I ran some quick simulations on the Universal EQ circuit, just to get a feel for how it would work. My tests were far from comprehensive as I only tested the three inductive bands in boost mode. That said, the results were very positive.
I started with two boosts maxed and the third set flat, all with their Q-pots set to their most narrow positions and with no Q-max resistor. With the bands set far enough apart that their operating frequencies would not overlap there was no noticeable interaction between them. A pleasant surprise was that, even with the bands set close together (the mid band at ~1.5k and the high band at ~3k), there was negligible frequency shift (the lower band moved downward and the upper band upward by about 40Hz each - less than a 3% shift) and the two bands retained their individual peak shapes rather than combining into one massive peak - both things that did occur with the Poor Man's mid boost. Adding the third boost at ~800Hz caused no further frequency shift in either of the other two bands, and was itself only shifted downward by about 6Hz in comparison with the position of that boost by itself. In summary, having three narrow boosts at ~800Hz, ~1.5kHz, and ~3kHz, all maxed, produced minimal interaction aside from some narrowing/distortion of their curves.
This brings up another great advantage of using the Pultec Hi-Boost/Helios Mid Boost circuit (the two are identical except for the presence of the Q pot in the Pultec circuit and the cut switch in the Helios one, and in fact Ian's drawings include the Q control in the mid and low bands as well which I think is a great idea) over the Poor Man's Mid: not only does it solve the issue with minimum boost, but it also allows for narrower curves on the boost side (and I would suspect the cut side as well, though I haven't tested it yet), which in turn allows for the mid and low bands to be located in more useful frequency ranges in conjunction with the 10k input impedance.
Of course, as we decrease the Q or reduce the amount of boost (each of which also accomplishes the other, though unlike with the MEQ5 circuit the two are not redundant and do actually produce different Q-values at the same amplitude), we see the peaks combine as expected. However, the locations of the peaks remain essentially constant and the shape of the resulting curve predictable. Not that interaction is necessarily a bad thing - the Trident A Range has significant interaction and is still considered one of the greats, albeit a bit finicky to use - but it's nice that this EQ appears to respond in a predictable manner.
One thing that I did find is that I actually preferred the response of the circuit with the three boost pots at a size of 47k rather than 100k, for a few reasons. First, the smaller pot size allows for narrower possible boosts, which again translate into lower useable frequencies for a given inductor. For instance, going with the 47k pots allows for a 1.5k boost in the mid band which very closely matches that of the Helios 1.4k at lower boost levels (which IMHO are the most important considering that one rarely boosts anything 18dB but quite frequently boosts at 0.5 to 6dB levels), whereas the 100k pots made the curves too wide even at the narrowest setting. To be more specific, at 1.5k using the 2H tap, 47k boost pot, and 10k Q pot we can get a 4dB boost with Q ranging from about 0.5 to about 1, which allows for both a nice gentle curve and a harder, more surgical peak. With the 100k boost pot, on the other hand, the narrowest 4dB boost that can be produced has a Q of <0.5, and it gets wider from there - fine if your goal for the mid band is cuts and/or very wide boosts, but not as useful if you'd like to get a nice peaky 1.4-1.6k boost for guitars from it.
The second reason is related to the first in that it is also due to the narrower curves possible with the 47k boost pots. Using the 60Hz inductive low boost with a 47k boost pot, and combining it with the 60Hz shelf cut from the Pultec, I was able to achieve curves very similar to those produced by the standard Pultec low boost/cut (with the exception of the roll-off below the 60Hz point due to the bell vs. shelf boost, which IMHO is actually an improvement as it should solve the problem Pultecs can have of the low boost sometimes getting too muddy). This combination of the Helios's inductive low boost with the Pultec's ability to dip the low-mids above the boost would be incredible useful. Unfortunately, when we go to the 100k boost pots the resulting 60Hz boost becomes too wide and we can't get those sort of curves any more.
And finally, the 47k pots give us a maximum boost of about 13dB which is still plenty for just about any conceivable purpose (APIs are limited to 12dB and I don't think anyone has ever complained about that being limiting.)
That said, I don't know what sort of other issues might present themselves by keeping the boost pots at 47k (what sort of effect that would have on input impedance, for example - calculating i/o impedance is a weak area of mine.) And I suspect that the choice of pot values would change anyway depending on which and how many bands the individual chooses to build (the idea of three 100k pots in parallel producing 33k goes away if someone only wants to build two of those bands). So it probably isn't important in terms of developing the Universal EQ, but I just thought I'd mention it in case anyone found it useful. Actually, it would probably be a good idea at some point, once the design has been finalized, to put together a small spreadsheet showing the relationship between which bands are to be built, the pot values used for those bands, and the resulting impedances and insertion losses.
Anyway, those are my preliminary findings; I hope they're of some use. Later I'll take a look at the cuts and see what sort of effect they have.
