I've had a chance this weekend to make a test I've been wanting to for awhile. Tuning the ODA's snubber for the specific transformer being used, now that the WAU16-2400 is the pick for the standard +/-12.5Vdc power rails. Turns out changing the 110R resistor to 52R will result in being closer to critical damping with this specific transformer.Details:
The snubber in the ODA comes right out of Hagerman's famous paper:http://www.hagtech.com/pdf/snubber.pdf
0.01uF (C5) to lower the ringing frequency, then the series 0.068uF + 110R pair (C8, R13) to damp what is left.
The rest comes out of Mark Johnson's excellent Quasimodo thread:Simple, no-math transformer snubber using Quasimodo test-jig
As mark points out, a person could go to a lot of effort and math measuring the leakage inductance of the specific transformer winding and plugging it in the equations. Or...simply replace the 110R with a pot, as Mark did, and adjust for critical damping. Then read the value off the pot. Simple and easy! In my case I used a 500R linear 25 turn taper. I think Mark used a 1K in his writeup.
My appologies upfront for the fuzzy photos below. I should have had the room lights on instead of just relying on the scope screen's brightness.
The scope is across one of the ODA's power rectifier diodes, D1.
* The first photo shows the transformer's secondary leakage inductance ringing in the lower right corner of each half-cycle, that small part that fades out at this level of resolution. This is with the orignal 110R resistor.
* The next photo is a closeup of the ringing with the scope's timebase stretched out. This is still the original 110R resistor. I can see the ringing dying out after one cycle. Not bad, but we can probably do better.
* The next photo shows the test setup, with the 500R linear trimpot soldered onto the SMD pads where the 110R R13 used to go. The blue wire is going to the end-munted end of D1 for the scope probe. I pre-set the trimpot to 110R on that side so it would start out the same as the original resistor. I've soldered the pot directly on the board like this to keep from introducing any additional inductance or capacitance from using wire leads.
* The next waveform shows pretty close to critical damping at 52R. For each of these I unsoldered the pot, measured it, then soldered it back in. Note I've increased the vertical scale on the scope from the first two pictures to show the ringing more clearly.
* The next waveform is with the pot in the other direction at 190R. Clearly we have ringing! I count three cycles before it damps out.
* Next I just turned the pot until the waveform was damped inone cycle and measured the pot: 102R. Nearly the same as the 110R.
* The final waveform is back to the original 110R resistor.
If you use a different transformer than the 16Vac 2.4A in the BOM and have access to a scope, just use this procedure to tune the snubber. I would suspect though that most transformers in this general voltage and amperage range will be snubbed adequately by something in the 52R range. 51R or 49.9R would be standard values