A rugged operational amplifier

Discussion in 'General Electronics Chat' started by ptuomov, Feb 23, 2009.

  1. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Hello group --

    A virgin poster here.

    I am looking for a rugged, fully packaged operational amplifier that can be installed under the hood of a car. The temperature ranges, noise, etc. might make this a challenging environment.

    The component that I am looking for is simple. Input voltage 0-5v. Output voltage 0-5v. Hopefully, input and output voltage should be close to the same, but I don't need huge precision. Then the ground connector and a connector to the car's main electric circuit that runs the power windows and whatnot. So I am looking for four wires altogether.

    Here's more about the application. The input 0-5v comes from a MAF signal. The output is going to the fuel pump control signal, 0-5v. I don't know how much current it draws, but probably not much -- it's just a signal for the pump controller, and it doesn't drive the pump itself. Because I have not been able to find out much detail about how the pump behaves and because MAF signal is used for many other important things, I don't want to screw with the MAF signal too much and was thinking about putting an opamp between. But now I need a component that can survive inside a car (hot, cold, wet, vibrations, interference, etc.) If what I am trying to do here makes no sense, please, please let me know before go further with this project.

    Anyone know a good component to use?

    Thanks in advance for your help!

    Best,
    Tuomo
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    An LM2904 dual opamp or LM2902 quad opamp should meet the environmental requirements, and will get close to the negative rail.

    But fooling around with your MAF sensor signal without knowing just what you are doing will mess with your fuel economy, emissions output, and could quite possibly destroy your engine.
     
  3. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Thank you for the suggestions. I'll check those out.

    This opamp is needed precisely so that I can't adversely impact the main ECU's. At least if I have understood the issue correctly, the opam will act as a firewall between the MAF and the pump controller, protecting for the case that the pump controller tries to draw too much current.

    I can't hook up a toaster, so all advice appreciated!
     
  4. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Thanks again for your answer.

    Can I push this a bit further?

    Is there a simpler, four pin version of this or similar opamp? The 0-5v signal in, 0-5v signal out, 12v, and ground. I understand that this would not be nearly as flexible as LM2904, but I just need this for a very specific purpose.

    Does there exist a cable for automotive use that has the opamp integrated? Three wires in from one way (0-5v signal in, 12v, and ground) and one wire out from the other end (0-5v signal out)?

    Suppose that I go with LM2904 and wire it to produce those four connectors. How do I protect it from interference, water, heat, etc? Do you guys have some standard tricks, like dipping the whole thing in an epoxy bath?
     
  5. wilson479

    New Member

    Feb 23, 2009
    5
    0
    I don't think you should be messing with this!
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Not that I'm aware of.
    Opamps have a minimum of five connections; +V supply, -V (and/or ground) supply, noninverting input, inverting input, and the output. Some opamps also have offset adjust inputs. Reading up on opamps in our E-book might help you to understand them a bit more:
    http://www.allaboutcircuits.com/vol_3/chpt_8/1.html

    Not that I'm aware of.

    Proper grounding and bypassing of the supply pins using capacitors are certainly issues to be dealt with. Automotive engine compartments are severe environments, both thermally and electrically. Correct routing of cables and shielding can be problematic. Protection of circuits against fluids and corrosion is sometimes done by encapsulation in epoxy or simply by conformal coating.

    If you have a tough time with a toaster, I suggest that you have a considerable amount of reading and bench experimenting to do before attempting surgery on your vehicle.
     
  7. italo

    New Member

    Nov 20, 2005
    205
    1
    AMPS are typically specified as 125 degrees c° I never bothered to check my engine compartment but on a hot day and the engine just sitting and idling i bet it surpasses that for sure or maybe not. So to begin with find the enviroment temperature first then worry about epoxy and other things.
    LETS NOT FORGET THE 40KV LEAKING All over the place. take a look under the hood at night you will get the the idea.
     
  8. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Thanks for your help.

    I have read many of the allaboutcircuits tutorials, including the above linked chapter. I think they are excellent, by the way. Still, I can't pretend to really know what I am talking about.

    I was looking for an opamp which would have the inverting input connected to the output by default. I think this would allow one do with four pins. All I need is isolation, making sure that the pump that uses the signal can't draw too much current from the MAF. Perhaps there's a better solution.

    Yes, I agree it is a difficult environment. That's why I am asking for help! The spark firing, turbochargers heating, etc.

    My hope was to find a simple, packaged opamp that is (or more likely was) used by OEM car manufacturers in the engine bay for some similar purpose. The underhood air temperatures for turbo cars can climb to 150F if one drives hard and then shuts the engine.

    For the hell of it, I used copy pasting to put together the attached proposal, but really what I am looking for is a packaged solution specifically for automotive use.

    Sure, I'll wait untill I am having perfect lightly toasted bagels before tapping that MAF! ;-)
     
  9. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    The device you wish -
    - is simply a piece of wire.

    A very interesting question is the relationship between the pump, running on a separate circuit and not controlled in any way, and the mass airflow sensor. Nor do they share any common power line.

    The sensor signal is an input to the ECU. The fuel pump always runs at full speed, with a diaphragm regulator controlling pressure. Fuel injection is done by the ECU varying pulse width to the fuel injectors.

    Do you know that your car does not follow the above controls?
     
  10. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    I would like to respectfully disagree. A wire AFAIK would allow malfunctioning pump to alter the MAF signal. What I am looking for is something exactly like a wire, but with isolation / protection.

    I would again respectfully disagree. Although you are correct that most fuel pumps in conventional fuel return system run at full speed all the time, it is not true for all pumps and cars.

    I have Fuelab Prodigy pump which takes 0-5v control input signal. With 5v input, the pump pumps the full 130 gallons per hour against a 100 psi pressure. At lower input signal voltages, the pump flows less.

    If I run the pump at full speed when the car idling, the return line will have to return (I am guessing) about 127 gallons per hour. This heats up the fuel and requires me to install a new fuel return line. Waste of money and hp.

    Instead, what I am planning to do is, with your help, to wire the pump to use the MAF signal as its control input. As the MAF signal increases with the mass air flow (and required fuel), this arrangement (with some tuning of the signal levels) would result in relatively little fuel unnecessarily pumped to the fuel rail, heated in the process, and then returned.
     
  11. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    In that case, you use the op amp as a buffer. You splice a wire into the line with the signal from the MAF and run it to the op amp separately. I would want to mount the op amp on a printed circuit board, and house it inside a rugged case like a Hammond R100-060-000. That case is thickwall cast aluminum, and watertight besides.

    There would be room for a voltage regulator in there, so you could power the op amp with 9 volts derived from the nominal 12 volt battery supply. A 78L09 would be plenty.

    With 100 PSI on the fuel rail, I hope you never lose a connector. I've seen several cars burn with just 20 PSI behind the rail.
     
  12. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Hmm.... I'm thinking that basing your fuel pump pressure/flow purely from the MAF may not be such a wonderful idea; you may wind up with fuel starvation/detonation and holes burned through your pistons when running WOT at low RPM. Under those conditions, your MAF signal output will still be fairly low, but fuel demand will be rather high. You might need to integrate MAP sensor output in with the MAF readings; even that probably won't be enough.

    There's also the question of response time from the pump; it's not going to be instantaneous. Any delays in increasing fuel flow will decrease performance, along with the risks of detonation/engine damage.
     
  13. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Thanks, I'll take a look at the regulator.

    By the way, do small "buffer" boxes like that exist on the market? It sound like a common problem, take in 0-5v at very high impedance and send out 0-5v with a buffer or "firewall" in between.

     
  14. ptuomov

    Thread Starter New Member

    Feb 23, 2009
    7
    0
    Apologies for drifting a bit off topic, but here's a quick response.

    I will tune the pump to provide enough fuel for say 5:1 air fuel ratio based on the MAF signal. The car will ran at about 10:1 at its richest during transient conditions, and return half the fuel. The pump will be supplying a flow margin of safety of 100%.

    MAF signal is very close to a sufficient statistic of fuel needed, and manifold absolute pressure doesn't add much above that.

    I was originally thinking about measuring the fuel pressure in the rail or fuel return from the regulator, and then running a splatco controller with their PID software to control the pump. This has the downside that if a fuel line is cut, the PID controller may turn the pump full on and 130 gph gets pumped to the fire. MAF is safer, as simply lifting the accelerator in case of a fuel line bursting will slow down the pump to a low level.

    The engine accelerates much slower than the pump, so I don't see a problem here.
     
Loading...