Operation IDIocracy: Left Snail

With the IDI oil cooler repair completed last episode, now I could fire Snekvan up and warm it (very loudly) up without internally trading fluids. And that means the fun part begins! While watching glue dry and hiding from the intermittent cold after-work nights, I was shopping around online trying to orient myself in the world of terrible Chinese turbochargers: The Chinese Choo-Choo Conundrum.

I like to think my history of state-estimating the commoditized Chinesium product cloud made this search pretty straightforward. I had upwards of 15 or 20 tabs open across Amazon, eBay, and Aliexpress trying to sus out what the “main bloodline” part was. In other words, if you put absolutely zero effort into buying a generic Hong Kong Hair Dryer and basically hit the first one that was presented, what would it be?

Based on my clicking around, it’ll probably be a:

  • Garrett T3 / T04E hybrid (using a smaller frame turbine housing with a larger frame compressor housing)
  • 0.50 A/R compressor housing and 44 trim compressor wheel
  • 0.57 or 0.63 A/R turbine housing and 73 trim turbine wheel
  • 5-bolt turbine outlet flange

What the fuck do those numbers mean? Hell if I know. Actually, about 1.5 months ago, I didn’t know at all. Apparently if you blow into them really fast, they make whooshing sounds, and if you point the whooshing sound into your van it goes faster. And that’s the goal, right!?

I had to read a bunch of “how tubblecharger” guides, like the ones (101, 102, 103) put out by Garrett, to make ass and teakettle of those specifications. Once I established what the ‘dimensions’ of adjustment are, things made more sense. The takeaway was that a proper combination of The Numbers is needed so the turbo is able to supply the needed air flow at the needed pressure at where you want to operate the engine, whether it’s high RPMs and loads (racing and being a hoon) or down low (towing and street driving, which I have a feeling is where I’ll be operating).

Now, the majority of these Beijing Boosty Bois seem to be sold as part of kits for small-block V8s… but the same ones are also found advertised for inline-4s! So what gives, besides the Hangzhou Hoon Honker sellers just writing whatever gets them more hits on websites?

I mean, ultimately it’s up to the end user to not blow everything up… but then again, half the reason I’m doing this is to maybe blow something up. Ultimately, I’m kind of backdriving this whole process: From a set of given Tsingtao Tranny Twisters, see what kind of boost and performance I can get out of this clapped out school bus engine. Most “how to” guides in this realm seem to help you size/select a Wuhan War Whistle based on a certain engine and a performance target. My performance target is “lol”

But first, I’ll go through the motions to make sure it won’t explode outright. Luckily, Garrett has wrapped up their how-tos into their “Boost Advisor” webapp that lets you punch in all of the optimistic bad assumptions and it should, in theory, spit you out some part numbers you can look up and buy.

I say theoretically, because it seemed like no matter what I enter that approximates the operating conditions of the 7.3 IDI that I think it will be under….

I tried using the twin turbo config as well as the single config, putting in unrealistically high numbers for horsepower and RPMs, and so on. The IDIs really only rev to ~3300-3500RPM without extensive rework inside that I am not intending on doing. I wasn’t out to run (hopefully…) insane boost (pressure ratio) levels as well. I also didn’t really know what it was expecting for a “Mid range” RPM – while the guides say the RPM of peak torque, for the IDI this is allegedly at 1500 RPM and that resulted in…. negative pressure ratios.

Lower than atmospheric pressure. Pulling a vacuum. So I had to up it to something like 2500-2600 RPM, which would be the approximate RPM it will sit at while on the highway based on prior experience with the ven.

Maybe it’s just that this app doesn’t expect anyone to try and turbo a potato.

Nevertheless, I wanted to just do a first-order sanity check. So I decided to go digging for some boost maps. The compressor side is the T4 frame, so I looked one up that had approximately the same compressor trim. Using 1/2 of the numbers provided by Garrett, since I’m using two (and was told it’s okay to just combine them, remember…), I got….

Well, the points are at least on the map. I (having no prior intuition or knowledge in this realm) was certainly delighted.

Now, my understanding is that i’m well under-utilizing the potential of this size of Shanghai Singing Snail. That my “Max power RPM” operating point is almost center, if not towards the lower left, in the map means they technically has a lot more to give. But, giving thought to their provenance… perhaps a light load is beneficial to survival.

Keep in mind this is only “half” of the engine, so to speak, and one of these T3/T4 units would seemingly not be able to serve the flow required by itself.

No matter what, this setup almost seemed too sane. Maybe it will suffer from obscenely long spool-up (lag). Who knows!? I don’t even know what turbo lag feels like, and I’m kind of hard pressed to believe that I’ll be able to tell a difference with an engine like the IDI.

I decided to play around a little with other available compressor maps and see if there’s maybe a future path if I decide I know what I’m doing and have a specific setup in mind. For instance, I got curious about getting a “purebred” T3 sized Ying Yang Spinny Thang, which have smaller compressor housings and may be able to 1. fit more places in the chassis, and 2. be better utilized.

I found this map for a 60 trim wheel in a T3 compressor housing and plotted my same airflow and pressure ratios in. This seems quite reasonable to me, with my “max power” estimate being closer to the choke line but the “Mid-range” point being squarely in an efficiency peak zone.

However, it seems that the non-hybrid T3 size Wu-Tang Whirlygigs are a little harder to find and more expensive. And the whole goal, as I introduced this whole project, was to find the worst solution possible. The turbo in front of me is the one I need.

I ended up picking and choosing the “Least Common Denominator” spec as I outlined above, two of them for the absolutely eye-melting cost of…. $260 with shipping from somewhere in Kentucky (presumably a domestic fulfillment warehouse as is very common now for eBay stores based in China). Here’s the “d a t a s h e e t”:

Yeah, alright. We get it, you can enter numbers in an input field on a screen.

Turbone…. are we for real here, guys?

Fast forward a few days, and….well, what do you know. It IS the item that was shown in the picture above!!

As much as I might be making fun of these Pearl River Pinion Poppers, the fact that this object exists, in my hand sent from halfway around the world, is superficially of the correct shape and material, and for ~$125USD apiece, is a testament to the arrogance of Man.

Excuse the college term paper tier run-on sentence there, but I say this often about the Chinesium goods populating the hobby and maker sphere. Chances are everything about them has been optimally value engineered, and they Won’t Not Work, but will do so only under certain favorable circumstances. And it’s up to YOU to find that out! This won’t last forever. Enjoy it while it’s here, and while we still have oil left in the ground.

When I received these things some time in early October, I was still in the “Yup, that’s there a turbo” stage of knowledge gathering.

The left snail shell is the compressor housing, with intake on the left end and the output looking at you.

The bearings and seals are in the center, known as the “cartridge” section. The exhaust turbine housing is on the right, made of cast something magnetic and ferrous in nature for heat resistance.

Spanning the two is a pressure-actuated pushrod canister for releasing the wastegate at a set pressure (you can see it connected to the output of the compressor with a black hose on the left).

The two halves can be “clocked” any which way by releasing the center hex-head screws that lock them in place with clamping rings. This will be important later on because I’ll need to set this clocking before making new mounts for the wastegate actuators, which won’t line up any more if I need the outlet to point some other direction. All this seems to need is the oil drain hole (the rectangular port) being approximately vertical such that the post-bearing oil flow has somewhere to go immediately.

Interesting aside here – usually you’re supposed to mount these high up in the engine bay so the aforementioned oil return can simply flow back down to the crankcase/oil pan. However, because I’m doing what is called a “low mount” or “remote mount”, I’ll eventually need this return path to go to a lowest-point-in-the-system sump of sorts, from which it has to be actively pumped back upwards to the oil pan.

I broke down the compressor side to take a look at the compressor wheel itself. This is a cast wheel that’s post-machined only on the outside to match the profile of the housing. For more money you can get “billet” wheels that are presumably what all the 5-axis, mill-turn, and combination machine manufacturers make when they demo the things at trade shows.

On the other side, the “5 bolt” internal wastegate lid unbolts to reveal the wastegate valve itself. Now, I think there’s supposed to be a MLS or stainless steel gasket on this mating face, but this didn’t come with one, just the two cast iron sections screwed together. Nothing that isn’t also sold online everywhere, but I figured it was a point-of-manufacture cost-cutting measure.

The wastegate valve is just seemingly a stamped washer loosely riveted to an arm. It depends on the wastegte actuator (the silver cans) holding them shut with their internal springs, until boost pressure overcomes those springs.

I mean, if it gets the job done? It doesn’t have to be a vacuum-perfect seal, just so long as it leaks Substantially Less than the total exhaust flow.

The two housings come off pretty easily. On the exhaust side, there’s a big loose-fitting steel hat (silver thing between the cartridge and the turbine) that I presume is a heat shield.

I got extra curious and decided to see if I could dismantle the cartridge to get to the bearings.

These things are either cranked on very tight, or have threadlocking adhesive in the middle, because I couldn’t get the nut loose without almost stripping the cast hex head of the compressor wheel. It doesn’t help that one face of the thing is machined down for balancing purposes.

So, not wanting to completely destroy these things before I even use them, I decided to stop here. I mean, I get it. Spinny thing go wheeeee and it hopefully looks the same as a diagram on the Internet inside.

The worse part is…. there’s two of them.

Alright, let’s get to work. With the exhaust having been removed for the oil cooler surgery, I went ahead and did some placement ideation on the left side.

Now, this side is actually the easiest. There is PLENTY of space here, even if I were to do a single larger turbo with a crossover pipe underneath the transmission like the OEM exhaust.

The reason is that the engine and transmission are actually shifted a few inches to the right in the van chassis. They’re not dead in the center, and the engine cradle/crossmember is asymmetric to reflect this. This was done to give the driver some semblance of legroom, but it results in the passenger having to be basically sitting sideways.

Hell, I’ll go as far as to say that this is the most prime piece of undeveloped real estate in the entire Ford VN chassis.

To hang the turbo on this side, I had to start out with a mating element to the exhaust manifold flange, which was vaguely ball shaped. Turns out this is indeed called a ball-and-socket flange, and is to give some compliance in assembly for the engine and exhaust pipes, which would be mounted on different squishy things (engine mounts vs. those rubber loopy things)

I found a matching part number with the correct ball diameter (the Walker 41725 seen there on the barcode) and made a few guesstimates on the length needed to put the turbo at a height that I could escape the exhaust straight backwards, and that wouldn’t mess with the brake proportioning valve mounted nearby on the frame – that’s the square-bent tubes on the upper left.

There was an immediate huh moment when I received the T3 adapter flanges. Round peg, rectangle hole.

What gives? Well, after pestering Car People friends, iseems like I was either supposed to spend bigger money on a “Transition Flange” that someone CNC machined, or…. deal with it.

Well, the faster path to donuts and violent ejection of head studs is DEAL WITH IT. I decided to just cut a hole in a piece of steel and weld it on. I had a piece of 1/8″ thick cold-roll barstock that was just wide enough to cover the rectangle, so I had to put a 2.25′ hole in it.

Now this operation was “sketchy” to use a very mild term. I think there was less than 1/16″ of wall remaining on the sides of the steel piece where the hole saw, which isn’t a perfectly true cutter, was swinging through. But it held!

The ball flange adapter gets cut down to the height I think is correct. I went 0.5″ over, actually, just in case I needed to trim more off later.

So this is what it actually looks like. Not bad, I suppose.

In the “Final Version”™ I think my plan is going to just be having correct-hole flanges laser cut from some thinner steel (seriously, this thing is 1/2” thick… You’re supposed to weld that to skinny exhaust pipe?)

I decided to be weird here and did all my welding on the inside of the pipe and flange adapter. The 1/8″ “transition plate” was already rather close to those bolt holes, so keeping the weld on the inside helps with being able to actually use nuts or screw heads later on.

I then got extra weird and pointed the MIG gun down the center of the pipe stub and filled this last area in from the inside of the inside. Look, all it has to be is somewhat hot-air-tight.

Here’s the prototype left side … downpipe? Exhaust runner? ready to get some test-fit bolts.

Here’s the first test fit! I absolutely loved how well this turned out. If I clock the compressor outlet upwards, it basically points straight through the gap in the floor to run the charge air hose. The intake would need to do a quick 90 degree turn to avoid the engine mount crossmember, but it can go anywhere from there. And, the exhaust exits basically straight back.

I mean, I couldn’t have hoped for a better fitu….

Oh, I forgot the oil filter.

Alright, everything’s ruined. This still works, though the exhaust will need to make a funny U-turn between the outlet flange and the engine mount crossmember. There’s plenty of space to do it and then to run the pipe down the outside of the frame rail.

UGH. I thought I had it all figured it out. Just like my life.

The next episode will focus on the other side, which has a lot more interesting piping involved!

Operation IDIocracy Season Opener: The IDI 6.9/7.3 Oil Cooler Repair

Alright! So before I can go about my snail farming, I had to address a pressing issue with Snekvan: That the oil cooler assembly was causing engine oil to enter the coolant circuit any time it was running.

This manifests as blobs of engine oil floating on top of the coolant. The amount of grunge built up indicated it has been leaking for a long time, and was just neglected.

I drained the terrible soup into a large under-bed storage bin, upon which I had to invent a method of disposing of it that didn’t involve the storm drain. Then I refilled the system with clean water, idled it, and watched the open radiator cap as more oil blobs began surfacing. No matter what, I wasn’t out to gauge the severity or mediate it, since once it happens, it’ll keep happening. It was time to remove the oil cooler and do some surgery!

But first, I spent a good hour just cleaning and scrubbing the front row of this thing because it felt disgusting just getting in. I blew out or vacuumed up a lot of biological detritus, including dozens (or hundreds) of stinkbug husks from under the dashboard, and bleach scrubbed all the surfaces to get rid of the mold and… stains.

In this process, it spawned yet another snake skin. Certainly living up to its name!

As I have harped on so many times as of late, such as the Crabmower post, I’m happy that other people on the internet also document their arcane machine-enhancing adventures in static and searchable ways still, despite the ease of social media ecosystems. In a way, I feel like this is the karmic energy balance shifting back towards me after documenting all of my asshattery here going back years, and fielding a ton of questions from other makers. Now, I’m following the directions of van wizards, and I’ll be happy to act as a second-source as well; the “Well I did it this way and it STILL worked… somehow” followup act. Almost this entire “build” will freebase knowledge from Nick Pisca’s IDIOnline website, beginning with this oil cooler repair.

So let’s begin. The oil cooler here is actually the oil filter head as well, so the filter has to come off. The exhaust downpipe is also in the way. Since I was going to have to re-engineer the exhaust system anyway, I decided to take this opportunity to just remove it wholesale.

Luckily, Snekvan didn’t serve time in the Salt Belt (it has practically no rust of meaning) and so the manifold flange nuts released without resorting to drastic measures.

I did end up cutting off the pipe right at the muffler joint, so there wasn’t a way to get it out otherwise. The entire OEM exhaust is installed in the chassis as one peice.

Surprisingly, all of the bolts holding the oil cooler headers on are reachable with only a regular 3/8″ drive short extension. The front bolts are 5/16″-18 threaded with 3/8″ 12-point heads, and the rear bolts are a 5/16″-18 hex head with 1/2″ hex drive.

As soon as the gasket are cracked, remnant oil and coolant will drain everywhere, so I made sure to apply the forces with the good ol’ Drain Bedpan underneath. There is no need to drain all the oil, as a good amount will remain in the cooler tube, but the coolant must be drained to below the level of the assembly (or it’ll keep coming out anyway)

I did discover that the oil cooler assembly was on the verge of being able to escape, but barely could not due to the presence of one of these exhaust studs. I mated a crow’s foot wrench to a breaker bar and extension (to clear the adjacent engine cradle crossmember from the bottom) and broke it loose, and then a deep 5/8″ socket to drive the stud out.

The whole assembly comes out “upside down”, so to speak. It has to be rolled about 180 degrees to clear the front header.

Better make sure you have a bucket to land it in, since the remnant coolant and oil in the tube assembly is now going to pour all over you. No matter what, this is an oily operation.

The weakness of these is the internal O-rings that harden over time. The problem is, there’s no good way to pull these apart. The prybar tabs that are provided are very weak and soft, probably meant for use with a cooler assembly that isn’t 30 years old and “cured” in place.

When the header sections are exposed like this, I used some wooden paint stirring paddles as the pressure applicator to avoid scraping up the O-ring land surface. Once it gets here, it’s easy to pull off the rest of the way.

This other side, though. The tab broke off almost instantly, so I just resourced to chucking the header in the vise and giving it a good ol’ twist-and-pull, wiggling the whole thing out little by little.

Check out these delicious aged-to-perfection O-rings. The small one seals coolant from oil, the large one seals oil from The Environment™. These both tore and cracked in some way as I removed them.

Overview of the entire removed and disintegrated oil cooler unit. The only thing that really needs “fixing” is these O-rings! What a series of unfortunate events to get here.

I decided to go ahead and deep-clean the assembly as I’ll need to do it eventually anyhow, so why not – iIt’s Already Taken Apart. There are two large fittings on the rear (oil filter head) header which have NPT threaded plugs. I suppose if later on I need to take high-pressure out of somewhere, for something, I can do so here.

One comes off with a 3/8″ square drive, the other is just a hex head.

Alright, this is absolutely terrible. I began by using a wire brush and terrifying chemicals to scrape and brush the grunge off. But it was absolutely baked on, there was a lot of it, and I’m positive the Chemical Romance was impacting my future genetic makeup.

So I gave in and decided to acquire( ) a wash basin from Facebook Marketplace. First one I saw nearby, offer 60% stated value in cash, and off we went.

I put in a gallon of Purple Colored Degreaser and a few gallons of water to make this delicious raspberry flavored van tea.

Within an hour, the raspberry van tea had turned into cold van chocolate. AND IT ONLY GOT WORSE FROM THERE.

From end to end, I let the parts marinate for 2 days, periodically turning them around or pointing the nozzle at different trouble spots.

So I actually put these in my dishwasher for the final rinse…. with my dishes

Huh…. so that is the color you’re supposed to be. I suppose I could paint these now, but cleanliness at the moment was enough.

I also had the small pipe fittings and the tubular assembly itself in the same bath – I positioned the tubes right in front of the pump intake, so it would be forced to pull Van Chocolate through the cavity.

So here is where I ran into some shenanigans. Rock Auto lists like 7 different part numbers for the oil cooler gaskets:

Some of these are wrong. I have no idea what the upper set of gaskets I received even are for, but guess who had to run back and order another kit once he found out. I ended up getting the cheaper Victor Reinz parts since chances are this will have to come off at some point soon anyway.

Big O-ring go first onto the large flanged seat (the “inner” flange, as I call it), then the little O-ring on the “outer” flange. I made sure to gently grease all the O-rings to avoid rubbing/tearing when installing them; any petroleum-based grease will eventually get dissolved into the oil once things are in place.

The next challenge was reassembling the oil cooler, which involves pressing the whole thing back together once the O-rings are slid on.

Now, Nick used a tall shop press to do this, but I didn’t have one, just a smaller 3-ton arbor press with not nearly the height needed. Well, then I remembered: The Trap House came with a random log splitter in the basement.

Whatever, it’s a hydraulic jack on a stick.

I lined up an aluminum bar chunk on the “splitter” end so I don’t “split” anything accidentally, and very carefully advanced the jack. What was handy was having a coarse fast-travel pump on this jack, so I was limited in the amount of force I could actually apply. The point is not to actually PRESS it on, but just slide the O-rings onto the lands. There should be practically no actual force applied, and I would periodically nudge the headers to stay in alignment with the tube.

Once it’s popped in, the headers can rotate on the tube, so they can be leveled off with a bench top or long straight surface.

Cleanliness is key for the gaskets, so I used brake cleaner and a 3D printer scraper to fully clean off the mating surfaces on the engine block.

I don’t have photos of this process in entirety, but I followed Nick’s advice of pasting the gaskets onto the mating surface with silicone sealant (applied in a thin noodle around all the bolt holes and on top of all the passage outlines), squeezing it down just a little, but then letting it cure slightly – still flexible, but keeping everything in place.

This ensures that I didn’t have to jiggle 2 gaskets and a (not very light) oil cooler assembly….

…while still having to rotate it 180 degrees back into position.

I fed through one bolt on each side while holding it up, and let it hang there while I inserted the other ones.

The procedure here was to gently snug every bolt a little (watching the silicone mush out some), then walk away and come back the next morning to do the final torque.

And then I decided to REALLY amp up the Full Send by pouring another gallon of Purple Colored Degreaser into the radiator and chasing it all with water as the engine warmed up. This will now become its “coolant” for the remainder of the build. At some point the future, I’ll drain it and repeat the process again, and hopefully the cooling system will be de-grunged enough.

Now that this repair is complete, and I could boot up the thing any time without potentially causing more damage, the fun part can begin: Sizing the turbos, and doing placement testing.