Ribbon Mic High frequency

I took my first crack at re-ribboning some chinese ribbon mics (working on an apex 210).  I practiced a few times with 0.6 micron silver leaf, before moving up to 1.8 micron aluminum foil (or whatever it is, 1.8 micron for sure).  I got it in and tightened up as well as I figure it should be based on what I've read.  (short description, not too loose or too tight.  if I bump it there is a resonance around 40 hz when looking at the frequency response in my daw).

Anyways, my question is,  how is the high frequency supposed to be with these?  I have never heard in person one of the great ribbons (coles, RCA, royer, etc).  My mic is significantly rolled off starting around 2K I believe, and down from there.  However,  it responds astoundingly well to EQ.  I'm boosting the high end like crazy and it sounds delicious, so I'm thinking all is well?

I'm a bit confused by the frequency response plots I see for the "Fancy" ribbon mics - all of those make it seem like mine should be much flatter than it is.  I believe I did something crazy like a +6 DB shelf at 1.8K and a low pass with 6db slope at 250hz, but I got nice vocal and acoustic guitar sounds after adjusting the eq to taste for each.

Yes, it should be much flatter, and sounds like your resonant point is higher than many.  The only resonant note I have off the top of my head was a pair of Fatheads with the Lundahl transformer option, the original ribbons in those had resonance around 17Hz, bad enough someone could open the control room door from the live room and you'd get feedback from mic to monitor speaker, at not very loud volume.  You could look at 'still' air pickup on a spectrograph and see the resonance jiggling around.  I had Samar re-ribbon those, and no longer detect a resonance.  The treble got better too.  Couple dB at 10K is about the most I'd be adding to them.  The Samar ribbons I have need no treble boost 90% of the time. 

Ribbon HF response is determined by the front to back path length, diffraction,and a few other factors.
When the path length equals 1 wavelength there is a null, but a slow decline in response starts long before that.
Combining a fairly short path with a 10 dB diffraction peak and wave plates can allow a flat response to 20kHz with
greater than 1mV/pascal sensitivity. If you allow lower baseband sensitivity arbitrarily high bandwidth can be obtained.

With cardioid and pressure ribbons (like ours) there is no null from path length and 25-30kHz bandwidth is possible.

Les
http://lmwattstechnology.com/