Operation ENDURING BROWN: Well, This Smells Familiar

Work on the beleaguered U.S.S. BROWN C. STENNIS has been continuing at a rapid pace! For what really is the first time with one of my projects, I called for extensive backup from friends willing to lend some elbow grease. First, because some of them are more “car people” than me, and second, because everything on this thing is heavy. It’s like the Chevy of trucks: cheap and heavy.

Wait…. Hang on.

This post recaps all the events of about 2 weeks ago following its return from my mechanic; I elected to take it back since it had not yet been consistently diagnosable (i.e. it was becoming a “throw parts at it” situation) and I didn’t want to keep running up labor charges. I ended up going through a series of cross-checks and inspections and discovered the problem was all too familiar, but obfuscated by a compounding issue. In the end, it was literally entirely my fault, so I’m very satisfied; unlike most people, I’m only happy if I caused the entire mess in the first place.

Let’s begin.

Prodding Ford Truck Bro forums and groups led a lot of Internet mechanics to suspect an ignition problem. The ignition coil on these things is tucked snugly into the valley between the two V8 cylinder banks (seen above) and its electronic control module can also succumb to temperature-related failure. As I described it, where the engine will start and run for only 10-20 minutes, it sounded like a temperature-dependent electrical issue.

I had my doubts, because it would only some times start and run for 10-20 minutes, and other times have a hard time starting at all. But seeing as these parts were all fairly cheap, I decided to outright replace them just in case.

 

That’s the Ur-ECU ignition control module. I like how it’s just bolted into the wheelwell stamping like a good ol’ retrofit system that got put into production. “Looks good here! Build ’em like this!”

So this didn’t resolve anything. I started wondering about other electrical systems in the path of the fuel pumps. For some reason, there are a whole bunch, as illustrated by this handy diagram I drew in a fit of frustration:

I managed to locate and test the fuel pump cutoff relay but couldn’t find any trace of the fuel tank selector relay. Based on some more sleuthing, I deduced that the DPDT selector switch might have been wired directly to the fuel pumps. The object that says “NEW!!!!” all over it is this contraption:

Another one of those “Well why the hell did you do that?” parts on this thing is the fuel tank selector valve. It’s basically a small hydraulic solenoid valve that connects one circuit to two, but shittier and plastic. And it was the source of one of my troubles.

See, while bypassing the fuel pump power safety cutoffs, I did numerous impromptu fuel pump volume tests (because I wanted to see if it was pumping fuel with shortcut wiring). This led me to discover that some times, the fuel volume was low or nearly nonexistent. It would start off good, and then taper off. This actually corroborated a weird behavior I noticed where some times when the engine would start sputtering, I would wiggle the selector switch repeatedly for a few seconds and it would gain some run time. This was in fact one of the symptoms that made me think a power supply to the fuel pump problem.

Uh oh. Suspecting that this valve had been bad from the beginning despite me testing it on the bench  listening to it turn over, I performed another volume test upstream of the valve. Both pumps were giving the correct volume, so I scrambled back to Pep Boys and picked up a new valve and spliced it in. I don’t know exactly what kind of failure mode it is – maybe the mechanism inside is sufficiently worn or damaged such that it might travel fine, but can easily be pushed out of position by fuel pressure. Maybe that causes it to backflow into the un-powered tank, or maybe it just plugs up and sits there.

The recurring lesson I’ve been learning from VANTRUCK is “BTDMIW” :  But That Doesn’t Mean It Works. Just because I jiggled the component on the bench, or in isolation, doesn’t mean it actually is working correctly when installed.

But there was more! The fuel being pumped from the front tank, the one I personally serviced with a new pump and float  – was rather BROWN. It was darker than what I usually knew gasoline as. Even more telling was that there was sediment on the bottom of the jar.

I instantly knew what was happening. The next thing I did was run around to the front and start unscrewing the carburetor fuel inlet line:

…and I sheared it off in the process, because I untwisted the first big nut looking thing I saw, but it was really the nut that went into that nut. Brilliant!

I couldn’t even blow through that little metal sponge. It’s the “filter of last resort” for the carburetor, and it had trapped in it all of the rust slurry being pumped from the front tank. You can see some of the visible grunge on top.

Dedicated readers might recall that I also replumbed the fuel tanks in the same operation that I replaced the front fuel pump in. Even more dedicated (or observant) readers will also notice that I did not install an inline fuel filter on the front tank. Why? Who knows?! Maybe I’m just traumatized by fuel filters. Maybe I thought the fuel pump’s jizz sock thing was enough.

I patched this section of steel fuel line with some rubber hose and called it good.

Here it is installed behind the zinc-chrome part in the lower center. That is a “vapor separator valve”, or as I kept calling it, the vaping valve. It was bypassed earlier this photo series by a chunk of fuel hose. Unlike a modern fuel injection system where there’s a fuel pressure regulator bleeding off fuel into the return line, this is just a little pipe with a hole in it. Its nominal purpose is to prevent vapor lock by having fuel vapor escape through the little hole, along with small amount of fuel. Ford sells these in several hole sizes depending on how much you want your engine to vape and also to help modulate fuel pressure. What the hell? Almost every system on this vehicle in some way is analogous to touching a variable resistor to something. None of this is okay.

And you know what? It ran GREAT!  Rev for days! Piss of the neighbors!  A M E R I C A N  P R I D E. Oh crap, what’s that smell??

This is a giant puddle of BROWN  that was slowly increasing in magnitude as I was having too much fun firing gun-wielding Bald Eagles out of the exhaust. Which, by the way, was backfiring (afterfiring) like crazy. It was clear the thing was running super rich. I now had literally the opposite problem as before. Hurray!

A little research showed that all of this BROWN was coming out of the vapor canisters. So much fuel was getting into the carburetor that it was coming out of, and flushing, the vapor collector lines and washing out the canister. Great! It’s like clearing your sinuses!

(The parking lot is still BROWN and smells like a freshly coated gym floor there, to this day)

So what was the cause? I dug into the carburetor manual a little after observing it with the air cleaner removed. That’s where I discovered another “tighten the incorrect nut” problem.

I found a little nut on the carburetor that I thought was left loose by the mechanic; since, you know, I kind of bailed it out of their hands before they could put it back together. So I tightened it. All the way.

Well it turns out that’s the fuel level adjust screw and accompanying locking nut for the secondary throttle (which only opens after about 50% throttle travel or so). All the way down is all the fuel, all the time, forever and ever. I noticed the secondary throttle discharge was completely wet even while idling.

This screw was adjusted more correctly.

 

And off we go! I went on a “Lap of I-95” test. On the whole, performance was excellent. I made a point to exit and then re-enter the freeway repeatedly to do full throttle pulls from ~25 to 70mph. It still tended to be a little backfire-y when letting off from high throttle demands. Given that the mechanic didn’t have a chance to properly tune and road test, I think something’s still a little whacky, but that is now an addressable problem.

Hello, Mr. Tesla. Get in my belly.

In the end, I concluded that there had to be a very specific series of events and misjudgements for the running condition to get this bad. Here is my assessment of what happened:

  • I replace the front fuel pump. Given that I had no lift and was working entirely on the ground without help, mostly in the dark, I didn’t drop the fuel tank fully and inspect it. I only lowered it enough on the straps to grab the fuel pump. Therefore, I didn’t see how much rust was inside.
  • When the front fuel pump went back in, I neglected to install an inline filter on the output side.
  • Filling of the fuel tank with new fuel, plus the force of the collision, likely washed a lot of loose debris into circulation, where it was picked up by the pump.
  • I specifically used the front tank around town and during the Motorama trip attempt to ensure the system was operational.
  • The carburetor began clogging with this rust slurry, causing me to abandon the Motorama trip as problems gently surfaced.
  • Gradually, with additional around-the-town usage, it became worse as more rust slurry blocked the caburetor inlet screen.
  • The particulate debris MAY have also affected the operation of the fuel tank selector valve – I am uncertain if it played a substantial role in the earlier problems.
  • The mechanic was able to test for proper fuel volume delivery because the valve problem was a some times thing, which incorrectly ruled out the fuel system as a source of trouble. The Ford official player’s guide makes you only time the delivery of 1 pint of fuel as the test. I was pumping into a 5-gallon gas can and was purposefully holding the pumps on for a while in case something caused them to lose power.
  • After the carburetor rebuild, I told him to continue testing using the front tank, because the rear tank was in an unknown state to me and had a non-functional fuel gauge sensor, whereas I said I had replaced the front fuel pump and fuel gauge sensor.
  • Furthermore, for debugging systems in isolation, they bypassed the fuel tank selector valve completely and directly piped the front fuel tank to the carburetor.
  • The carburetor began clogging with rust slurry again, leading to much the same symptoms and to the frustration of all involved.
  • This timed well with me electing to cut my expenses and asking them to stop work.

So there you have it. Once again, one of my vans is stymied by the uncertain nature of fuel delivery. You know what? Electrons don’t need fuel filters! What are you gonna catch, some neutrinos?

How bad is the front tank? Probably very. Before I turned anything on again post inspection, I went and bought the biggest clear inline filter Pep Boys had to sell and dropped it on the front fuel tank’s output line. That’s what it looked like afterwards. Delicious, delicious BROWN . I’m not likely to do any fancy fuel tank treatments to stop this. New replacements can be had for $100 or less, and as long as the bed has to come off, replacement will be easy. Plus, unlike Mikuvan’s filter, these two things are easy to reach, so I am likely just going to keep changing big plastic pubbles first and foremost.

Next time, the action starts for real:

 

Reassembling a Bridgeport J-head with Uncle Charles! And More About Hooking Up Your Annoyingly Chinese VFD

You know what? I’m tired of having sweet-ass machinery sitting around not hooked up. Last time in “Charles takes forever to set up his own shop because he’s sick of setting up shops”, I did some battle with a generic Chinese VFD and completed what the damn factory couldn’t be buggered to by adding the dynamic braking components.

Though Bridget ( <3 ) ran since then, there were some issues. The spindle brake was so worn it was difficult to change tools, and the head made the “Bridgeport Clack” from the high/low speed dog clutch being worn. The motor’s V-belt was also severely worn. I wanted to tear it down for a rebuild of sorts, so I spent a little while watching “How to rebuild a Bridgeport head” videos. I decided that all of these videos sucked, and that I was really only interested in repairing the brake and replacing the timing belt and V-belts.

So here is my documented take on how to take apart a Bridgeport 1J head. In it, I discover that it wasn’t as terrifying as I had thought originally, and that old-school American engineers might commit some abominations but damn they’re good abominations. I guess this is kind of a Beyond Unboxing, too.

Step 1: Dismount the motor, which is retained by two studs, one with a set of two jam-nuts to let it move a little for belt tensioning, and another that’s the ball handle (you unscrew the ball handle and then untighten what it’s attached to). Then, crank the head about the Y axis (roll) 90 degrees.

Six socket head cap screws live underneath the belt cover casting and retain it to the steel back-gear housing. You can take all these off; pins retain the belt cover afterwards, and it needs to be yanked off. Don’t worry, it’s not heavy. But there’s one catch:

The back-gear timing belt pulleys both have flanges. To remove the belt cover means taking off one of the pulleys with it, and that means removing the belt with it. You have to remove the four slotted head screws that keep the pulley flange on. Once it’s gone, the belt slides off with everything, like this:

This setup is quite the abomination. The timing belt has no tensioner – it relies on good will and good spacing. Mine was getting a little loose from the years. While I haven’t run the machine hard in back-gear range to see if the belt skips, I ordered a new belt anyway since it’s a “Might as well” item. The belts, and other rebuild components which will be seen, came from H &W Machinery Repair.

While the cover was off, I cleaned off the thick layer of congealed rubber dust and spindle oil. I didn’t break into the back gear cavity, however – if you do, remove the nut on the big pulley and use a gear puller or Three-Phase Prybar to pop it off, then undo the remaining screws. Some times the gear cavity is filled with grunge; if your machine had multiple owners, chances are it has both grease and oil in it.

I loosened the cover and a lot of remnant oil started pouring out, so I’ll likely keep it together but drown it through the front oil port later.

The second step pulley and back gear timing pulley live with the belt cover and has a large bearing carrier assembly under it. To undo this, I need to remove the shifter mechanism.

The pins that ride in the shifter groove also help retain it completely. Problem: One of them was completely stripped and wobbly. Due to the pressure exerted by loading springs underneath the pulley, I couldn’t get the pin to bite on its remaining threads and back out. So I drilled straight down the center and threaded the hole for a #4-40 screw that I could then grab with pliers and pull on:

The stock machine has slotted head pins; H&W sells a replacement with a hex wrench drive. Here’s the victim screw driven in…

And a few tugs later, the shifter ring is freed.

The pulley then flies off the other side, since there are loading springs underneath it.

And here we have the brake assembly. The brake is simply a phenolic drum brake setup that crams against the interior of the pulley. Nothing sophisticated at all!

To remove the brake, you have to remove the 3 slotted-head shoulder screws holding it down. However, to do that, beforehand you have to undo the three hex nuts on the top side (underside in these photos) – they prevent the shoulder screws from loosening.  After that, the brake can be wiggled off gently. It will snap closed, due to its own return springs, so watch your fingertips .

The small tongue on the upper right of the bearing bore is the cam that toggles the brake shoes.

Many times, when a Bridgeport spindle brake is worn, it means two things – one, that the brake shoes are worn down, but what I found is that the cam had also dug a little trench into the brake shoes where it makes contact. So this has reduced the effective travel length and caused the brake shoe to lose engagement. In fact, it seems like the harder you wail on the brake lever, the quicker you induce this 2nd failure mode.

Also, Brigeport brake shoes are expensive. Speciality exotic part, sure, but I can do all 4 brakes on Mikuvan for less money using nice ceramic pads too! So I wasn’t going to replace these, but simply make the cam bigger.

Returning to the top side, the brake cam escapes if you untighten the set screw holding its handle pin in place. The pin slides out and the whole thing falls apart.  The cam and shaft assembly are on the upper right.

The fix? Make the cam bigger by welding repeatedly over it, building up more metal, then sanding and filing it down! This was after the rough-sanding stage. I filed a gentle round onto the engaging edges so it doesn’t cause further erosion of the phenolic laminate brake shoes.

Alright, we’re now on the reassembly path. The brake cam is going in back in…

Secured up top, along with installed brake shoes and re-tightened locking nuts.

I reassembled the shifter ring after cleaning the whole area and thoroughly greasing it. In Bridgeport maintenance, you’re supposed to oil the shifter ring daily in production use. I think I’m fine with putting in a few greasewads where it needs to be instead of having to clean up even more crusty oil grunge down the line.

The belt cover is remounted now.

Before final assembly, make sure to thread the timing belt and V-belt back onto the pulleys. Then as you line the belt cover on, wiggle the timing belt onto its large pulley.

When finished, you can then replace the small screws and pulley flange.

Putting this motor on was the precarious part, since it involved holding something pretty heavy and wiggling it from an awkward angle! I threaded the two jam nuts onto one side in order to hold it in place for….

Final head tilt. Here are the newly installed parts! And there we  have it. Shifts great, runs smoothly. Still makes The Bridgeport Clack, but further research showed me that is all in the quill spline drive and there is not really a way to R&R that short of replacement. I’m fine with it.

Moving onto controls! I can’t use this thing from a potentiometer dangling by its wires forever. You may, but I have standards.

I put a little money on eBay into some more machine style switches and buttons.

I had two buttons left over from a project long ago, so they were going to be used as the Run and Stop functions. The same potentiometers got transplanted into a panel mount which I screwed into the housings. Knobs were a matching pair (rare! legendary!) found at MITERS.  The two-position switch will control forward vs. reverse.

The wiring was concocted using disembodied Ethernet cord, which is one of my favorites for pirating cables from their intended purposes. The VFD’s Use of Manual™ just showed a bunch of normal looking switch symbols connected to the forward/reverse, start/stop/reset, etc. inputs.

This is where I discovered another great undocumented feature of Use Of Manuals. The diagram was a lie, but only enough to get you in trouble.

I had problems with it accepting my switch configuration. I found that the VFD didn’t want to read my stop button at all, and it accepted any flip of the direction switch as a “run” command. That is, I can toggle the forward-reverse switch for it to change directions, but it wouldn’t take my stop button input. I’d have to hit the STOP button on the control panel of the VFD. After that, I couldn’t start it by using the start button, but just changing the state of the direction switch would let me turn the knob and increase speed again. Well, all of my settings seemed to be correct for the job, so I was a little confused and figured there must be Undocumented Behavior. This was certainly inconvenient to use the damn thing intuitively, and I certainly wouldn’t let anyone else touch it in this condition.

It took a few friends with experience in industrial controls to point out what I was doing wrong.

 

That is a diagram for a normal industrial magnetic contactor, showing how Start and Stop buttons are typically wired. In these things, the STOP switch is always closed unless something causes it to open (either by accident or on purpose). The Start switch, on the other hand, briefly powers the contactor coil which pulls in not only the main contacts, but a little auxiliary contact that keeps the coil energized and hence the contactor latched. You can see how any number of interlocks (e-stop systems, overload detection, etc.) can work its way into the STOP circuit and turn the machine off when needed.

The VFD is technically designed to replace this setup, so it’s expecting the Stop button to be normally closed. Well, all my switches are N/O type (close when pressed). So the VFD was waking up in an unexpected mode, I guess, where it seems to default to treating any forward/reverse switch inputs as “Okay, start running”. Well this seems a little scary of a failure mode.

Anyways, the Use Of Manual shows all switches as N/O, so it definitely assumes you already know industrial control practices to use it. That’s another endearing characteristic of Chinesium… you better know exactly what you’re searching for, or else you might find it.

Well that’s quick fix. I didn’t order modular contacts with my switches, but luckily they’re manufactured modularly enough to use the same set of contacts, just internally turned upside-down, to become N/C. Now my control panel works as expected – the stop button puts the VFD into slow-down-and-brake, then start will ramp the motor back up to the previous speed it was at. In run mode, I can change speeds at will, including braking down to zero speed manually.

And here’s the test video.

Now that I understand this setup (or do I….), I can build the second control box accordingly. It’s also easy now to add an anti-face-eating emergency stop mushroom button anywhere in line!

The next machine to go online will be Bridget’s cute Japanese friend, Taki-chan!

how about no