Yeah, I got one.
I got a really cheap battery powered mic pre from Markertek. 'Rolls Pro-Mix Plus'. It's got a sort of phantom power but that's best suited to powering electret condensers directly- only about 18v. It's three channels mixed to two outputs, transformer output. Turns out that almost all the thing is, is op-amps. There are almost no other parts. The op-amps are 4560s and there's two of them in the signal path- one powering two mic inputs, one powering the remaining third mic input, the mixer network, and then another doing make-up gain to the point that you can have a headphone out, or drive the output transformers.
Here's what I did:
Replaced op-amps with Burr-Brown OPA2132PA. Noise level got way quieter, and the general tone got a lot better and less vague and fuzzy. 4560s aren't very good, are they?
Replaced six ceramic 471 (470pf) caps on the inputs with 220pf PST caps (cheap, but decent- no special kind, just 'polystyrene film')
Replaced three 47uf power supply caps with 1000uf. One had to go to the underside of the board for them to fit, and I lifted a trace by mistake and had to solder the cap's leg to a resistor leg to cover for the mistake. This did a great deal as far as permitting huge bass. Some weird electret mic configurations I built had a problem of motorboating when turned way up, but it turned out that this still happened whatever the powersupply cap value- it just sounded bigger when you could actually hear it clearly because of all the bass ooomph. I stuck with the 1000uf caps (I have a boatload of them sitting around)
Finally, the oddest mod- I wanted to bypass the input electrolytics with PST film. There wasn't a convenient way to do this, so I followed the circuit traces only to realize that on one end they went straight to the XLR jacks, and on the other end, there was only a 0.33K resistor between the cap and the op-amp pins. I'd experimented with running the 2132 with no stabilizing cap, and thought that it was a bit too tizzy and detail-happy, hence the 220pf to subdue it but not as much as the original 470pf. At which point I went HMMMM- and tried taking another 220pf and running it directly from the XLR jack to the op-amp pin, bypassing not only the electrolytic that handles 99.9999% of the audible signal, but also the loading resistor. I figured, okay, this will produce a response irregularity, except the stabilizing cap will exactly cancel that out. In theory it would go up to very high frequencies nicely, but without being tizzy, because the stabilizing cap is still there.
That's what happened- to my ear, and as far as I can tell. I'm quite happy with the results and have already used the resulting pre- and it's rather exciting since basically everything in the circuit is now different from the original Rolls cheap mic pre. I figure that if I went to discrete transistors (which I can't yet do, too dumb to figure it out) I'd get another level of performance, but one thing about using the op-amp directly is that it's gotta have headroom from hell, taking just a mic input like that. All in all a fun and pleasing mod- and I have 22 more of those chips just waiting to be used for something