Adding a Fuel Pressure Gauge - PeachParts Mercedes-Benz Forum

mach4 · #1 12-04-2014, 08:13 PM

The trials and tribulations of OM617YOTA in a long but interesting troubleshooting exercise around solving a rough running engine in his Toyota repower project (see 617 weird behavior diagnoses beginning at around post 101 for background) has led me to the conclusion that a fuel pressure gauge can be a very valuable monitoring, early warning and troubleshooting component to a set of engine instruments. I currently have installed a suite of engine monitoring instruments that include TIT (turbine inlet temp), intake temp, oil temp, transmission temp, engine bay temp, boost pressure and vacuum pressure. A thread describing the development of my Arduino based data collection system and bluetooth enabled Android display system is described in the following thread - Engine Instrumentation Project The Android display also provides HUD functionality and navigation capabilities. Here's the current system in use.

I've decided to add Fuel Pressure to the system (as well as Voltage, but that's another subject) and this thread is a documentation of the thinking and the design and the installation of the components.

Since I'm already running boost, vacuum and temp sensors that use a 5v supply, it's trivial to add a three wire pressure transducer that uses the same wiring configuration. It's literally adding one wire to the Arduino to take the pressure signal. The code changes are probably a dozen lines of code on the Arduino and a more complex matter to add gauges and/or digital readouts to the display program.

My current research and thinking points to using the following pressure transducer which should be about $25 or so.

As a 30 psi device this will give me excellent resolution for a nominal 14psi fuel pressure. The device has a less than 1ms response time so there will be very little delay in the signal. Knowing that fuel pressures vary, especially at lower rpms, the plan is to average 10 samples every time a dataset is sent to the display, which means I shouldn't need a pulse modulator as might be required with a manual gauge.

The installation plan is to tap into the supply line to the IP to get actual fuel pressure at the IP. This will be downstream of fuel strainer, primary and secondary filters. Fuel pressures that deviate from normal will be due to a filter problem or a lift pump problem or a pressure relief valve problem and can be easily troubleshot. The pressure transducers are inexpensive enough that I might install a second one on the supply line to the primary filter thus giving absolute confirmation of the need to change the primary filter by showing pressure drop across the filter. (I've got 62k miles on my primary filter and there is currently no way to know when it will need to be changed - other than a drop in performance and potentially a side-of-the-road filter change)

I mocked up a potential mounting method on my spare engine. I drilled and tapped the banjo bolt for a 1/8 NPT fitting. The image shows a capillary tube from an oil pressure gauge installed as might be used with a remote mounting of the pressure transducer. More likely I'll mount the transducer directly, probably using a 90 degree adapter due to clearance to the dipstick.

That's the current thinking....subject to change as research and design continue - all of which will be documented here.

Diesel911 · #2 12-04-2014, 09:14 PM

I am curious to see what is going to happen.
Just for test I tapped into the same Banjo Bolt but uses a 1/4 Inch Fuel Hose to a liquid damened Gauge. The needle still moved around too much.

I think it was Funola that connected a Mechanical Gauge to simliar sized capillary tube run to inside of the Car and said it did not bounce. So there may be a need for a restriction.

Besides the Piston in the Fuel Supply Pump the Elements in the Fuel Injection Pump each expel Fuel back out of the same hole that the Fuel is fed into the Element with but it is expelled at a lot of velocity and causes pulsation and turbulence inside of the Fuel Injection Pump Housing.

So there is a lot of Fuel pulsation going on.

Sensors often have a restriction on them so that may be enough to dampen the pulsation.

mach4 · #3 12-04-2014, 09:32 PM

Quote:

Originally Posted by Diesel911

I am curious to see what is going to happen.
Just for test I tapped into the same Banjo Bolt but uses a 1/4 Inch Fuel Hose to a liquid damened Gauge. The needle still moved around too much...

So there is a lot of Fuel pulsation going on.

My working hypothesis is that the electronic sensor is a lot less sensitive than a mechanical gauge which is affected by inertia as pulsations occur. I've got at least three ways to dampen pulsations if they affect the readings. The first is in software as I mentioned above, by averaging a series of readings over some time period. The second is with a long capillary and/or a small orifice. And the third is a pulsation dampener such as Toyota uses on the fuel rail.

I, like you, am interested to see what happens. If it doesn't work I'm only out the cost of the transducer and a few hours of research/testing.

OM617YOTA · #4 12-04-2014, 09:56 PM

Fantastic. Ill be following this thread closely.

Lucas · #5 12-04-2014, 11:48 PM

Very cool. I've dreamed of dropping out my cluster and putting a complete digital lcd display. A nice juxtaposition to my mechanical motor. But Java is beyond me, I can do the classic "hello world" in c, not much else.

dieselbenz1 · #6 12-04-2014, 11:50 PM

Have you considered measuring the fuel pressure on the return line. I did that in the past but never tried on the supply side. Could double to tell you when your filter is plugging off.

mach4 · #7 12-04-2014, 11:59 PM

Quote:

Originally Posted by dieselbenz1

Have you considered measuring the fuel pressure on the return line. I did that in the past but never tried on the supply side. Could double to tell you when your filter is plugging off.

There should be no appreciable pressure on the return line as that just dumps back into the tank. On the other hand if I measure the pressure on the supply side after the filter, a reduction in pressure would indicate the probability of a clogged filter, though it could also be the lift pump or the pressure relief valve. If I elect to measure both before and after the filter I can discern the pressure drop across the filter due to clogging and eliminate other factors.

mach4 · #8 12-05-2014, 12:03 AM

Quote:

Originally Posted by Lucas

Very cool. I've dreamed of dropping out my cluster and putting a complete digital lcd display. A nice juxtaposition to my mechanical motor. But Java is beyond me, I can do the classic "hello world" in c, not much else.

I don't do Java either. I cheated and used Processing to code the Android app - tons easier and works very well.

mach4 · #9 12-05-2014, 03:28 PM

I was looking at my IP this morning thinking about pulling the fuel pressure relief valve to check the spring. On my spare engine the spring had "weakened" to 21mm and I was guessing that the primary engine might also need a "spring adjustment". Unfortunately, it would appear that someone previously had removed the valve and replaced it with a plain old banjo bolt.

As I understand it, the valve is designed to maintain 14 psi of fuel pressure at the supply line of the injector pump by restricting the fuel flow out of the return line. If the spring weakens the pressure drops. In my case, with no pressure relief valve it would seem that the pressure should drop significantly and approach zero and the engine should not run, or run very poorly.

Now I'm really wondering what my fuel pressure reading is. And wondering why the engine runs so well. The only anomaly I have is a tiny lope at idle until the engine gets up to temp - putting the car in neutral restores normal idle.

I'm also wondering what logic would lead someone to the conclusion that removing the pressure relief valve was a good thing.

Everything appears stock i.e. I see no evidence of an aftermarket pressure relief valve or anything else out of the ordinary.

Any thoughts???

dude99 · #10 12-05-2014, 03:39 PM

Unlikely that it was replaced. Most OM617 turbos don't have the adjustable fuel spring. I seem to remember that very early 617's might have had it. All om616 that I'm aware of have it though.

mach4 · #11 12-05-2014, 03:45 PM

Nevermind....

Mystery solved. This is apparently a replacement pressure relief valve. The number on the valve is 14174 13046 which brings up the part and also indicates that it is used on MAN, Renault and Fiat as well as Mercedes.

I feel better now.

OM617YOTA · #12 12-05-2014, 03:55 PM

I was going to say, that doesn't look like an ordinary banjo bolt.

Glad it turned out that you do have a pressure relief valve after all.

Junkman · #13 12-05-2014, 03:56 PM

We use a needle valve as s snubber in the Cummins world to install a mechanical FP gauge and avoid destroying the gauge via "water hammer". My understanding is that electrical gauges have the same issue.

For plumbing use 1/8" air brake tubing. I have 100' in the shop. I'll look for the source later but it was cheap. I could put some in the mail if you just need some but it is handy to use as vac lines also so you may want to get a roll. I think it was under $30 delivered. The place had constant tension clamps also.

OM617YOTA · #14 12-05-2014, 03:58 PM

http://ww2.justanswer.com/uploads/JCTech012/2010-05-25_215739_How_To_Build_a_High_Performance_Cummins_Diesel.pdf

Here is the article I read on building high performance diesel engines. Quote from the bottom of page 4:

"Now we will concentrate on pyrometers, manifold pressure, liquid-filled
fuel pressure and air and fuel filter restriction gauges. These five gauges work
in unison to monitor your engine as you drive. It's like having a portable dyno
in your cab at all times. With the gauges, we can diagnose ninety-five percent
of all power problems over the phone."

mach4 · #15 12-05-2014, 04:11 PM

Quote:

Originally Posted by Junkman

We use a needle valve as s snubber in the Cummins world to install a mechanical FP gauge and avoid destroying the gauge via "water hammer". My understanding is that electrical gauges have the same issue.

Interesting piece of data. I just ordered the pressure transducer this morning so we'll see what it performs like.

I'm thinking that the construction of the pressure transducer is so different with essentially no moving parts as compared to a relatively fragile traditional gauge, the thing should survive well in our environment....we'll just have to see.

I'll add the notion of a needle valve (essentially an adjustable orifice) to the options for attenuation/dampening as needed.