- March 12th, 2018
Trials and tribulations of the 68RFE. Part One: The Valve Body
BD's approach to this challenging transmission.
Although cited as the weak link in the 2007 to present 6.7 Ram trucks, the 68RFE transmission can be a decent unit if you can correct some of its shortcomings. One of those is cross leaks in the valve body.
It is tribal knowledge among transmission builders that valve bores need to be inspected or vacuum checked and repaired and that flat surfaces need flat sanding to reduce cross leaks, but this "old wisdom" does not always result in a "fix" for newer transmissions. The most common cross leak observed with 68RFE transmissions is leakage into the overdrive (OD) hydraulic circuit; this is in fact the only chronic cross leak area we see on 68RFEs. We, like most others, increase line pressure in our 68RFE transmissions to increase clutch holding capability. The caveat of increased line pressure is increased cross leaks, so it is important for us to correct these leaks when making a performance transmission.
Leaks into the OD circuit will cause the annoying P0871 code (set in 2nd or 3rd gear) but are also partially responsible for the burnt OD clutches we see on a daily basis. This code is triggered when the pressure switch in the solenoid pack detects pressure in the circuit when it shouldn't be. Our experimentation shows this switch triggers at 16PSI, this just so happens to be right around the pressure required to apply the OD clutches (air check this if you don't believe us). This means by the time you're setting P0871s, you've already been dragging the OD clutches.
Certainly there could be minor leaks from many areas, but our exhaustive testing shows there are only two areas we need to concentrate on to correct this leak. 1-The Solenoid Shift Valve (SSV) and 2-The Valve Body Separator Plate.
Part I "The SSV"
Begin rant… The reason behind the existence of this valve is nothing short of ridiculous and Chrysler should have done away with it decades ago. This is a small assembly of valves (the ones we're worried about look like little pucks) that direct the output of the LR/TCC solenoid to the torque converter clutch when in 2nd gear or higher and direct fluid flow to the low/reverse clutch when in first or reverse. This was done to use one solenoid for two purposes and originated in a time when Chrysler transmissions had only 4 solenoids and solenoids were expensive. Today one would think this had been done away with and the cost of 7 vs 8 solenoids to be negligible but here we are in 2017 with the SSV valves being backwards compatible with those from a 1989 Caravan. Rant over.
The problem with the SSV is the valves aren't well supported axially causing uneven wear, and worse yet since the transmission solenoids are pulsed on, these valves can oscillate rapidly during clutch apply leading to increased wear. This allows fluid from the other clutches (2C, 4C, OD) to leak into each other. It is perhaps no surprise that the OD circuit is between the other two, so it is the most susceptible to leaks.
The common fix for years was to ream this valve and insert oversized valves (remember this is our old wisdom talking). This worked great, we would put a valve body on our custom 68RFE valve body tester (equipped with a digital OD circuit pressure readout to boot) to measure leakage, "fix" the valve bore and put it back on the tester. All better! The problem with this fix that the surface finish from reaming/honing the bore is nowhere near as smooth as factory, leading to accelerated wear – faster than the first time around. We observed that after 1-2 years of heavy use, the leakage in a repaired valve bore was right back to where it was when we started - Not a permanent fix.
The fix? The early model 68RFEs had normal cast aluminum valve bodies with no coating on them, the later model valve bodies (circa 2010 and up) came with a hard anodized coating on them, identifiably by the darker color of the part. This hard coating substantially reduces wear in the valves and makes a remarkable difference in service life of the SSV valve bore. Through our 68RFE development program we have found that hardly any of the core transmissions sporting anodized valve bodies have worn through this hard coating in the SSV area. ALL BD 68RFE transmissions get new hard anodized valve bodies, don't fall for the "fix", reaming is a cheaper option but for this valve it is not the right one.
Part II "Valve Body Gaskets"
The valve body separator plate is a simple stamped or laser cut sheet of steel sandwiched between the halves of the valve body. It directs fluid flow between the two valve body aluminum castings, contains orifices and acts as the seal between the two halves.
It is normal for valve bodies to leak a little bit in this area, indeed a small amount of cross leaking is inevitable in all transmission valve bodies due to their design. The problem with this transmission is that the leakage at this surface is not negligible. In some cases it can get to be enough to actually generate fault codes.
Scan tool showing pressure switch oscillating occasionally, then more rapidly. Note the error counter must reach maximum for a fault code to set. Here you can see its teetering on the edge, but hasn't gone into limp mode. This means even if the code isn't setting, it doesn't mean there isn't a problem cooking – err, clutches cooking.
The leakage seen due to the separator plate is not something we can accurately mimic on the valve body tester, not even on the transmission dyno. The only way we've been able to observe real world results is by installing a battery of sensors into the transmission hydraulic circuits and strapping the truck down to the chassis dynamometer.
Here we have modified a transmission oil pan to accept bulkhead fittings, installed pressure transducers (sensors) and connected very small nylon lines to the valve body to get accurate and fast response pressure readings. Analog gauges with rubber hose are far too inaccurate for this type of testing and length of rubber hose acts as an accumulator or buffer to the readings. Our sensors are wired into a high speed data logger that is also tapped into engine/TCM sensors.
To best replicate the fault, we observed the conditions that caused a P0871 to set in a late model truck (that already had the anodized valve body) and recreated those on the dyno. To keep all equal, we held the truck at 2000RPM, 3rd gear, put between 200-240 wheel horsepower of load on it and held that for 1 minute intervals with brief cooldown periods in between. Each time the transmission temperature would climb somewhat and we would continue repeating the tests until it was at least 200 degrees Fahrenheit in the transmission.
To our surprise on only the fourth test the problem began rearing its head. It was then possible to replicate every run after that while the transmission was hot. The graph shows line pressure in the 230psi range with a leakage into the OD circuit (orange) of about 16psi. The green line is attached to the TCM pressure switch wire. Note it starts flickering just as we get to 16psi from 15psi. What happens when it's hanging out at 15psi continuously and lightly dragging the clutches without setting a fault?
We tried fourteen different tests to work at eliminating and reducing this leakage. These tests included overriding solenoid control to eliminate the periodic pulse tests, staking check ball seats, flat sanding various pieces, changing separator plate surface finish, changing fastener torque values, and combinations of the above. Finally, we custom made a set of 1-off paper gaskets cut with our laser etch machine (wow what a cloud of smoke that made) note the blackened edges.
Once we aired out the smokey new gaskets, pressure sensitive film was used to check clamping force on the sealing surfaces. The gaskets provided a much more even sealing surface and gave us a strong indication they would be an improvement.
This picture shows the before and after of pressure sensitive film sandwiched between the separator plate and the valve body casting. On the left is a stock valve body with separator plate, on the right is the same valve body with gaskets installed. Note there are no longer gaps where there is insufficient pressure on the valve body separator plate. This means we expect it to seal properly.
Early production gasket taken back out of service for inspection. Note the very clean indentation created from the mating surface.
Gaskets for the win! Dyno testing showed we were able to get our OD leakage pressures down to a mere 5psi. What a difference from the 16+psi we started with. Remember, this is 5psi at high temp, high load, high line pressure. This is well within the realm of normal cross leaks from valves and other areas of the valve body and well within our comfort zone.
Data from testing with the gaskets. Note the shown point is near the end of the last (hottest) run, we are only at 4psi of leakage at 252psi line pressure.
All BD 68RFE transmissions come equipped with our special gaskets and separator plate, a combination which is designed to handle some of the highest 68RFE line pressures in the industry. Our transmissions are intended for 250-260psi at full load. These gasket and plate kits are also available separately from the complete transmission in a kit with a pressure controller and as a bare kit. Keep in mind if you use these we strongly recommend you upgrade the valve body to the late model anodized unit to eliminate SSV leakage as well.
Links to parts applicable to this article:
1030362 - BD 68RFE ProTect68 Pressure Control Kit Cummins 6.7L Dodge 2007.5-2017
1030373 - BD Protect68 Gasket Plate Kit Dodge 6.7L 2007.5-2017 68RFE Transmission
1062023 - Build-It Trans Kit - Dodge 2007.5-2017 68RFE Stage 3 Performance Kit
1062025 - Build-It Trans Kit - Dodge 2007.5-2017 68RFE Stage 4 Master Kit c/w ProTect 68
1064262 - Transmission - 2007.5-2017 Dodge 68RFE 2wd
1064262B - Transmission - 2007.5-2017 Dodge 68RFE 2wd c/w Billet Input Shaft
1064264 - Transmission - 2007.5-2017 Dodge 68RFE 4wd
1064264B - Transmission - 2007.5-2017 Dodge 68RFE 4wd c/w Billet Input Shaft