Top Gears

I'm getting all pumped up for my vacation in a week when all of the parts that have been specified have been delivered. I'll have mostly clear decks to go into the garage and work as hard and as fast as I can on this project that has gone through several retrenchments. Someone pointed out a nasty side effect of switch to Subaru power eight months ago and I've been looking at it since. It's called gearing.

While any number of adapter plates can put all sorts of engines into your VW, that doesn't mean that your VW transmission is going to appreciate 2-3 times the horsepower than it was designed for. Not only that, but you might not much like the experience either.

The VW transmissions were built with a basic idea about their power plant that goes all the way back to the Nazi era KdF-wagen: Top speed is cruising speed. In the KdF-wagen 100kph was as fast as the vehicle was expected to go, so everything about the engine's design was made to meet that goal and not a bit more. That means that the RPMs required are quite high by comparison, not for acceleration, or passing, but just for cruise: maintaining a mostly constant speed. So the engine is designed to produce its peak power and torque at the RPM needed to maintain cruise. At any RPM faster than max power and torque, both fall off rapidly: You can turn the engine faster, but not more productively.

So let's use an example to wrap our heads around this: The 1977 VW Bus has a designed top speed of 75MPH. With the right size tires on (185R14C which are a type of truck tire) the tire stands 25.7 inches tall. Every rotation carries you a certain number of feet forward. The taller the tire, the further forward you travel per rotation. So, pass the diameter of the tire, the engine RPM and the various gears in play in the transmission, and you can get a nice chart of what speed you'd be going in 4th gear for any given engine RPM. Confused? Relax. Here's what we've got, given a completely stock engine and transmission setup from the factory. (A 1 to .82 gear ratio in 4th, e.g. one rotation of the tire for 0.82 rotations of the engine.)

Speed (MPH) per RPM
2000	2500	3000	3200	3400	3600	3800	4000	4200	5000	5400	6000	6500	7000	7500	8000
38	47	56	60	64	68	71	75	79	94	101	113	122	131	141	150

So the top speed of 75MPH is achieved by turning the engine at 4000RPM in 4th gear. Coincidentally, maximum torque for the stock 2 litre Type4 engine is achieved at 3000RPM, and maximum power at 4200RPM. So 75MPH is a good spot for that engine: it can work at that RPM effectively and efficiently for a long time. It's designed that way. The maximum torque starts falling off after you've reached freeway speed at 3200RPM (60MPH) and then it's all about horsepower to keep things turning that fast or a little better.

As vehicles got heavier over the years, different gearing was installed in transmissions to match it to the peak efficiency of the engine, trying to stay in the sweet spot. The engines and their output kept changing too, as manufacturers struggled to comply with emissions regulations, usually resulting in engines that were less powerful: A constant dance between the weight of the vehicle, the power of the engine, and the transmission in between trying to mediate the constantly flexing relationships between the two.

So why the lesson in ratios? Because the EJ22 SOHC Subaru has a very different power and torque profile. It produces so much more power and torque than the poor old Type4 many people just assumed that it's a win to just swap the engine. What they don't count on is that the whole rest of the drivetrain, from the clutch all the way out to the tires, is expecting 4000RPM for cruising speed. This means that the EJ22 is capable of wailing away under the rear deck at 4000RPM and still have more power available. But it's LOUD. The RPM required to make the engine happy (where the EJ22 makes the best balance of torque and power) is much lower.

My EJ22 is from a 1997 Impreza Outback Sport which has an unladen weight of 2915 pounds, 127 pounds lighter than the 3042 pounds of a 1977 Deluxe Bus. So what RPM would the EJ22 cruise at if it was at home, coupled to its favorite Subaru 5MT transmission in a similarly weighted vehicle?

Speed (MPH) at RPM
2000	2500	2850	3000	3200	3500	3800	4000	4200	5000	5400	6000	6500	7000	7500	8000
43	54	62	65	69	76	82	86	91	108	117	130	140	151	162	173

Aiee! 4000RPM, while well below redline for the engine, is still flailing away making a lot of useless sound and fury for no good reason. To produce good cruise behavior, this engine only needs to be doing 3450RPM. Something has to give: either the engine needs to be completely redesigned or at least get new valve cams at a minimum, or the transmission in the Bus needs to be a closer match for the engine. Messing with the engine's behavior is expensive and a little ridiculous. Putting the Impreza 5MT transmission in the Bus is do-able, but very expensive. Having the 091 transmission in the Bus rebuilt with different gearing is not cheap, but it's less expensive than either of the other options. But there's a third option that has real merit...if you're willing to embrace some 'lifestyle changes.'

Remember I said that the tire diameter is part of the equation, moving forward a certain distance for each rotation? If you increase the diameter of the tire, you go further for the same number of rotations. That means, working backward from the tire, through the drivetrain and gearing to the engine...the engine turns more slowly. So it's only a matter of playing with the tire diameter until you find the (pun intended) golden ratio. Make the tire too big and it will have rubbing, grabbing and interference issues with the suspension and the body. So there's an upper limit. Thankfully, there's a tire that will fit, and produce an improved change of total ratio: instead of the 185R14C at 25.7 inches tall, we replace it with an Offroad/Onroad multi-purpose 27X8.5R14LT. (Yes, the sizing format changes to an older truck tire standard. It's roughly the same as a 215/70R14, a nick bigger at 26.8 inches tall.) 

So what do we get for running a big, scary off-road/on-road truck tire?

Speed (MPH) per RPM
2000	2500	3000	3200	3300	3600	3800	4000	4200	5000	5400	6000	6500	7000	7500	8000
39	49	59	63	65	71	75	79	82	98	106	118	128	137	147	157

Now we're getting somewhere! It isn't perfect, and 3800RPM is still much higher than it needs to be relative to the 3450RPM the Subaru 5MT gearbox would need for 75MPH. If you want to do better, you're going to need to regear: either a different transmission, or the same transmission worked over with different ratios of gears.

Fortunately for me, I don't have any desire to go 75MPH in my Bus! It is a 40 year old vehicle now, and has a reputation for far too much 'float' in the front end at high speed. I'm not in such a hurry to meet my Maker that I want to spend most of my time at that speed. So a John-Law approved 65MPH will put me at 3300RPM with gobs of headroom in torque and power if I need it. That's about the best I'm going to get without the $1700 for a rebuilt and custom re-geared 091 Bus transmission, or a rebuilt 5MT specially modified to join a VW Bus: $4150.)

I have to buy new tires anyway: the last time the tires on my Bus were healthy was when Clinton inagurated. They hold air...barely. So it's a bargain at $500 for four tires (General Tire - Grabber AT2 27X8.5R14LT / Load Range C) that will provide me many years of roadworthy travel and may lead me into adventuring in places that street tires might not take me. Everybody wins.

Decisions, decisions...

I've joined a few new forums recently to fan the flames for VolksarU. I don't want to do this solo; that's boring! But I am finding the number of people who have both experience *and* opinions to be thin on the ground. Opinions....you know what they're like. When it comes to the 'Conversion Perversion' crowd, the proportions change for the worse: 5 people in a room, 9 opinions.

I turned to the Vanagon crowd who have been doing conversions to their already liquid cooled vehicles for years. They've been very helpful, though all struggle with imagining the challenges of having to design the whole coolant system themselves. They're already plumbed for it. Burp the baby, and hit the road! A Bay window conversion? Not so much.

I've found a quite appropriate parallel to posting my solutions online only to have it shot at by those who've never worked on a Bay conversion in their lives: On the Discovery Channel, the program Mythbusters has a maddening fanbase. On the one hand, they're faithful to the idea of Confirmed, Plausible, or Busted. Where they differ is in the excruciating details of making a pop-science show that has to honor the scientific method on a modest budget while producing 42 minute episodes that entertain and educate. The pseudo-scientific second guessing (they're not taking drag into account! They've called it complete wrong!) got to be so bad that Discovery finally started producing an inexpensive 10 minute video blog to defuse the yelling. "Want to know why we did what we did, and didn't do what we didn't do? Log on..."

I'm inevitably in the same pickle. I can at least consider the opinion about my wiring from someone who has converted a Vanagon. But they're not qualified to comment on my radiator or coolant loop. You have to have tried a Bus conversion before I'll consider listening to comments about cooling. If you've actually succeeded, you have my attention. For everyone joining this blog in progress and who doesn't want to roll backward through two years of torturous decision making, here are the edited highlights.

But first, a word about international 'short-run' products: For the enthusiast in the USA ordering from the UK, the exchange rate and shipping usually winds up increasing the price of the component by at least a factor of three...or more. Add on your Federal, State and local taxes (all of which you are responsible for) and an 8 lb component from England can cost more in shipping than in the purchase price alone. Unless you are already shipping mounds of goods from abroad, you will get murdered on the shipping.

Therefore, most of my kit is from Rocky Mountain Westy in Fort Collins, Colorado, USA. They had two things going versus the UK conversion components: Their solutions required very little injury to the body of a vehicle being converted and they produce very high quality product at a reasonable price. They are a production shop, not a custom shop. If you ever want to standardize (one of VolksarU's chief goals) you have to have standardized interchangeable parts. I compared the quality of the following products that come from them with their competitors. They're also the only game in town in the Americas for production engine carriers.

So gaze adoringly at the T2B Bus EJ engine carrier, because its the only one on the American market. It's made of high quality materials, is mandrel bent for improved clearances and powder-coated for longevity. Sure, you can have your fabricator over and he'll weld up something that you can drill holes in the frame and bolt on. That's the solution everyone else offers for the Split and Bay bus. This makes every conversion an outhouse: put together with parts-on-hand to fulfill the bare minimum of utility. Parts from one aren't compatible with another. I've written elsewhere that this is the difference between an outhouse and an indoor bathroom. You're willing to read the newspaper in a standard bathroom, but an outhouse is 'minimal utility to do the job.' Not a place you want to linger.

So if you are doing a Late Bay Window as I am, the RMW hanger is the only design that locks the carrier into place in three degrees of rotation, and does so without requiring any cutting or welding on the body, re-using stock mounting points. In it's own way, it is the perfect demonstration of the VolksarU ideal: Installation is DIY friendly, no body mods needed and is compatible with two decades worth of Subaru EJ series engines. A lot of smart engineering, jig and fab work went into this carrier.

To my knowledge, only two other companies which offer carrier bars "cash and carry" rather than a custom fabrication each time. Both are both in the UK. (RJES & Fellows Speed Shop.) While good products and being weld-in compatible with many different models, this requires someone to do the welding, so they aren't as DIY friendly (unless you already weld.) Also, watch out for the international shipping killer costs if you aren't local to them.

My transmission adapter is made by Outfront Motorsports for Rocky Mountain Westy. Again, this is a CNC fabricated part, not a 'bespoke' one-off. Sure, Kennedy Engineering has been machining these adapters for years. Yet in my experience, they're also trapped in the technology dark ages, have a 20 year old web site, don't respond to email inquiries and often not to phone calls. They won't commit on delivery times or shipping. By contrast, there are several competitors who concentrate just on the Subaru to VW market and have taken the fit and finish to the next level. Why would I want to pay more for Kennedy's product when their quality and service has become demonstrably less? (This is my experience contacting them. Your Mileage May Vary.) It left a bad enough taste in my mouth that I decided to buy from someone else. Back to my friends at Rocky Mountain again.

I also needed a throttle valve reverser (TVR). While I can have this fabricated locally, why would I want to do that if I'm looking for standardization? RMW ships me one off of their shelf. When they run out, they fabricate new ones on their jig. Rinse and repeat. That jig can produce thousands of TVRs before it will need to be refreshed, and all of them will be as alike as pennies pouring out of a mint.

One area where some folks try to save some money is doing their own wiring harness. I took one look at that and while it is in my wheelhouse (My father taught me to do high conductivity electrical solder joints at the age of 7) I concluded that I wanted someone who had done a number of these and knew some of the pitfalls of the Subaru harness instead of finding them all myself...and ruining the harness in the process. So I put my ear to the ground and the name that kept coming up was Jeff Robenolt. Not just because he had done many Vanagons, but because he was willing to work with something unorthodox like a Bay bus and work WITH me. Jeff's contribution to the harness was invaluable.

So now I've got a backlog of product that I'm going to need: A Vehicle Speed Sensor (VSS), an OBD2 error suppressor, not to mention coolant tubing for the engine compartment and a complete exhaust system. When I went shopping for the VSS, I looked at the product quality offered by the vendors and shrank back with horror at some of them.

This vendor can look me straight in the eye and tell me that
I'm supposed to pay money for this ghettoo-rig engineering?

"To adjust the VSS distance to the trigger wheel, bend the bracket under the trigger wheel until you get a good signal."

Great slithering crow! I don't want that bracket to bend! I don't want it to even move! In any solution bolted to a vintage VW transmission, the VSS detects vehicle speed by watching a trigger wheel bolted to the Constant Velocity assembly on the outside of the transmission. On a Subaru, the VSS is inside the transmission where it's protected. Sticking a VSS on the outside of the transmission means it's implicitly more vulnerable, even if it is the only way to get a reading on a vintage transmission. Putting it under the CV, closest to the ground, is plain crazy. By contrast, the RMW unit is plasma cut from 10 gauge steel sheet, both the trigger wheel and the bracket. You don't bend the bracket to adjust it (you couldn't without tools: 10 gauge is stout.) The bracket bolts high on the transmission body to avoid debris that might remove it, and you adjust it using locking nuts on the VSS body.

Guess who's VSS I'm buying, even if I have to pay 30% more?

Right. Because I'm getting more than 30% more value.

Finally, I need an exhaust manifold that will fit a T2B Bay, and that won't crack on me. Why do they crack? Because often the chap whacking together his exhaust doesn't realize that for a structure this short, when one end is attached to the engine and the other end is attached the body (or just left hanging!) a crack is almost a guarantee. The engine vibrates and the body doesn't want to wiggle in time with it: crack.

A fully emissions compliant solution for both Vanagon and Bay,
and the components (muffler, CAT, sensors)
are all Commercial-Off-The-Shelf. (COTS)

The RMW solution connects both ends of the exhaust to the engine, so when it vibrates or rocks back and forth, so does the exhaust. It all moves together as a unit. No cracks.

Certainly, the RMW solution costs more than various fabit yerself solutions. It is built to last, not merely just to work until "It's Miller Time." Compared to what some 'write a check and walk away' vendors want for their solution, the RMW design is incredibly DIY friendly.

In the final analysis, I chose to source 80% of my solution from Rocky Mountain Westy because they were the only ones making solutions that were:

1. Engineered for Reliability
2. Cost Conscious (Not cheap, but not a price on the moon, either)
3. DIY Friendly
4. Cross-Compatible Solutions (More on this in a moment)

These attributes come so close to the ideal for VolksarU (Whenever possible, buy NEW specified parts. Fabricate from plans when necessary. Only fabricate without plans when otherwise unavoidable.) How could I not support their business and their willingness to support both T2B Bays and Vanagon?

I'll close with a riff on the cross-compatible solutions that I alluded to above. In my business (high performance computing) vendors are constantly pushing some amazing solutions. There's a dark side to it though: Instead of conforming to standards, where components from different vendors can be mixed and matched by the customer to scratch their special itch, you are sold a monolithic stack of products which all work together, but allow for no substitutions. It's called 'vendor lock in' and it's reprehensible when it is done only to cripple customer choice. Speaking as someone close to engineering, its also understandable. Getting everything to work together is enough of a challenge when you own all of the pieces, so it takes a manufacturer a lot of effort to go above and beyond to try to make their products compatible with some of their competitor's components, even when there are supposed to be standards everyone is conforming to.

While Rocky Mountain Westy isn't 100% compatible with Vanagon solutions from Kennedy or SmallCar, they take every opportunity to be compatible. For example, If you're unhappy with the exhaust you bought that cracked, you're not going to want to go back to the same vendor to buy another one, fully expecting it to crack again. RMW gives you the option to substitute in some of their parts for sub-standard parts from their competition. This is more than just smart business, this is a considerable engineering effort. They're so far out front with conversion products for the T2B bay Window, other vendors who want to get into the space will have to follow them. An arbitrary standard is often better than no standard at all. A well thought out standard (as RMW has managed) is a pearl of great price.

So don't assume that self fabrication is better because it's faster. Even when it is faster, it is what's referred to as a 'point' solution: A solution that is only good for one set of circumstances: one vehicle model configuration, one engine, one transmission. When you do the second one or someone else wants to follow in your footsteps, and any variable changes, another point solution has to be reworked from scratch.

That's why Modularity and re-usability is where its at. That's how you get Reliability, Cost-Conscious, DIY Friendly and Compatible Solutions. That's how many people collaborating can get usable solutions and keep the cost down at the same time. That's why I'm giving away my radiator design: Because anyone can buy that standard part, add some minor fab and have a working solution, a solution that others have time and experience with and can be vouched for by others in your community of enthusiasts.

So don't settle for 'it runs.'

Don't settle for 'It's Miller Time!'

Don't settle for anything less than 'reliably useful.' If that's your fundamental demand, it will guide your buying habits in ways you've never dreamed.