The gearbox is used to bring the prop speed down to an acceptable limit – the engine crankshaft will turn between 4000rpm and 5500rpm to be in its optimal torque range, whereas the prop needs to turn at a max speed of around 2500rpm. This speed is determined by the diameter of the prop- you want to aviod the tips of the prop going faster than the speed of sound, so the bigger the diameter the slower the prop must rev.
In the Crossflow gearbox two spur gears are used, with the larger gear having the teeth cut on the inside of the gear circle, and the smaller gear has its teeth on the outside in the more conventional manner. The main shaft is supported by two big bearings and at the end of the shaft, near the propeller, there is a collar that is used to intoduce high pressure oil from the prop Constant Speed Unit (CSU) or governor, through a hole in the prop shaft into its centre. From there it pushes up against the base of a piston in the propeller and this then changes the pitch of the prop, which is controlled by the CSU.
The CSU is mounted onto a pad or base on the gearbox, which is machined to a specific, fairly universal, pattern (AND20010) that is known to both the gearbox and CSU manufacturers. The CSU takes engine oil at 3-5 Bar and pressurises it to 22Bar and then either delivers it to the prop or returns it to the engine, depending on what is required by its speed setting.
In the picture the oil is delivered from the engine at the top of the pad (the silver/ aluminium coloured bit) via the blue AN fitting. It then exits out the pad into the CSU via hole A. After being pressurised it goes back into the gearbox body through hole B and is channeled to the collar via drillings. The collar is to the left. At the bottom of the pad, opposite the oil inlet pipe and hole A is the return line from the CSU to the engine. The flexible line that is to the left and below the pad (the one with the 90 deg bend) is what gave us our first problem. This is an oil return line from the collar back to the engine sump. Now that is just fine if you are not using a CSU as it circulates oil to the collar and back to the engine, or simply drains any oil from the collar back to the engine.
However, it does not allow any pressure to build up in the propshaft and so the piston does not travel and the prop will not cycle (change pitch). After a while I figured that out and blocked the return line, like so:
Great, now the prop cycled and we ran the engine and started tests on it. But after around 10 minutes of running oil started pissing out of the gearbox oil resevior. The gearbox has its own oil bath, which has oil that is completely different to the engine oil. Trouble was, the oil that was coming out of the gearbox resevoir looked to be engine oil… With no support from Crossflow we decided to strip the gearbox down to try and figure out what was going on. Here is the ‘naked’ engine in the front of the aircraft, now sans gearbox:
So now I opened the gearbox carefully, dreading perhaps hearing ‘sproing’ as some spring or equally vital part jumps out and dissapears into the same place that single socks go to; wherefrom they never return…
But it turned out to be fairly simple and the problem was immediately obvious. Some oil seals had burst. After going to SKF and getting a specialist to identify these it turns out that they are rated for a maximum of 5 Bar. So 22 Bar popped them in short order. Why the heck did Crossflow put 5Bar seals in an area that would see 22Bar? Who knows? I think that perhaps they were hitting a cash flow crunch (familiar to most in the aviation business) and shipped the box out, maybe hoping to sort it out later…
Here is the collar with the damaged seals. The bottom left arrow shows where the oil comes into the collar from the CSU; it then turns and comes out the hole indicated by the right top arrow. This hole aligns with a hole in the prop shaft and, even though the shaft is spinning around it is spinning in a pressurised chamber so the oil enters the centre of the shaft and pushes the piston. You can see the damage to one seal between the two arrows. These seals now have to be replaced with custom made high pressure seals that are being made in Austria for us.
Hopefully that will sort out the gearbox issue and we can get back to fine tuning the engine…
Posted by Jay Hyde
Rawhyde's Blog- a subsection of HorribleHyde 
Rawhyde,
I am currently working on a 350 hp crossflow engine with the same PSRU as you. I think you will find that the return line from the collar needs to be installed to allow proper flow and pressure control. We had a hard time bleeding the air out of the propeller hub during first engine runs but now have successfully bled and the propeller is cycling properly and quicky. If the engine is installed on a tail dragger as it looks like, then you will need to make sure the oil is serviced properly by levelling the aircraft, service the oil to the proper level then putting the aircraft back on the tail wheel and remark the dipstick so that you can service the engine to the proper level in normal configuration. I am not sure what the original seal part number was and if ours is the same, but we have had no leakage problems to this point.
Please let me kow if I can help in any way as I have learned alot in the last few months.
Shawn
Can anyone tell me what the part number for the oil seal between the re-drive housing to the prop shaft is?
Hi my name is Jon and I enjoyed this article. I too have a crossflow engine And I was wondering how its performing for you.
I have a crossflow 24T mounted in a Zenair 801. I lost faith in the engine’s reliability some years ago after flying it a little bit, particularly the ECU and its program. Never started well, ran poorly, sputtered and coughed, especially on take-off, yuck! It came with a funky crank position sensor, no knock sensor, no igniters (!!! – tho’ maybe the voltage output from the old p8 was enough to spark and thus it didn’t require one, I’m not sure), and no cam position sensor. All very unusual. It had the old DTA Fast P8 Pro ECU that was no longer supported by DTA, the tuning program (MAP programs) seemed barbaric to those skilled tuners I showed them to, and in any event the ECU itself had severe programmability and connectivity limitations, very 1990’s. I installed an all new VIPEC V44 ECU, new stock crank trigger wheel, new crank sensor, added back a cam position sensor, added a knock sensor, and added a second complete set of engine monitoring sensors since I lost trust in the ones installed by Crossflow, and so that I could compare readings. I rewired the entire engine with an all new harness following the VIPEC spec. I had the ECU MAPed (programmed) by a Subaru rally engine tuning genius (Cameron Koole, Lightspeed Innovations, Red Deer, Alberta, Canada) starting with a conservative stock MAP file set for the Subaru EJ-20-G (which I am ALMOST convinced is what my engine is) and then thinking hard about failure modes and limits and how airplanes differ from cars, we tweaked the MAP for what we expect and hope is a reasonable set of failure modes and limits for airplanes. This engine now runs so much better, starts easy in all conditions (cold, very cold, hot, very hot), runs cooler, better fuel efficiency, can handle lower octane fuel, has far more power, and the modern ECU lets my tuner monitor the engine and tweak the tune in real time during ground runs and high speed taxis. (obviously we don’t do that in the air). I am also carefully watching the crossflow PSRU as I have heard horror stories about those. Mine seems well-built, and I do not have the governor option so my PSRU is the simpler version with lower pressures. I have replaced all seals, and I added a PSRU oil cooling system with a thermostat control and manual override, as well as a PSRU temp sensor. The PSRU cooling system is basically a turbo scavenge pump (Exapump) on the firewall linked to belly mounted transmission cooler from JEGs. It works great. I also added an engine oil cooler to the cooling system mix.
Do you still have the Cross Flow engine installed in your 801 and do you still fly it? If so have you had trouble with the PSRU? I believe I may have the same engine which I have not started yet. Just looking for Info on this.
I have a crossflow cf4-20t that was delivered in 2003. It has not been started yet. Trying to decide what to do with it . Try it or scrap it.