Question on automotive fuel pump relay control, need to convert pulsed signal to steady +12v

Thread Starter

thebvk

Joined Nov 1, 2018
18
Hi All,

I have a basic problem that im sure someone here could very easily help me solve. For a long list of reasons not worth going into I've been having hard start issues on my viper. It's something of a known issue on these, the car (and most cars) have a safety feature that only allows the fuel pump to run for a brief period of time on initial ignition on, then only run if the PCM (computer) detects engine rotation, either cranking to start or running. In my case that is a 1-second prime for fuel pump, and its not long enough to start easily without priming it multiple times (switching key on/off etc).

The PCM triggers the fuel pump by connecting ground to the fuel pump relay. Other than the initial 1-second priming time on initial ignition, it uses the crankshaft position sensor to determine if the engine is rotating and thus provides ground to this circuit. This is a safety feature, so that when you crash the stupid thing into a wall and the engine stops the fuel pump doesn't keep pumping fuel, fireball etc.

I've extended that initial prime time by using an ELK-960 delay timer, works great. However, the PCM is now no longer providing that ground circuit, even when the car is returned to default state (no delay timer etc). I'm certain its a PCM issue and I don't want to buy a new one, as this seems to be the only issue with the car now.

I can make the fuel pump work by just grounding that pin such that the fuel pump is always running with the start/run circuit, however that is now an issue in the case of a crash, and i want that safety feature.

What I want to do is to replicate that safety feature. Crankshaft position sensors put out a waveform output only when the engine is rotating. I want to use that input as a signal to connect the ground to the fuel pump relay, effectively mimicing what the PCM should be doing - supplying the ground to the fuel pump relay as long as there is engine rotation.

I haven't measured mine (just have a voltmeter, dont have an oscilloscope) but I assume the output of the crank position sensor is something like described on this page: https://www.picoauto.com/library/automotive-guided-tests/crankshaft-inductive .

damn, its still a wall of text, sorry. Can I use another ELK-960 to solve this? datasheet for it is here: https://www.elkproducts.com/product-catalog/elk-960-delay-timer-module. I can set it to free run as long as there is a signal on the trigger input.

trigger voltage can be as low as 4.5v and <1mA on the ELK-960. I'm not sure if that will do the trick or if I need something to use that same trigger signal to push out a steady +12v that I can use as a signal, like a basic resistor/cap setup in front of a relay? (yes I know thats more or less what the ELK-960 is).

sorry for the long post, any help appreciated. I'm not an auto electrician, nor an expert in electronics, as you can tell.
 

drc_567

Joined Dec 29, 2008
922
... just curious as to the condition of the fuel injectors. Inefficient fuel flow may be contributing to the hard start problem. Either ineffective spray pattern or low flow volume could be checked by an injector specialty company.
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
i have changed fuel filter, and just spent 30 bucks a pop on getting all 10 injectors fully serviced, cleaned and balanced.

these were the details I didnt need to get into, but its a common thing on these cars (Gen II GTS). Essentially what happens is either the pressure check valve (on return), the fuel pressure regulator, or the pump itself goes bad or leaks a bit and then this issue shows up because you only get JUST 1 second prime time on the pump. I know this as its not fuel, fuel filter, or injectors (all replaced etc) and I've installed a fuel pressure gauge. its clear whats happening, its just the initial prime time that needs to be a touch longer, once running there are no fuel pressure issues.

why not fix the check valve or the fuel pressure regulator you ask? Well, on these cars those are all contained with the fuel pump in a module that fits inside the tank. To get to it you need to cut upholstery and drill out rivets on the cover plate, and then there are no servicable or replaceable parts in the module. You need to replace the whole module. They don't make the module any more. 2nd hand ones go for around $2000, if you can find them. At that point you might as well re-fuel the car with a new system.

Hence wanting to extend that primer time. Then I broke something and the PCM stopped doing it, and here we are in this thread. Thanks for the reply!
 

gerty

Joined Aug 30, 2007
1,292
Not an expert on this. but I believe if you try to tap onto the crank sensor you'll start another series of problems by distorting the sensors signal, If I had to do this I would opt for triggering the timer off the "start" signal to your timer. This would possibly cause a little longer crank time, as the pressure wouldn't start building until the starter was actually turning, All the rest would remain the same.
 

shortbus

Joined Sep 30, 2009
8,056
Are you sure it is the crank sensor that tells the ECM that the motor is running? I know that on GM cars it is or was the oil pressure that did that. So possibly your looking in he wrong direction? I'm a GM guy not a Mopar guy so just a guess.
 

ArakelTheDragon

Joined Nov 18, 2016
1,350
I am not that good on inner combustion engines. But if you want a permanent "12VDC", that means a capacitor or bettery rechargable battery between the input and output.

If you have the possibility of using the unified diagnostic service, that can help.
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
Are you sure it is the crank sensor that tells the ECM that the motor is running? I know that on GM cars it is or was the oil pressure that did that. So possibly your looking in he wrong direction? I'm a GM guy not a Mopar guy so just a guess.
yes its the crank position sensor. I have all the workshop books on the car. I suppose you could use any number of other indicators - camshaft position sensor, possible oil pressure, but that wouldn't work when its cranking


Not an expert on this. but I believe if you try to tap onto the crank sensor you'll start another series of problems by distorting the sensors signal, If I had to do this I would opt for triggering the timer off the "start" signal to your timer. This would possibly cause a little longer crank time, as the pressure wouldn't start building until the starter was actually turning, All the rest would remain the same.
I'm not sure, the trigger input on the ELK-960 doesn't seem like it would interfere with the crank sensor output... but I don't really know. You need to prime the fuel system before cranking, if you don't have it fully pressurized before you start it just cranks and doesn't fire, as the injectors are pulsing also when you crank and not allowing pressure to build up behind them.
 

Alec_t

Joined Sep 17, 2013
11,749
Is the crankshaft position sensor an inductive type or a Hall effect type?
Is the ELK configured to be triggered by a positive pulse or a negative one?
Will your insurance company accept a modification of this sort?
 

crutschow

Joined Mar 14, 2008
25,675
I'm not sure the ELK-960 will do what you want (a steady output as long as pulses are detected).
What you need is called a "missing pulse detector" circuit.
Can you build a simple circuit or do you need to purchase something?
 

shortbus

Joined Sep 30, 2009
8,056
yes its the crank position sensor. I have all the workshop books on the car. I suppose you could use any number of other indicators - camshaft position sensor, possible oil pressure, but that wouldn't work when its cranking
Like I said, not a Mopar guy. GM keeps the pump running until the fuel regulator reaches full pressure, and if it doesn't start and make oil pressure it will signal the pump on again after the regulator pressure goes low again. Makes more sense (to me) to do it that way rather than shutting down pump when motor just cranks.
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
Is the crankshaft position sensor an inductive type or a Hall effect type?
Is the ELK configured to be triggered by a positive pulse or a negative one?
Will your insurance company accept a modification of this sort?
I believe its induction, will dig through the service books and check
The ELK can be configured for either positive or negative triggers
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
I'm not sure the ELK-960 will do what you want (a steady output as long as pulses are detected).
What you need is called a "missing pulse detector" circuit.
Can you build a simple circuit or do you need to purchase something?
thanks. I'll do some research on that. It's been 20 years since i've built any circuitry, but I can if I have to, preference is to buy something off the shelf if it exists.
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
Like I said, not a Mopar guy. GM keeps the pump running until the fuel regulator reaches full pressure, and if it doesn't start and make oil pressure it will signal the pump on again after the regulator pressure goes low again. Makes more sense (to me) to do it that way rather than shutting down pump when motor just cranks.
really? that seems dangerous. what happens when you crash and the engine stops but ignition is still on, and the fuel line into the manifold is ruptured? In that case the fuel pressure would drop,does the fuel pump keep running?

but yeah it would probably have solved this issue
 

Thread Starter

thebvk

Joined Nov 1, 2018
18
ok, new question, new idea.

an oil pressure sensor would effectively work the same way as an indicator of engine rotation. linear output voltage, usually 0-5v

could I use that signal, or amplify it to serve as a trigger input to an ELK-960? minimum positive trigger voltage is 4.5v, >1ma
 

crutschow

Joined Mar 14, 2008
25,675
ok, new question, new idea.

an oil pressure sensor would effectively work the same way as an indicator of engine rotation. linear output voltage, usually 0-5v

could I use that signal, or amplify it to serve as a trigger input to an ELK-960? minimum positive trigger voltage is 4.5v, >1ma
Will the oil pressure be sufficient under cranking to give a signal that you can detect?
 

Picbuster

Joined Dec 2, 2013
1,024
Hi All,

I have a basic problem that im sure someone here could very easily help me solve. For a long list of reasons not worth going into I've been having hard start issues on my viper. It's something of a known issue on these, the car (and most cars) have a safety feature that only allows the fuel pump to run for a brief period of time on initial ignition on, then only run if the PCM (computer) detects engine rotation, either cranking to start or running. In my case that is a 1-second prime for fuel pump, and its not long enough to start easily without priming it multiple times (switching key on/off etc).

The PCM triggers the fuel pump by connecting ground to the fuel pump relay. Other than the initial 1-second priming time on initial ignition, it uses the crankshaft position sensor to determine if the engine is rotating and thus provides ground to this circuit. This is a safety feature, so that when you crash the stupid thing into a wall and the engine stops the fuel pump doesn't keep pumping fuel, fireball etc.

I've extended that initial prime time by using an ELK-960 delay timer, works great. However, the PCM is now no longer providing that ground circuit, even when the car is returned to default state (no delay timer etc). I'm certain its a PCM issue and I don't want to buy a new one, as this seems to be the only issue with the car now.

I can make the fuel pump work by just grounding that pin such that the fuel pump is always running with the start/run circuit, however that is now an issue in the case of a crash, and i want that safety feature.

What I want to do is to replicate that safety feature. Crankshaft position sensors put out a waveform output only when the engine is rotating. I want to use that input as a signal to connect the ground to the fuel pump relay, effectively mimicing what the PCM should be doing - supplying the ground to the fuel pump relay as long as there is engine rotation.

I haven't measured mine (just have a voltmeter, dont have an oscilloscope) but I assume the output of the crank position sensor is something like described on this page: https://www.picoauto.com/library/automotive-guided-tests/crankshaft-inductive .

damn, its still a wall of text, sorry. Can I use another ELK-960 to solve this? datasheet for it is here: https://www.elkproducts.com/product-catalog/elk-960-delay-timer-module. I can set it to free run as long as there is a signal on the trigger input.

trigger voltage can be as low as 4.5v and <1mA on the ELK-960. I'm not sure if that will do the trick or if I need something to use that same trigger signal to push out a steady +12v that I can use as a signal, like a basic resistor/cap setup in front of a relay? (yes I know thats more or less what the ELK-960 is).

sorry for the long post, any help appreciated. I'm not an auto electrician, nor an expert in electronics, as you can tell.
Please correct me if I did not understand your question.
You have a pulse the moment the pulse exists you want an output like 12V.
When no pulse the output will go to low after a certain time (say 10 seconds).
This implies that:
When a pulse arrives within the selected time the output is high.

If this is what you want take a counter and count to whatever seconds you want.
Reset counter with the pulse.
That's all.

Picbuster
 
Top