You need to do two things: (1) preserve the straight-through path from the microphone to the mixer so that phantom power can flow, (2) isolate the parallel path to the recording interface.

Looking at it this way shows you that the minimum you must have in every splitter channel is a 1:1 transformer to isolate and feed the interface. The transformer's primary is wired across the signal pins (XLR pins 2 and 3). Be sure to connect the transformer can to chassis (mains earth) to avoid hum problems.

You should prepare for a signal level drop of a few dB when routing through a splitter.

If you have a mix of line-level and microphone feeds, you should not use microphone transformers in the line-level inputs. You need to work out what your numbers of each is likely to be and whether to have separate splitter boxes of microphone transformers and line transformers.

Lots of patch cables!

For me, it's always been simple passive splitters. Over the years the most useful box I built had 3 circuit jacks that went into the analogue desk inserts and in went to out, with a tap to go to another mixer. I built the splitter into a rack with more inserts so you could still add your effects with the usual connections, but there was a nice simple split to the other mixer's line inputs. It also solved the issue with splitting condenser mics where the phantom might cause some grief. I've never noticed any quality difference. I suppose I could have bought transformer splitters but the price was a bit steep.

I am new to this forum and anticipate getting (and sharing , where I can) some handy studio advice. I had a modest home studio for the past few years but recently had to move...ugh...  that is a pain ,however, my new digs allow me to set up a new and improved studio, I am attempting to take the lessons learned in my previous experience and integrate them into the new set up from the get go. To be clear, I am not made of money and am a solid EM tech so I can troubleshoot design and build, but have no want to re-invent the wheel , or spend days at a work bench soldering connections (unless I have to), I have located a couple of "affordable" passive splitter units and am reviewing their capabilities and product reviews, they seem to address the points you guys have made (issues I would otherwise have been ignorant of) Thank you so much! I will update my progress on this string, I will provide model and manufacturer info of the unit(s) I have gone with and attempt to provide a down the middle assessment for your consideration. Again Thank you!

With splitters, there are two different things to consider - architecture and implementation - and they should be taken in that order.

The architecture includes, for example, whether the splitter needs to a ensure that a recording keeps running unaltered if the FOH sound engineer makes a simple change (e.g. trim pot adjustment) or a screw-up (e.g. presses mute rather than pre-fade listen on the main vocal channel). When I've been in charge of recording on large gigs, I've had this and much more from so-called professional sound engineers on the desk, and have been very relieved that I chose mic splitting and not recording from insert sends.

Recording from insert sends is easier, as you get mainly line levels (balanced or unbalanced depending on the grade of mixer), and you are never going to be sent any phantom power. The downside is that everybody who ever listens to your recording has to put up with the often less-than-stellar quality of the mixer's pre-amps and the results of the sound engineer's choice of gain staging. With a mic split, you are the one who gets to choose what make of pre-amps are used in the recording for the vocals, the guitar cabs and the drums and so on. That's getting on for being as good as able to choose the microphones as well.

In terms of implementation, the posts above outline the minimum needed for a split from an insert send versus one done at the microphone inputs.

I've talked mainly about live gigs, where you absolutely only have one shot at getting it right, and there are no overdubs possible. By contrast, in the studio, particularly where you are both the recordist and the mixer, you have considerably more flexibility. However, audio quality is even more important, so you should keep that firmly in mind in your architecture decision making.

It wasn't completely clear from your first post how you were thinking of using your analogue board and tape machine. Do you aim to record sub-mixes to go to tape? Does the tape machine have only unbalanced I/O? What make/model is your 20-channel audio interface?