Fuel Rail Pressure Gauge is here for bombers

[Matt400]

So how many BOMB without gauges? Most will agree it’s a no-no. With fueling boxes turning up the pressure its stands to reason you should keep an eye on it. Also for duration box guys you will want to know if pressures are dropping too low.



Just the other day I made a WOT run on a box and recorded over 27,900 psi. which has until now gone un-monitored- yikes!

For me I want to see 26,000 as a comfort zone and never go past 27,000 on a pressure box.

On the other hand I recorded over 22,000 psi on a WOT run with the box turned off so for duration box owners I would think it should be a concern to see pressures drop below what the factory intended at those throttle and RPM settings, time to stack then.



I mentioned in another thread how its possible to use a SPA dual gauge as a fuel rail pressure gauge using the OEM transducer without taping into the fuel system and got a bunch of PM’s on how it works and how to do it so thought I would start a thread for all interested to chime in.



For newer members not familiar with this kind of gauge you can learn more here and they look like this:

#ad




First I have to thank dan. poitras for posting the OEM transducer specs from his Cummins wiring diagram which is:

• 0 psi = . 50 voltage return (VDC)
  
• 5801 psi = 1. 39 volts
  
• 10153 psi = 2. 06 volts
  
• 14504 psi = 2. 72 volts
  
• 20305 psi = 3. 61 volts
  
• 26107 psi = 4. 50 volts
  

The OEM transducer is fed by a 5v supply so its output is limited to 29,367 psi.

I contacted engineers at SPA Design in the UK (not www.spatechnique.com) that is distributor here in the US to find out what their gauges impedance is and learned they use an internal 100K ohm pull down resistor. I was concerned if this was not high enough so I measured the transducers output voltage with my DVOM at idle and various throttle settings all the way to WOT.

At each of the various throttle settings I introduced a 100K ohm resister and noted no change in the transducers output voltage confirming the SPA’s microprocessor type gauge would not skew the voltage signal being sent to the ECM.



In working with the folks at SPA I found it was possible to configure the gauge settings to calibrate it to the OEM transducer. For my testing I used a temperature channel but you can also use a pressure or boost channel to read rail pressure.

SPA assured me that while transducer manufactures may list “in print” a spec range such as above that they typically will operate at double their intended range and that the printed range is simply a guaranteed accuracy range. I don’t know about everyone else but all I care about is reading another 1,000 psi. still well under the 5v supply.



All this has SPA contemplating building a one off “custom gauge” if there is enough interest.

For those interested in one just post here and we will use this thread to gauge interest, from there I will contact TDR staff and possibly set something up.

I should hear from SPA this week on that possibility and its cost.



In the mean time I would like to address current SPA owners who would like to “convert” one of their gauges they already have.



Keep in mind the “custom gauge" would be a DG216 Boost / Fuel Pressure dual gauge with only its face plate changed to read Rail Pressure / Lift Pressure because that’s the only change needed for a real nice look.

If you are creative enough and your gauge is out of warranty you could bomb it by pulling it apart and making a custom face plate on your PC or have a print shop do it.



Ok. . here are the settings to do this with the Boost, temp, or pressure channel of a SPA dual gauge:



You will need your manual, if you can’t find it a copy can be downloaded

here



In the “Special Menu Section” of the Manual- the oFF and SPn are incremented in 0. 01V (10mV) steps.



Set the oFF to 0. 50V which equals 0 psi on the OEM transducer, set the SPn to 4. 0V which equals 26,107 psi and then FSC to 261 representing full scale, set dEC to 1 and it should be spot on. It does not really matter if the span and full scale figures are not at the very end of the range. As long as it has two known points that’s all that matters.

Because the SPA’s have only 4 digits you won’t be able read 1 lb. increments, only 10 lb. increments so 26,107 lbs. will be shown as 26. 10

Since you can’t program in the last 7 lbs your gauge will be in error by 7 lbs under or 0. 027%

If you want your led warning lamp to go off at 27,000 psi then set it for 26. 99 and the warning led will flash at a true 26,997 psi.



My tests were done with a VAC3. 1 and a Ramifier (not stacked). I wanted to also do a Quad box but got passed over again on the beta tests

If you are using another box you will need to make sure the box is not skewing the transducer signal all the way back at the transducer or your gauge will also be fooled.

To do this- measure the voltage on the center pin of the transducer while at WOT and insure it will go to at least 4. 0v with your box on. This requires a DVOM (Digital Volt Ohm Meter) with at least 100K ohms impedance, most are 10M. Place it inside the truck and just run your Leeds under the hood and drivers wiper arm. I grounded my DVOM under the dash.

Obviously anyone with no mods that tap the fuel pressure transducer circuit don’t need to do this.



Once you establish your box is not pulling that signal down you simply hook the SPA’s “sensor signal” wire to the center wire on the transducer, be sure to do it right at the transducer.

Channel one (top gauge) is pin #9 and Channel 2 (bottom gauge) is pin #3 on the SPA’s connector. You will find the location of those in the manual. The other 2 wires associated with that channel are left unused. My truck is an 03 and I don’t suspect they changed the transducer or connecter on newer trucks but you can double check. The wire to tap should have about 1. 36v at idle. Just use a “T” tap- does it void your warranty? That’s a question for “your” dealer.



This will work with a boost, pressure or temperature channel. The high level alarm on a boost and temp channel work great for a high warning led but if you use a pressure channel those only have a low pressure warning this is why a Boost / Fuel Pressure gauge is the best choice because it will allow you to set a “Rail” high level led alarm while the “Lift” channel would get a low level led alarm.



Now if all you have is a pressure channel to use and you want to also have a high warning led that can be accomplished using an external led, the gauges external alarm and a relay. I have that diagram in my head at the moment and will post it if someone needs it.



For duration box owners a low level led alarm for the rail be it internal or external won't work because these trucks idle at 5K psi and the pressure goes up from there so your alarm would be on under part throttle use unless you got creative and set up a relay to enable the alarm circuit only under WOT conditions. Probably more simple to glance at the gauge @ WOT and high RPM's.



Lastly take note of the viewing angle in the manual. I have found this to be very criticle (especially at night) for a nice crisp display. I have one in the SRT-10 A Pillar and is marginal as it points directly at my eye, I plan to add 2 more SPA's and move them all to an overhead pod, the pod is higher but the face of the gauge should be pointing below eye level making a nice improvement.

 

[rbattelle]

Outstanding work, Matt. I'm surprised that the 100k ohm impedence on the gauge is enough to prevent skewing the pressure signal to the ECM! I had planned on builing a circuit using a unity-gain amplifier in series with the gauge whose impedence is 10^12 ohm. Obviously, that's not necessary. I would suggest that soldering might be a better choice than a "T" tap.



I'd be interested in an SPA gauge that does this, but I don't need a dual gauge. Maybe I'll try duplicating your work with a Dakota Digital gauge...



Again, outstanding work.



-Ryan

 

[Lonestar]

I for one will definitely get one if they are made available. I don't know what all you just posted up there, but if I can get a rail pressure and fuel pressure gauge in one I'm all over it. COUNT ME IN Oo.



KEEP ON TRUCKIN'

Lonestar

 

[rbattelle]

On second thought, I'd buy an SPA gauge factory-labeled with Rail pressure on top and lift pressure below...



... on third thought, I just checked the prices on SPA gauges. I think I'll make my own instead. But they sure are nice lookin'!



-Rya

 

[JCleary]

I'd be interested in one. Sounds cool.

 

[Tomeygun]

heck ya, count me in... . the idea of a gauge that does both pressures would be awesome!!!!

 

[dan.poitras]

Matt

Count me in for a custom guage

 

[Rams-n-Hogs]

I like the idea of a gauge that does both pressures, also. I would think it would work fine with the BDDL, since that's not an external box.



Jim

 

[Matt400]

I would think it would work fine with the BDDL, since that's not an external box.

Good point, since the transducers out put wire is not taped into and is secure all the way back to the ecm that signal will be left unchanged.

 

[DLeno]

great stuff matt. my only caution here is that the sending unit is specified as a . 5 to 4. 5v device. since Cummins does not specify calibration data above 4. 5v, the accuracy of the gauge above 26,100 is really not proven. we can assume linearity above this point, but I wouldn't count on the accuracy.



the other thing to keep in mind is that the SPA gauge peak detector is very fast. Since the rail pressure is also VERY pulsy, you may be recording peak values and not average DC values. But a quick change of the SPA averaging will cure that question -- set it up for 1 second averages, and read the gauge real time instead of depending on the peak detect feature. Thats the only reason I haven't hooked mine up that way -- I learned from measureing low side fuel pressure (with the boost channel) that things jump around a lot and that the peak detector may not tell the whole story.



good for you on the on-off! When I asked SPA they wouldn't do that. so I went to Diprocol



just use averaging, keep in mind that numbers over 26,100 won't be accurate, and dont' use the peak detector for meaningful data. its the average rail pressure that is of interest anyway. the pressure relief valve responds to the average and won't blow in response to the short peaks.



The other thing I recommend is to understand the noise level present on the line. what AC component of the DC signal is there. consider working with shielded, twisted pair wire, which should reduce the noise considerably. especially for the digital gauge. that should at least lend more credibillity to the peak detect mode of the SPA.



ultimately, though, if you are getting consistantly higher numbers from the SPA peak detector than you read while driving and using 1 sec averaging, then you know that those high numbers approaching 28,000 are not accurate, either because of noise picked up by the SPA gague, or by the pulsy nature of the rail, reflected by the sender output voltage.



If you have any further info on the nature of the sender, pls share. my experimentation shows that the sender appears to saturate above 4. 5v measureing average DC values (not peak). and Bosch information I have been able to locate thus far also specifies the . 5 to 4. 5v limits of the sender. yea, its fed from a 5v supply but I haven't been able to verify accuracy above 4. 5v.



keep up the good work!

 

[rbattelle]

The other thing I recommend is to understand the noise level present on the line. what AC component of the DC signal is there. consider working with shielded, twisted pair wire, which should reduce the noise considerably.

Is there really that much noise?



How many bits is the SPA gauge A/D convertor? If it's 8-bit, then the least-significant bit is 0. 016 volts, which is a pressure of 96 psi. That means you can really only trust the readings to within 100 psi, no matter what the gauge says. Question is: is the electrical noise greater than 1 significant bit?



DLeno, am I missing something?



-Ryan

 

[Matt400]

great stuff matt. my only caution here is that the sending unit is specified as a . 5 to 4. 5v device. since Cummins does not specify calibration data above 4. 5v, the accuracy of the gauge above 26,100 is really not proven.

True but I did confirm it continues to read / climb in a linear movement upward as I turned the box up, I then backed up testing with my DVOM using shielded test cables and read the same data the gauge reported.

you may be recording peak values and not average DC values.

I did reference the peak values on my DVOM but did not think to change the gauge setting to 1 sec averaging, thanks for pointing that out.



I will give the last . 33v representing just over 28,000 psi a closer look, thanks for your advise. I might make mention also that the box I recorded 27931 psi with is reported capable of that from the Manufacture, it was just at that point on a WOT run up to 60 mph when I saw it, I backed off.

 

[GWoody]

I'd be interested in a rail pressure gauge, put mt down when they become available.

 

[ceaman]

Matt,

Glad to see you are making progress on the idea and I hope Im on the list. How about sourcing a couple of plugs from one of the pressure box makers and making this a direct plug in with no warranty issues? Id be happy to help research if youd like... .



My new truck is in and I should be hitting the road this week... let me know if you need some high mileage beta testing for any issues!

 

[Matt400]

Matt,

Glad to see you are making progress on the idea and I hope Im on the list. How about sourcing a couple of plugs from one of the pressure box makers and making this a direct plug in with no warranty issues? Id be happy to help research if youd like... .

That is an excellent idea! Please go ahead and see what you can come up with. The plug once sourced could be sent to SPA and added to their harness for a true plug -n- play set up.



SPA would supply the lift sensor and some ideas for mounting would be good if anyone would like to comment. I have read where vibrations are hard on fuel sensors, SPA stuff is high quality though.



No list yet as this is just the thread to gauge interest however I am keeping track of everyone.

 

[DLeno]

Is there really that much noise?



How many bits is the SPA gauge A/D convertor? If it's 8-bit, then the least-significant bit is 0. 016 volts, which is a pressure of 96 psi. That means you can really only trust the readings to within 100 psi, no matter what the gauge says. Question is: is the electrical noise greater than 1 significant bit?



DLeno, am I missing something?



-Ryan

I don't know how many bits SPA uses in their A/D. I was refering to analog noise from two sources:



1. electromagnetic noise introduced onto the line itself between the sender and the gauge.



2. natural, pulsing of the sender output itself due to the pulsy nature of the rail.



digital multimeters may read artificially high. I'm only pointing out that my own experiments suggest that the sender is not reliable above 4. 5v. short excursions above 4. 5v are possible but we really can't rely on a digital reading without first characterizing the AC component of that signal.



and by shielded twisted pair, I don't mean using shielded test leads. I mean using twisted pair from signal ground and signal at the sender, then connect shield to chassis ground at the gauge and open (not connected) at the sender. this separates signal ground from chassis ground from gauge power.



I actually hope that further work here will result in reliable data from the sender above 4. 5v. I'm just not convinced yet without going through the effort.

 

[DLeno]

True but I did confirm it continues to read / climb in a linear movement upward as I turned the box up, I then backed up testing with my DVOM using shielded test cables and read the same data the gauge reported.



I did reference the peak values on my DVOM but did not think to change the gauge setting to 1 sec averaging, thanks for pointing that out.



I will give the last . 33v representing just over 28,000 psi a closer look, thanks for your advise. I might make mention also that the box I recorded 27931 psi with is reported capable of that from the Manufacture, it was just at that point on a WOT run up to 60 mph when I saw it, I backed off.

well, lets be careful not to use circular reasoning. if the sender is not reliable, then the box mfg's numbers cannot be trusted either. unless of course they used independant mechanical pressure measurements by modifying the rail.



the noise issue can be settled by analyzing the DC line for its AC component. only then can the DC reading by the gauge or by a DMM be trusted.



the sender I think is a bridge resistive network with an active device to acheive linearity. the supply is 5v but I'm not yet convinced that the sender can go up to the 5v rail. some senders do and some don't.



I hope you converge on this. I came up short going down this road, but I would like nothing more than to have an accurate way to measure rail pressure. keep up the good work.

 

[rbattelle]

I'd bet the AC component is less than 1 significant bit. I think without knowing the number of bits in the A/D convertor in the SPA gauge it's not possible to make a determination whether the noise is significant or not.



I'm just speculating. It's all very interesting work.



On edit: the least significant bit is 0. 018 volt assuming an 8-bit convertor. I had used a range of 4v instead of 4. 5v.



-Ryan

 

[Matt400]

FYI the SPA uses a 14 bit A/D converter. As far as I know the Boost and pressure channels are configurable for longer averaging times, I have neither. The gauge I used for testing is a temp / temp dual gauge and the At is not an option.

The gauge to use would be a Boost / Fuel Pressure which you can increase the averaging if noise were an issue.

 

[rbattelle]

FYI the SPA uses a 14 bit A/D converter.

No wonder they're so expensive! That's quite impressive. So the least-significant bit is 0. 00027v. As usual, DLeno is right and you'd probably do well to consider some twisted/shielded pair.



There is a circuit that can strip off the DC signal, but I forget what it's called now. It's generally made as part of a low-pass filter arrangement. You might be able to use it to eliminate the AC signal altogether, since you're mostly interested in the DC component.



-Ryan