Well you're right, I am using the Ocelot and Jay Hogg's IR tools (score 100% for your ESP! ;). As for the protocol carrier frequency, I'm able to learn the remote's commands reliably and I'm specifying 38kHz as the frequency when sampling with the Ocelot. I know that this isn't a definite identification method however. If you think you can see what type of encoding it represents, here is my learned code (converted to Pronto format using Jay's tools) for the power on/off toggle:

0000 006c 005d 0000 0084 0042 000f 0010 000f 0032 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000d 0012 000f 0012 000d 0012 000f 0012 000d 0012 000f 0032 000f 0012 000f 0032 000f 0032 000f 0032 000f 0032 000f 0010 000f 0012 000f 0032 000f 0010 000f 0032 000f 0032 000f 0012 000f 0032 000f 0010 000f 0032 000f 0b1e 0085 0042 000f 0010 000f 0032 000f 0012 000d 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0032 000f 0032 000f 0032 000f 0032 000f 0012 000f 0010 000f 0032 000f 0012

I also tried the code that was supposed to work with PV-9662 found in another thread:

0000 0071 0000 0032 0080 0040 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0030 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0010 0030 0010 0010 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0030 0010 0010 0010 0030 0010 0030 0010 0010 0010 0030 0010 0030 0010 0010 0010 0a9c

and even "adapted" it to look like the learned on/off code (with the different header style and adding the equivalent repeated section which gives:

0000 006c 005d 0000 0084 0042 000f 0010 000f 0032 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000d 0012 000f 0012 000d 0012 000f 0012 000d 0012 000f 0032 000f 0012 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 000f 0010 000f 0032 000f 0032 000f 0012 000f 0032 000f 0032 000f 0010 000f 0b1e 0084 0042 000f 0010 000f 0032 000f 0012 000d 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0032 000f 0010 000f 0012 000f 0010 0010 0012 000f 0010 000f 0012 000f 0010 000f 0012 000f 0032 000f 0010 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 000f 0032 001f 0012

but it still doesn't work, although I'm still very new to IR encoding so this might just have been a shot in the dark. If you see an obvious fallacy in all this, tell me!

I know that as an Ocelot user, I can always make a current draw detection circuit (I made one using the core of an old tv flyback transformer) but I'm trying to avoid that if at all possible. Thanks for any suggestions.