Extraordinary shift control

All things mechanical eventually wear out or break down. Everyone knows this. It is the most well established of all scientific laws. It’s called the law of entropy. The law or principle of entropy is actually an extrapolation of the second law of thermodynamics, but you may know it better as Murphy’s Law.

Remember Murphy? “If anything can possibly go wrong, it will.” There are a thousand humorous variations of this reasoning, but it really is a scientific reality. Simply put, everything moves toward an increasing state of disorder. This is not speaking only of large scale systems, it happens on the atomic scale. And it is amazing to realize how many of the mechanical failures we repair actually begin on the molecular level. Now, where am I going with this?

Very early in my experience with transmission repair, I began to realize that in the automotive industry, we are dealing with multiple levels of disorder. On the one hand, we have to deal with the inevitable failure of parts like bearings, washers, bushings and so forth. But there are systems in transmissions that have built in engineering flaws. That is, the system itself is designed with some feature that, if done correctly, would extend the life and functionality of the transmission. Yet the manufacturer continues year after year, and even model after model, to build in the same defects for several generations of vehicles.

Some call this “planned obsolescence,” but I don’t think it is always deliberate.

In fact, sometimes it happens quite incidentally. Design and field service engineers come and go, get promoted, leave for a better job (sometimes with a competitor), quit, get fired and so on. When they leave, the wisdom and experience they gained goes with them, and when the new guy that comes in, the learning curve starts all over again. Had they continued on the job, they might have passed a field service recommendation up the line to the design engineers, but due to the break in the organizational system (entropy again), the problem never gets addressed. Let me give you a real and perfect example of exactly what I mean.

Earlier this year, I completed a series of valve body and accumulator upgrade kits for the Honda/Acura 4 and 5 speed transmissions. These kits (available through Superior Transmission Parts) were engineered to correct several chronic defects in the Honda/Acura transmissions. There were some major areas of concern, but one area we worked on was related to shift control and shift quality. That is, fixing erratic shifts (all gear changes), along with correcting the notorious slide-bump specifically on the 1-2 shift. First, lets talk about the erratic shifts.

It is natural for transmission and engines to produce microscopic particulates that get deposited in the fluids, which of course is why we change fluids and filters. In the case of engines, other contaminants like carbon from combustion and small amounts of fuel find their way to the sump. In the case of transmissions, these particles can interfere with valve operation. The particles get lodged between valve lands and the bore and along with expansion and contraction, remove the bore clearance causing valve drag.

In the case of a shift valve that moves one time per shift, this is not as critical. But when it comes to high speed regulating valves like the clutch pressure control (CPC) valves, that work in concert with the computer and linear (PWM) solenoids to ramp up clutch pressure during shifts, contamination and loss of bore clearance can disrupt the regulation cycle and cause erratic shifts, slight slips, flares bumps and even clutch failure. And of course it is impossible to prevent any system, whether engine or transmission, from producing at least some contaminants, even if only a very small amount of linen fibers from the converter clutch during the break-in period. To make matters worse, we’re talking about a trans with a non serviceable filter. It’s inside! And as the system ages and produces other small particles from here, there, and everywhere, the system must be kept functioning.

So when we apply Murphy’s law (oh yeah, that guy again) to an OE-style valve as shown in Figure 1, it goes like this: “If there is a stray piece of carbon anywhere in the trans, it will end up in the CPC valve.”

Well, correcting this is not something a little bit of American aftermarket ingenuity can’t fix. But in order to defy Murphy, sometimes it is necessary to go out of your way to build in redundant fail-safes. Comparing Figures 1 and 2, there are some very obvious new design features. The OE style valve has three lands, whereas the aftermarket valve has 11. The spaces between the lands are oil galleys. These help more evenly distribute a film of transmission fluid across the surface of the bore, as well as provide a place to collect and pass along contamination as the valve cycles.

The new self-cleaning design restores correct bore clearance and improves flushing, and the obvious reduction in surface contact reduces inherent drag that results when fluid can not easily pass across the land surfaces. The retainer plug with O-ring furnishes superior sealing and eliminates plug leaks so that response to linear solenoid pressure changes is instantaneous and at true value. It proves to be a fantastic design that fully restores functionality with long term high reliability. It should be obvious to you at this point that in addition to doing an immaculately clean rebuild, successful repair on these new transmissions also depends on some parts with advanced design features that enable it to function reliably under adverse conditions. After all, the CPC valve is the most chronically sticking valve in the whole transmission.

Now let’s turn our attention to the slide-bump 1-2 shift. This slide-bump or slip-bump shift is a very common defect in these transmissions and is often associated with a burned out or blackened second clutch. The trans might come in with some other failure (typically converter lockup clutch burnout), but when the trans is disassembled the second clutch frictions are usually over 50 percent black, and the steel plates are very shiney with blister spots. The clutch has been operating with a controlled slip for quite some time. The blacker the clutches get, the worse the slide bump becomes. Some even get so bad as to be a long slip with a hard thump at the end. And it is quite uncomfortable.

It is very natural after tear-down, for the technician to see the clutches and think, “There is the cause of the slide bump! The second clutches are burned.” When in fact the opposite is true. It is the slide bump that burns the clutch!

But here is the ironic part. The valve body upgrades I mentioned earlier fit 1998-2007 vehicles, but this slide bump 1-2 shift problem has been around as long as I can remember. This is not a recent development. When I first started rebuilding Honda/Acura transmissions, back in the days of the two shaft 3 speed F4 and G4 trans (late 1980s and early 1990s), they would come in the shop with this issue. So for the last 20-plus years, I have been routinely fixing this problem. It was actually the first Honda defect I learned how to remedy. But Honda did make one more recent change that amplified the problem.

Figure 3 shows the 1-2 and 2-3 accumulators from an Odyssey B7TA. Notice that the 1-2 accumulator piston is slightly larger in diameter than the 2-3. The earlier two-shaft units had the smaller 1-2 accumulator, and though it still was calibrated with the same basic characteristics, it usually would take longer for the problem to manifest. If the vehicle was driven mostly on the freeway, it might take 50,000 to 80,000 miles to blacken the second clutches. But if it was used mostly in town light to light, after 20,000 miles you could definitely start to feel the tail end bump on the 1-2 shift. The front end of the shift got longer (slide) due to the poor accumulation.

It’s my guess that Honda engineering, when introducing the 3 shaft 4 speed unit, and with the Odyssey (heavier class vehicle) in mind, made the accumulator piston larger with the intent of furnishing some extra absorption to prevent bumpy 1-2 shifts due to the wider average throttle opening during light acceleration because of weight. But in doing so they shot themselves in the foot. I think they realized this, because the subsequent 5 speed 3 shaft models again have the smaller accumulator piston.

But let’s take a look at the 1-2 shift to discover exactly how the slide bump occurs.

Take a moment and look at each detail, so as to understand what you are looking at in this graph:

The 0.710-inch measurement below the graph is the total accumulator piston travel. The top of the graph indicates this travel is divided over time. The vertical ascending numbers on the right side column represent clutch pressure in 10 psi increments. There are two sloped lines on the graph. The longer represents the longer outer accumulator spring, and the shorter steep angle line represents the inner. If you have rebuilt this trans and handled the accumulator springs, you may recall that the outer spring is not very stiff, and easy to squeeze in your hand, whereas the inner spring is much shorter and very stiff. Keep this in mind as you look at the graph. Those spring characteristics are clearly visible.

The lightest second clutch pressure switch is the green (28 psi). The Odyssey uses the black (32 psi) switch, so the clutch start pressure is a little bit higher than the dotted line on the left. Notice that the tension of the OUTER spring only increases from 21 pounds to 34.5 pounds through the first half of the accumulator stroke before the piston contacts the inner spring. But if the second pressure switch sets the clutch START psi at 32, then the accumulator rapidly sinks in its stroke and only gains 4 psi by midway.

This is essentially flatlined! Slide…and then the pressure jumps up suddenly when it starts to ramp up steeply on the inner spring. Bump. If you stop for a moment, sit back and just look at the graph you’ll see it is a very clear visual representation of the slide bump.

Now, prior to our lengthy research on Honda, we had heard some complaints of a light to mid throttle shudder on the 1-2 shift, and these complaints were allegedly associated with GPX plates. After a lengthy investigation, we’ve come to some conclusions. But first I want to explain the difference between a slide bump and a shudder.

The slide bump is a long slip due to insufficient clamping force, then bump occurs when the pressure spikes suddenly at the tail end of the shift cycle. Hence, “tail end bump.”

A shudder, in the simplest terms, is a rapid slip grab slip grab slip grab. It happens when the clamping force is not sufficient to hold against the applied torque. The co-efficient of friction reaches the sheer point (the clutch is no longer able to hold against torque under the supplied clamping force), the clutch breaks loose, but as soon as it does the tension is released and the clutch locks up again, and the cycle repeats. This is analogous to lifting a table on one end and dragging it across the kitchen floor. With no one on the other end, the legs will shudder across the floor surface. It also makes a resonant sound. This too occurs in the vehicle.

Now, here is the craziest part of this whole thing. The GPX plates have a pretty high thermal capacity (the clutch can stand more heat than is generated by the shudder), so you can do 20 or 30 consecutive 1-2 shudder shifts, then down the trans, pull it apart, and those clutches look perfect! Then you think to yourself, “What is going on here?”

But it is important you understand that the shudder only occurs because the first half of the accumulator stroke is flatlined. In fact, it is the excellent characteristics of the friction plates that keeps it from failing under such adverse conditions. A plate with a lower thermal capacity would burn up quickly.

A high quality good feeling shift is achieved by initially applying the clutch just below the psi at which it begins to pull the vehicle weight at minimum throttle (around 20 psi) then ramp it up quickly to squeeze it hard enough for good torque transfer without overheating the friction plates. The effect of the re-engineered accumulator springs shown in Figure 4 is illustrated in Figure 6 (the second graph). Notice how the steep pressure rise actually completes the shift in half the time, and passes thru the sheer threshold of the friction materials very quickly by ramping the clutch pressure in a linear fashion.

Concerning clutch plates: In the last two decades clutch plate design has taken a significant leap forward. I have been especially impressed with Alto. Their research and development lab with several full time engineers has made significant advances with clutch paper development. So I don’t hesitate to say that when the shift is properly calibrated, it doesn’t matter which plate you use. Fuji, GPX. Borg Warner, Alto, they are all high quality plates that work well beyond any reasonable expectations. The superb shift quality is proof these shift concerns are not in any way caused by substandard clutches. You can repair confidently and reliably when you know from where the real causes of problems come. 

Subscribe to Motor Age and receive articles like this every month…absolutely free. Click here