Consider the diode characteristics

These datasheets are available in a couple of different places. They were developed by Ben T. and Mike T.

They show a thing called Ideality factor or N on the chart. In theory 1.000 is ideal. Look at n and Ro.

I did a little cut and paste to isolate the diodes we have been using in out sims.

Using Avago's design program to compare n at 1,2,3,4

If you compare 0db you will see n=1 is the highest output. Looking back at the chart you can see the BAT46 has n=1.1 and the IN34a n=1.3 yet the 1N34A output soars? Look at Ro. The 1N34A is 35 - 40K while the BAT46 is 175-100K. Could it be the lower Ro? Look at the D18 with Ro of 160-170K it soars?

So that brings me back here. I put reference at L2 - L3 and measure to the other end of the coils and then to ground. The output is = the voltage across L2 & L3. When R1 is about 40K the output peaks. Lowering R1 will lower the output but raising R1 has no effect on the output. 

 Food for thought.