New replacement coolant pressure tanks for the Vanagon cooling systems are known to be not nearly as resilient as the OEM ones, they tend to last only a few years. The definitive, fix-it-once answer to that is a welded aluminum box tank from RMW. Competently designed and executed, they are truly bomber. Plus, VW’s exclusive and sometimes troublesome screw-on blue pressure cap is replaced by an old reliable bayonet mount Stant 2-way cap, available everywhere. That also means the standard pressure tester every shop and AP store has can now test the cooling system and the cap; few if any have the proprietary threaded adapters for the VW parts. So, big improvements in several areas at once, and knowing what quality welded parts cost to make, it’s actually quite a fair deal at $250. And if that’s not enough nice things to say about it, the proof RMW had a good thing on their hands is that after pooh-pooing the whole concept on their blog in favor of buying new plastic tanks every 2 or 3 years, GoWesty recently reversed themselves and copied RMW, trying to undersell them at $200.
.
Having used up one by one my stash of OEM plastic tanks, I finally had to buy a new one, made by Meyle, and that one was at two years when a stone threw the water pump belt and I had a minor overheat. The p-cap did its job and vented off a bunch of coolant, but then coolant was boiling to steam at a rate faster than the small orifice in the cap can let escape, whistling madly. The pressurised steam did this:
.
It still holds pressure, but I’m not gonna count on it to go another round.
.
So, time for something better, finally, and RMW’s tank beckons. But I’m doing a start-up, spare cash is very tight, and besides, I have skills and tools, why not roll my own?
.
I don’t have a TIG welder, though, so can’t weld aluminum. So I ordered a steel radiator neck, thinking I would braze up a sheet steel box tank, but rolling the idea around I knew there would be a lot of challenges and didn’t want to spend a lot of time on this. So I came up with a compromise where I spend a little money, do the rest myself, and get on with other things.
.
So, shopped the huge number of tanks at Summit, found a welded aluminum box tank that looked like a good candidate for only $100.
.
Up top it has a 1/8″ FPT hole in the filler neck, which took a 1/8” MPT to 1/4″ hose barb for the overflow hose (which no longer has to be pulled off to remove the pressure cap, I love that!).
.
On the sidewall is a 1/4” FPT bung at top and 1/2” at bottom. The hoses that connect to the side of the stock plastic tank are 5/8” ID top, 1” ID bottom. No one makes a 1/2” MPT to 1” barb fitting, nor a 1/4” MPT to 5/8” barb.
.
No worries, though, when you have the tools. The lower fitting I spun up on the lathe from some 1.040” thickwall bearing bronze tube stock, that extra mm was perfect to leave behind as the bead. The upper I made from two standard brass fittings, a 5/8” barb to 1/2” MPT adapter, and a 1/4” MPT hex nipple. Cut off the threaded end of the barb adapter, and the ID was perfect for the 1/4″ MPT tap. Bang-zoom, the hoses will fit right up.
.
This goes in my Syncro, so the engine sits almost 2” lower in the chassis. The vertical spacing between the two side fittings is wider than the hose nipple spacing on the plastic tanks, but that was easy to solve for the Syncro by using the 2wd elbow-flex hose instead, which is shorter. If this was for a 2wd Vanagon, you’d use that same elbow-flex hose but install the tank higher up and find a shorter formed elbow for the bleeder ring connection.
.
I don’t run with that giant, oppressive OEM muffler heatshield, I always found it unnecessary and with it gone engine access is so much better. You do need to shield the lower crossover pipe’s coupling hose, and the plastic part of the oil filler neck, but those are the only bits that ever suffer from the radiant heat. With the shield and p-tank support gone, my p-tank was supported by two metal bars bolted to the body instead, which give two bolt holes in basically the same location relative to everything else as the stock p-tank stand. So, leaving those bars in place, a short length of 1 ½” alu angle, a few holes in the right places, and the new tank bolted up in the same spot.
.
The new GW copycat tank features a sight glass, which the RMW one lacks, and some folks think it’s the cat’s PJ’s. Me, I could never say for sure what the coolant level was in the milky plastic tank even when it was brand new, and the clarity doesn’t improve as it ages (I anticipate same for their sight glasses, sorry to say), so being able to see the liquid level was never something I relied on. Instead I’ve always found that with a good p-cap I could count on watching the reserve tank to tell if the system was losing coolant, and the OEM level sensing system was always pretty reliable as well. But now I also lack a coolant level sensor. So I have a few choices: run as is, most cars don’t have a coolant level sensor anyway and they manage just fine. Or buy RMW’s metal level sensor and add a bung for it to the top of this tank. Or get GW’s little circuit board that adds an audible warning when the coolant temp gauge goes high. That’s only $40, easy to install, and I don’t hate GW or anything, they do come up with some original stuff and this seems like a good little gadget to have. In my experience if coolant level is dropping fast from a leak or rupture, overheating is seconds later, so in that event the audible warning is more useful than the visual-only warning anyway (as long as I can hear it over the Vanagon’s churning symphony of awesomeness). I’ll probably go that route. The coolant level sensing system I disabled by just jumping the two pins in the sensor connector.
Good on you! It seems your business is not VW parts, but Creative Improvisation!
Well Steve, I’m sure you get that problem-solving and designing are where the fun and challenges are. I never tire of cooking up and productionising new stuff. Unfortunately, afterwards I have to make it and sell it to get paid for the work, and that’s the not-so-fun part, doing the same things over and over. I wish someone would just pay me to do the part I enjoy!
I’m going to zerox your coolant exp-tank idea. And yes most vehicles get by fine without any visual ck for coolant level unless you open the system, cold of course. Most people never even look at the engine for anything so it probably won’t matter. Also VW left the “low coolant” warning system out after the early 1990s. I knew guys that were working at the dealers at the time and because of all the warranty hassles with the constantly flashing “low coolant” LED and the “low oil pressure” buzzer notifying you of something, but not low oil pressure, they were ready to quit and go clean toilets or something.
BTW the “low coolant” warning is ineffective. I’ve personally had several component failures that let out most of the coolant. The “warning” LED never flashed on. Oh Well.
Also the cooling system pressure caps are another problematic area.
Found this out 30 years ago when motorcycle road racing. Our team had switched from the first liquid cooled Hondas (absolutely stone-axe reliable) to the, then latest, Yamaha 600 cc inline 4. Those bikes should have come with a large envelope with several new head gaskets and installation instructions.
Long story short. We fixed the problem by installing a “Nascar” cool can. Which is just a tubular expansion tank, which the stock bike had none. This was mounted where the headlights had been since we only raced in daylight. With the outlets from the cylinder head plumbed into the top of the tank and the outlet plumbed to the radiator there were no more head gasket failures.
Two things were going on here. One was the stock pressure cap was the ‘mini’ typ and higher pressure ones were nonexistent at the time. One of our team decided to test the cap installed on the fitting on the bike. The cap did not go on the radiator, but on a remote fitting with several hoses in and out. The ubiquitous Stant tester had the cap opening at 15 psi. Exactly what it was rated at, but when installed as it would be on the bike, block off all hose fittings except one and hookup the pressure tester to the remaining fitting, it opened at about 9-10 psi. When the ‘cool can’ was installed we ended up with a Stant “racing” 24-27 psi cap. When that was tested on the can instead of the Stant tester, it opened at 19 psi.
Some experience for thought.
Also with the ‘cool can’ the bubbles/steam coming out of the engine got a chance to swirl out of the coolant, we ran just water per racing rules, so only liquid went through the radiator. This lowered engine temps considerably. With the higher pressure system the boiling point was raised enough to stop significant problems. BTW in the process of this drawn out ordeal we tested all the ‘racing’ water additives that were legal to run. Most of them were ineffective at improving cooling and the rest caused other problems like water pump leaks. Waste of money and time.
Hi Victor, thanks for your comments on this and my other posts, it’s great to get observations from someone in the trade.