Operation IDIocracy: Infrastructure Preparation – 38 Gallon Fuel Tank Mods, Engine Mounts, Suspension

We’re going to back up a little in time here, back before the full disassembly of Vantruck happened. This post covers a modification I wanted to make to it since I got it… and some other necessary behind-the-scenes work before the Really Heavy Thing happens. I’m structuring these posts more in a narrative package, deviating from my usual preference of chronological as-it-happens reporting from the line of contact.

Some of the stuff below happened during the fall, others in the middle of when the engine was propped up on wooden blocks, and so on. This megaproject has so many facets to it that I can’t just finish entire systems in one sitting like a robot build would have happen. I have to swap between tasks as parts arrive or as problems have to be solved, so more for my future reference, things are being documented more as groups.

How to use a 38-Gallon Fuel Tank with a 22-Gallon Fuel Gauge

One of the enduring “why did you do this” problems of older Ford trucks is the dual fuel tank system. Maybe the technology just wasn’t there yet to make bigger fuel tanks. Maybe there wasn’t a safe-for-back-then place to mount them. Or maybe the heritage of trucks as farm equipment meant there was supposed to be two different fuels used.

Whatever the case, if you search “ford fuel tank selector v….” the auto-completes all say “PROBLEMS” or “REPLACEMENT” or “NOT WORKING”. This is true: Not even Vantruck’s front fuel tank worked when I got it, and Spool Bus certainly doesn’t have a working one now either. The brilliant door lock motor in a plastic spool valve body that Ford used for this task (seriously, take one apart… ) tended to stick shut or leak, and the extra creative crossover wiring harness used to connect to it made for even more points of failure.

I guess I could have taken the easy way out and just left the single rear 22-gallon fuel tank while excising the front one and removing the valve from the equation. But that seemed awfully small for the long-distance cross-country loafing I am still pretending I will do.

Instead, I’ve wanted since day 1 to replace the dual tank setup with a single large one. This technically does exist, but not when Vantruck was manufactured. When Ford introduced the E350 cutaway chassis, it also introduced a 40-gallon single tank option. These OEM single tanks are very expensive and hard to find the specific fuel pump, bracket, and gauge assembly for. As per my usual M.O. for projects, the more people can enjoy them the merrier the whole party becomes. I decided to try and piece together a fuel system using parts you can get easily today.

On the left is a generic 38-gallon fuel tank for more recent vintage F-series trucks. That’s what they’re typically sold as, such as by Classic Diesel Designs. They’re also found on eBay and other random truck part vendors as well, so are probably made “overseas”. These are not made for the van chassis, which has a different frame rail spacing and straps, but that was part of my quest to find out what changes are actually needed and what parts have to be chopped up or made from scratch.

On the right is the stock rear 22-gallon fuel tank that I removed from Snekvan. Vantruck has the same one. I wanted to combine the 22-gallon fuel pump bracket/sensor gauge (still easy to buy) to adequately use and read the 38-gallon tank using the power of friendship and geometry.

I’m planning on taking some select measurements to determine where the fuel gauge floats are in each tank, and where they have to end up. Subsequently, I’ll be able to figure out where to bend the 22-gallon bracket to get it to play nicely with the bigger tank and read ideally the same fuel levels.

We begin by levelling the tank with some mill setup blocks. The 38-gallon tank has a bit of a micro-sump on the bottom, so it doesn’t want to sit level on a table. This level is important, since…

One of the things I had to do was collect the angle of each fuel pump mounting ring.

Next, I pointed a long ruler square down the center of the hole until it hit the bottom, on each tank. The machine parallel here acted as a good flat surface to take the measurement from. With the “origin heading and depth” of each fuel pump bracket known, I moved onto measuring the bracket itself.

This is the 22 gallon pump and fuel gauge bracket. The diesel fuel systems don’t use a pump in the tank itself, so I just didn’t order the one with an integrated pump.

It might be becoming clear that my goal was to measure the X-Y displacement of the fuel gauge float’s pivot point from the center of the fuel pump mounting ring. Then measure the angles that it swings to. With the previously collected “origin heading and depth”, I could draw up where inside each tank the float will end up, and then calculate from there how to bend it around.

So that’s what I’m doing. Depth measured downward from the bottom center of the metal cap (where it would sit flush against the surface I put the ruler square on the tank). Then displacement sideways to the pivot. Establish the chain of geometric custody!

All measurements were eyeballed and probably +/- about 1/8″ or so, by the way. Precision? What’s that?

Finally, the Angles of the Dangle were measured, once at lowest displacement (Empty) and once at highest (Full).

With these numbers, I made a sketch representing a 2D elevation-view of the two fuel tanks.

The green highlighted lines represent where the 22-gallon bracket sits in its own tank, and consequently where the float ends up. The dark purple lines are the measurements of the 38 gallon tank. The squiggle represents the bent feed and return tubing which I couldn’t be buggered to model correctly.

Both families of lines are referenced from the center of the opening (where I stuck the ruler square down as a depth gauge) and notice they’re set at the angle which the opening faces upwards at. That way I have a common point of reference between them.

I found out that when the float arm is straightened out all the way, it only measured 6.5 inches.

This was not enough length for me to solve this the way I wanted, which is to cover as much volume as possible. Short of remaking the entire arm and float, the angle of swing being limited meant I could pick between having “Off-scale Full” or “Off-scale Empty”.

In that situation, I’d rather have “Off-scale Empty” because of the extra padding it gives me if I actually do run it to the E line, such as not being able to find a station with diesel fuel. It would mean the gauge bottoms out, but I still have a bunch of fuel left in the tank.

How much left would be indicated by the ~2.3 inch dimension at the bottom. Given the size of the tank in length and width, this can be calculated:

Good ol’ Wolfram. In this case, I’d have what is supposed to be 7 gallons left at “Empty”. However, this measurement is to the bottom of the miniature sump-like bump at the bottom, and is not representative of the depth of 90% of the tank. So I think it’s more like 5 gallons.

Still a reasonable safety margin for a 38 gallon tank, I say.

To double check the dimensions I arrived at with the sketch, I punched in the X/Y size of the tank and how much total height of fuel there should be inside: About 2.3 inches at the bottom, coupled with the swing of the float, yielded about 11.7″ give or take.

I assume once again that the fuel tank is not fully rectangular (it has rounded sides) so I’ll just say the actual liquid volume occupies 90% or so of the full “bounding box”. We arrive at 37.4 gallons, which is close enough to 38 gallons . The 90% factor was solely to make myself look smarter on the Internet. So the tank will have ~5 gallons in it when “EMPTY” is reached, and the fuel gauge will read approximately 32 gallons of the rest.

Now, Vantruck’s fuel gauge already goes all over the place and is known to read as much 1/4 tank too LOW (causing me to feed it more, of course…), so NONE of this actually matters unless I replace the gauge with a known good one! But it made me feel better, and look smarter. That is really the central theme here.

Onwards! With the numbers roughly cross-checked, I went ahead and bent out the float arm to the length indicated by the design. I then applied a gentle bend near the base to change the “start angle” to what I sketched out as well.

Time for physical validation. I installed the modified 22-gallon bracket with the locking ring, and peeked through the fill hose port to see where things ended up.

The float is “EH, ABOUT 2 INCHES” from the very bottom of the micro-sump bump lump at the bottom (about 1 inch from the main bottom plane of the tank).

I used a string through the top vent port to pull the fuel gauge arm upwards to see where it ends up, and it was “EH, CLOSE ENOUGH TO 11.5 INCHES OR SO” from the same point as measured by staring at it, then at a ruler placed on the table vertically outside the fill port.

Look, I am doing some serious CERN and NASA level science here, you understand?

As a very last check, I made sure that I was indeed installing a 22-gallon fuel gauge for gas engined Ford trucks, which read about 10 ohms at the bottom and 70 ohms at the top.

Yes, gas engine fuel gauge. Because Vantruck was originally gas, not diesel! I wasn’t able to find a certain answer on The Internet of Car Guy Advice on if the OEM diesel gauge and needle calibration were actually different. The gas part is also infinitely easier to find and buy, so if I can make it work, then it’s a net win.

So what’s the final verdict here? Take the 22 gallon fuel pump and float assembly, straighten the float arm all the way (gently, ideally with a bit of heating because that spring wire is pretty work-hardened) to 6.5 inches center-to-center, then bend it downwards near the pivot 5 degrees from stock (or leave it alone… the different this makes is basically more pedantic and mathematical)

I used a Carter #P74544S for this exercise from RockAuto which cross-references Ford part number #E6UZ-9H307-BA. It seems pretty genericized, with even random Amazon off-brands available, so it should be repeatable by anybody else.

With that adventure over, I turned to installing the mini-sump system. This is also a part sold by Classic Diesel Designs and available as a generic medication as well. The OEM “pull from above” fuel system is more prone to air infiltration, whereas the sump feed system will ensure that the intake is under the fluid level at all times. You simply hole saw a 3″ hole into the bottom of the tank and it clamps from both sides with gaskets in the middle.

Editor’s note: In fact, it worked so well that on a local day trip with friends in June 2023, I ran the fuel system so empty it completely deprimed itself. This will literally pull every last drop out of the tank and the feed lines, and then empty the filter as far down as it can reach. It took two jump starts and a minute of holding the air purge solenoid open before it woke back up. I wasn’t actually paying attention to my hacked-up fuel gauge, completely ruining the purpose of making it.

This is one of the times when being a manlet works in your favor when servicing vans. Besides reaching awkward buried-at-the-factory screws with ease, no red-blooded grain-fed American bicep is gonna fit down this hole to install the clamping plate!

I pointed the feed and return fittings in the direction I needed and cranked down on the screws. That’s all there was to it!

Preparing the engine mounting crossmember

The 460 engine mounting brackets are held onto the frame rails by a weird combination of screws and rivets. What? Yeah, I don’t know why either. The front portions of each are riveted into the engine cross-member, but the backs are secured by hex-head screws going into weld nuts on the plate itself. I’m not sure how this was easier or simpler or quicker to do on a moving assembly line. But they had to come out in order to later install the 7.3 IDI mounting plates.

I had to drill those big rivets out, so I made a centering jig using an aluminum round. It has a 3/8″ hole drilled all the way through for my drill bit, but also a large dimple I made with a countersink on one end. This dimple roughly seats it on the center of the rivet head, allowing me to start drilling it out on a rounded surface.

When the rivet is drilled hollow, a whack with a cold chisel removes it completely. I then cleaned up the entire crossmember with teratogenic solvents and tool-mounted brushes.

I gave the whole thing a coating in rust converter compound, then later followed up with my favorite “Eastwood Goo” surface coating. It’s not properly painted or anything, and I didn’t really even treat any other areas. This is not a comprehensive chassis restoration, just making it slightly less poopy.

On the other hand, the engine and chassis brackets were hit with rust converter, cleaned, then actually painted over. Installation of these Bracketry will occur after the engine is roughly in place. They’re fully bolt-in now, so I wanted the flexibility of looking at how everything has to jiggle and shift to line up as I’m lowering the engine.

Upgrading to ambulance springs

The IDI engine weighs around 200 more pounds than the big-block 460. The E4OD is also heavier than the old C6. Internet Knowledge recommends upgrading the front coil springs to a heavier type and also upgrading the shock absorbers if you’re stupid enough to do this, so I went ahead and did both.

I purchased a set of chonkier springs sold for the ambulance/RV chassis (like Econocrane has). These were going to be paired with some Monroe shocks, also RV chassis spec. I used Husky Springs #RV860 for the coil springs and Monroe MAGNUM RV *flexes and bulges veins* shocks.

Check out the length difference! I’m sure Vantruck’s springs had sagged over the years too, and I hoped the extra stiffness would help raise the front ride height some. It otherwise has a bit of “drag rake” that makes it look a little silly.

I bought a set of coil spring compressors for the install, and let me just say that using these was the most terrifying thing I’ve ever done in my life… and I’ve definitely done terrifying things.

These were what showed up when I ordered “coil spring compressor for truck”, and I think they’re designed for little twinky springs like on Corollas or something. They were rather visibly bending. But they held on! I had some help come over for this and we made sure to wear face shields as a symbolic gesture, as if the 1.5mm thick acrylic will slow down a flying spring compressor shard. This was roughly the equivalent of hiding your face under a leaf as you eat a whole bird at once.

With these preparations complete, there was only one thing left to do. On the next chapter, THE STUFFENING will happen!