I had an engine failure in flight a few days ago.  We were about 3 minutes out of the field and about 1200 feet AGL when there was a loud pop and a moderate level of vibration started. A liquid appeared on the canopy but it did not look like oil. A quick 180 was done and power reduced to the lowest level to maintain altitude  until we were with in glide distance of the airport (perhaps 45-60 seconds). An emergency was declared and we landed downwind in the opposite direction.

The Honda engine behaved flawlessly even with what turned out to be an almost total loss of coolant. We were able to taxy back to the hanger leaving a trail of coolant behind.

The failure was that the "spider/plate" on the back of the gearbox lost one of its three sides. It exited the aircraft making a hole in the cowling and badly damaging the coolant pipe and radiator.

On looking on line, it seems that this failure mode seems to have been happening more frequently. Of course since this is in an experimental, there is no central place where all these failures are reported.

I am now quite sure that these failures have all been caused by miss alignment between the gearbox shaft and the engine crank shaft. The gear box is mounted on long standoffs which are non precision and do not offer any way to insure the alignment of the shafts. Jan provides an alignment disk when assembling the gearbox but that does not insure anything as if there is any misalignment, the disk will be slowly destroyed. Additionally the mounting standoffs are positioned asymmetrically so that the engine and prop torque will introduce additional alignment errors. Even a few thousands of misalignment results in a flexing force on both the spider and flywheel. This is happening 5000 times per minute. Only spring steel can withstand this but that alloy would not be suitable for these applications. The result is metal fatigue and ultimately failure.
I have noted to others that we have known for almost 100 years now to mount a transmission on an engine. It is done with various variations of a bell housing and alignment pins or the equivalent. Jan thinks he knows better and provided a system that has no alignment mechanism.

If this was a certified/production aircraft, I am sure that there would have been an AD grounding the fleet until there was a proper fix. Personally, I would ground all viking installations. For those who continue, I would frequently replace the spider, rubber isolator and flywheel, probably at intervals of no mare than 100 hours. Even that may not be enough as I believe there was one failure in under 100 hours.

I have included pictures of this event.

I am retiring the viking package and may put a UL or Rotex engine in.  I see no way to modify it to make it reliable short of designing a whole new gearbox. Even then since jan chopped off one of the Honda alignment pin holes, making anything that would align like a Honda transmission is probably not possible.