Mixing Balanced and Unbalanced mic input to HD-P2
 

I'm setting up some mic comparisons. Ideally I would like to have left channel / right channel comparisons. We'll be using an HD-P2 to record samples. There is a single switch that controls phantom power to both channels and an admonition in the manual to not use phantom power on unbalanced inputs. Keeping the input chain as simple as possible, is there a way to bring in a balanced and unbalanced mic level signal to the recorder?

The best way would be to use a balun (or unbal in this case). This is a simple circuit with an unbalanced input and a balanced output. It is a very simple circuit to make up with a small transformer but I would think you ought to be able to buy one somewhere. I couldn't find one with a cursory web search but perhaps someone else will know of a source.

A second way to do this is to connect the unbalanced microphone to the balanced input through DC blocking capacitors. These will solve the problem with the phantom voltage but the capacitors, if not large enough, will reduce the low frequency response.

Just discovered one the other day - the Rane "Balance Buddy" converts between balanced XLR +4dBu and unblanced RCA -10dBv. Not cheap though at a couple hundred bucks.

Exactly which unbalanced mics do you want to test?
I know that an actual simultaneous run would be ideal, but you could keep the conditions as similar as possible and record in two separate passes.
For mic comparison, I think that would be close enough to make a choice and might be preferable to easily available transformers that aren't specifically for this purpose.
There are other less expensive transformers available, although several are designed to work at line level and one isn't very high quality.
The one from Radio Shack that's high-impedence unbalanced 1/4-inch female in and balanced XLR male out will work, but may have some attenuation. If it does attenuate, that would seriously compromise your test. I don't remember exactly.
The Ebtech Hum Eliminator would also probably work and doesn't have any attenuation but it's designed for line-level signals. I haven't had mine hooked to a phantom input, so don't sue me if it blows up your phantom supply. It's only $60 online and it's great to have on hand for other uses.
There may be some active direct boxes that will let you use a mic in addition to their normal duty of using a guitar or other high-impedence instrument.
I think i'd just run your balanced and unbalanced mics separately under the same conditions for testing.
If you really have to operate in the field with two types, I'd get either a separate phantom supply, or a transformer/preamp specifically made for this job. Or just buy balanced mics.

This is going to be sort of an open comparison where people who have a mic they want to be included will bring it to a conference, so I don't have a clue yet which models might end up being evaluated.

I think I may be adding undue risk and cost by trying to facilitate side-by-side comparisons of XLR and non-XLR mics, so I think I'm going to restrict the side-by-side comparisons to one type or the other. Given that scenario then I think I'm correct in assuming the simplest input chain for a non-XLR mic would be to connect the mic to the XLR input via a 1/8 or 1/4 phone jack to XLR converter, assuming the mic provides its own power. Or would an impedance matching circuit provide a better sample?

You could use a transformer mic splitter. The transformer output would go to the HD-P2 the direct output would not be used.

Sam