isn’t really any other way to do it, because of the tight tolerances required in a high pressure diesel injection system. Modern diesels are pretty damn picky about the fuel quality, and you can’t run them on just any old lamp oil. All the old hat tricks like adding a quart of gasoline to a tank to dissolve waxes in the winter don’t work anymore and risk damaging the engine.

          Running a tank dry btw. can damage gasoline vehicle fuel pumps as well, which leads to weird behaviour like losing power at high RPM because the pump can’t keep up the fuel supply.

  2. I think once a piston broke free there would have already been a lot of damage done already – and the oil turning to chese – you just shouldn’t fill your sump with water.

    I’ll go out on a limb here and say a large proportion of modern car faults are failure of the sensors intended to detect faults

  3. I would add two parts to his circuit – a ten ohm, half watt resistor between the twelve volts and the C2 cap, and a tranzorb with a 25V rating across the C2 cap, to protect the 5V regulator, since auto/truck electrical systems can generate high voltage spikes under certain conditions. Without this, the 5V regulator could see spikes that exceed it’s input voltage rating.

    1. Adding parts ??? Roflol…. That guy uses a µC to read out a switch and activate a buzzer and you want to add even more parts ??? ? ? ? This is seriously going the wrong way….
      Connect a 12V buzzer directly to the switch… If that switch is not capable enough eventually add a relay or a transistor or similar.

  4. I used to have a Nissan that overheated consistently once certain conditions were met. It took me nearly two years to figure out the why of it but in the mean time I had to drive with one on the gauges once the weather turned warm.

    An alarm would’ve helped immensely and in retrospect would’ve been trivial to put together. As it was, no garage in that entire time could figure out what was going on unless I shoved more money at the problem than my college budget would allow.

    For the curious, it was a fan shroud behind the radiater that disentigrated.

    1. Ha – had an old Nissan that flooded easily when it was cold (yet fuel injected!). They tried to fix it in software but still about 3-4x per winter it’d flood. I knew to pull the fuel pump fuse, crank till she rumbled to life then died, then replace fuse and restart. When I sold the car to my sister that was an issue, so there was just a switch installed under the dash she could turn off to cut the fuel pump, clear the flood, and then turn back on…. that thing made it way past expected lifetime until rust got the better of it….

      1. On most early fuel injected cars there is a ‘flood clear’ mode where if you crank at full throttle (accelerator fully depressed) it will stop firing the fuel injectors. Real useful trick…

    2. “For the curious, it was a fan shroud behind the radiater that disentigrated.”


      a year ago during summer I helped an older lady with an “overheating” engine problem. First I listened, what kind of circumstances of the failiure she described, and she got to the point where she mentioned “.. on the highway during highspeed everything is ok, everything is ok when I manage to get out of the city before the warning ..”

      This was when it was clear what to check first – the fan. (I also took a look at the coolant level, that was ok – visible from the outside – hot engine -> dont open when hot!, also the condition of the radiator was checked)

      Root Cause:
      The bearing of the fan had disintegrated and the fan was really stuck, no quick & dirty repair possible.

      If an engine starts to overheat during “city” conditions but not on the “highway”(in this case the “Autobahn” – ~130-150 km/h) its a contradiction. Because during city conditions the power that needs to be disspated, would be much lower than during high speeds. But on the highway the fan’s importance decreases because the wind speed will push cool air through the radiator.

      Long story short, I told her that the repair should be simple and cheap, and gave her some hints on what amount of money/time the mechanic should not exceed to bill.

      Tell mechanic: radiator fan is stuck – need to be replaced.
      abs. max work time: 0.75 h (6 screws, 1 plug, wiggle out, wiggle in, 6 scews, 1 plug, check fan fuse, short test run)
      parts: 15-20 € (used-ebay), 40-60€ (new)

      If your car starts to overheat during high speeds, check the coolant level & for leakages.
      (to have a 1.5 ltr. bottle of non-sparkling water at hand is good, because it can supply you with water and your car with coolant)

  5. brucedesertrat: You are correct, this was a problem that needed to be addressed. And there were many after-market warning systems that were built for the 1998 1/2 -2002 Dodge trucks. The problem really is the Bosch VP-44 injection pump was a poor design. If it lost fuel for even just a second, it destroyed itself. I saw it happen in two trucks, both just outside of warranty. Both times the repair bills came back around $4k after everything, excluding the tow truck bills.

    Dodge truck owners got the shaft between 1998 1/2 -2002. The lift pump was junk and the injection pump was junk. The lift pump could barely supply enough fuel to the Bosch VP-44 fuel pump under normal conditions, let alone under a heavy load. And given that the lift pump was unusually prone to failure, well, a lot of injections pumps got smoked.

    Dodge ultimately addressed it – with future models. They got away from the Bosch VP-44 fuel pump and rethought their lift pump configuration. Not only did the lift pump fail often, it was located inside the fuel tank, making it a bit more difficult to fix.

    1. Aren’t all modern (heck, even the old ones will get killed if they run dry) high pressure diesel pumps designed with lubrication by fuel? I can’t imagine how difficult it would be to give it a separate lubrication system…

      1. I think the pump uses the weight of the fuel create better head pressure but the type used doesn’t have good draw strength.

        I think they’re similar to those water pumps that have to be under water to work properly. The can’t draw much water over a distance but have good head pressure.

        I took a water pump apart and it used an (if I’m using the right name) impeller blade to suck water from all round and drive it into a small outlet. The blade can’t move air so it has to be under water to ensure there’s no air in the system. Looking at a diagram for a fuel pump, it looks like it’s the same design with the difference that the fuel also flows past the motor itself. Probably as a way to cool the pump I suppose.

        A pump that pulls water has a different design I think. The design of choice for that type seems to have poor head pressure which is what fuel injectors really need.

        I suppose engineers can design some sort of compromise (if they haven’t done so already) but the incentive to reduce repair costs for the consumers probably isn’t enough for it to go into the cars. Especially since the pressure to reduce costs is a far bigger incentive.

        1. Those are centrifugal pumps. Diesel injectors don’t work by centrifugal action because it’s completely open to back-pressure. If you block the outlet the pressure essentially drops to zero – it works by the throw of the water and if the water stops flowing the whole thing stops.

          The lift pump works that way, but the actual injector pump is some kind of piston pump.

        2. And the impeller you saw actually draws in water at the center and throws it around the side, where it’s collected up into the outlet by a spiral shaped scoop. The blade can move air, but air is lighter so it doesn’t have the same throw and it just gurgles around inside the pump. Once you get it full of liquid, the pump can pull just as high as it can push.

        3. Just helped a mate solve this very issue on a petrol-powered racer – it would run all morning and then stall in the afternoon. Turns out it only has the high-pressure EFI pump, which is not designed to draw fuel, only to push it. Adding an old low-pressure lift pump in the line to do the drawing & force some fuel into the EFI pump seems to have cured it.

          Another mate didn’t realise this and burned two Subaru fuel pumps in 5 minutes because he’d mounted them high up above the tank – first run up of the vehicle, pump #1 dies in ~2 minutes, flick to pump #2, 2 minutes later that’s toast too…

          TL;DR: There’s more than one type of pump.

  6. I wonder if for a belt and suspenders approach if he added a surge tank before the injector pump he’d
    a) get a warning from that alarm he has
    b) have a minute or two to get to a safe spot before the injector pump ran dry

  7. On my 2006 Ford Focus C-Max duratorq diesel (van type car) when the fuel filter was clogged – (at ~190,000 km) the ECU would cut engine power and idle the engine at 1000 rpm or shut it off completly – protecting the engine – reignition was possible during run.

    Funny: the ECU did not log an error, when I obd2ed the car had no errors – however the error condition was easy to interpret.

    1. Agreed, its a factory feature of decent cars!

      BMW diesels made in the last 10-15 years will limit engine power or cut it off if the fuel pressure gets too low. The ‘lifetime’ fuel filters are not a service item so will usually clog up and cause those symptoms at some point if they’re never changed!

      1. They get clogged because summer diesel is usually lower quality and has longer hydrocarbon chains in it (wax), which are simply dissolved in the mixture. When the temperature drops enough, the fuel starts to cloud meaning the wax begins to crystallize and form particles which don’t dissolve back into the fuel and instead clog up the fuel filter.

        Service stations change the fuel mixture throughout the year with higher quality fuel in the winter to prevent engine damage, but people still have leftovers in the tank when the weather goes cold, and in places where the weather really goes cold fast there’s cars with blocked fuel filters by the side of the road every single year.

        1. It’s true that refiners change up the fuel composition between winter and summer but it’s not because they’re trying to pass off an inferior quality product to consumers during the summer (they still have federally and locally mandated specs they MUST meet). Summer blends do tend to have more longer unbranched hydrocarbons that have higher cloud points (i.e. they have higher distillation end points) but this is a good thing because it typically means higher energy densities and lower volatility. During the winter additives and mixes with lower boiling point hydrocarbons (e.g. kerosene) means less energy density and thus lower miles per gallon for the consumer.

    2. “Funny: the ECU did not log an error, when I obd2ed the car had no errors ”

      Probably because, by law OBDII is only required to log/warn of emission related failures… a “dead” engine isn’t emitting anything! B^)

  8. At the risk of identifying myself as an “old Foggie” I want to respond to all of the comments about “planned Obsolescence”, “Designed to fail” and the gloom and doom failure scenarios that have been outlined. I am not disputing that these failure scenarios are real but I think more than most places, this group would appreciate the balance of cost vs reliability. When I got my first car, my expectation was that it would probably work for around 70,000 miles before the cost to maintain it was greater than the cost to replace. Now I would be disappointed if the economic replacement point is not closer to 300,000 miles and a quick google tells me the cost of a new car vs inflation in the US has basically remained the same for the last 50 years. If you agree with my numbers (3x longer life and no cost increase) then you should also recognize that strong engineering effort focused on reliability that the auto industry has made.

    1. I remember that 60’s and 70’s vehicles required an engine rebuild by that time (70-100K miles), if rust did take them out first. Yet I drove my ’86 Caravan to 180k and a Chrysler Concorde to nearly 160k. Neither had any rust, though the paint starts to go on Chrysler products after 10 years.

    2. Agreed. My first car was a 1972 VW Bug, which was not even 10 years old when I bought it for $800. I can’t remember how many miles it had on it, but it was a 5-digit odo and I’m pretty sure it wasn’t more than 50k. It was a wreck, but I kept it going until it just dissolved into the salty New England roads.

      Fast forward to last year, when I bought a 2004 Sequoia for $8000. It had “only” 80k on it, and I didn’t even blink about that. I still consider it the baby of my fleet, which includes a 1997 4Runner and a 2003 Tundra, both with in excess of 190k on them. All told, I have almost half a million miles on the three vehicles, for a total outlay of about $58k (purchase prices; doesn’t include maintenance and repairs, of which there have been almost none). That works out to a cost per mile of a mere 12 cents. I’ll take engineering that can deliver value like that any day of the week.

    3. Oh thank you so much for that – I’m not really old enough to remember those days (although I remember my Dad’s trials with rusty cars) but I did happen on a pile of old (1970’s) motoring magazines in a drawer some years ago and was greatly amused to read used-car buying guides where 3-year-old cars were deemed passable if you couldn’t poke your finger through the larger rust holes, engines were expected to maybe struggle to 100k miles with a rebuild or two but ultimately the thing had a 50% chance of being in the crusher by about its 5th birthday. Oh, and if you were driving more than down to the shops and back you had to check the oil & water before setting off…

      All these people who bitch and whine about the Big Conspiracy need to step back and look at the economics of the whole thing – what the average idiot actually cares about from their car, what big fleet buyers care about, and what’s actually practical for manufacturers to sort out.

      We have vehicles which are like frickin’ spaceships compared to 20 years ago, the average family motor has stuff as standard that was at the top end of F1 racing a few short decades ago, and will run for 100,000 miles of utter neglect without missing a beat or dissolving into oxide. Oh and if you crash it, you’ll walk away.

      Wake up sheeple!

      1. As automobiles were first being developed, tires only lasted 50 miles, engines did not have air or oil filters, crankcase oil was not sieved for viscosity.

        But cars of 40 years ago generally had thicker sheet metal, and therefore less likely to rust through than todays foil thin fenders.

    4. Well, yes, “strong engineering effort focused on reliability”, that came after the Japanese automakers took away the lunchboxes of the Detroit auto makers and ate it right in front of them, even then Detroit was more interested in limiting Japanese imports than building reliable vehicles (I’m still not convinced Dodge is interested in building reliable cars/trucks).
      I may be “an old Fogie” but I remember trading my 1973 Ford F-100 with its monthly repair bill of $125 and 12 MPG for a 1976 Datsun 610 pickup that got 26 MPG and needed less than $100 in parts per YEAR!

  9. Wow, I’m surprised Bosch designed a mechanical injection pump that’d fail after a few seconds of no fuel! Their VE distributor pumps certainly don’t, plus they lift their own fuel. Of course, you know when they suck air or otherwise lose fuel supply — the engine stops.

    While you’re not supposed to ever run a diesel out of fuel, it does happen sometimes for various reasons, and in running a number of Bosch and Stanadyne distributor style pumps, as well as Bosch and Zexel inline pumps, we’ve never had one die from sucking air.

    1. My Datsun 720 (small pickup) had the SD22 diesel engine, it starved of fuel several times, like when the diesel fuel gelled in cold weather. It didn’t have any fuel pump or injector failures.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.