I'm actually geographically far from where I have my amps and all the needed to tests what I'm writing, but I would like to share some ideas about how to implement this concept in other areas, and get some more inspiration as well.
The easiest implementation would be in Engl-like amps, where the stages are mostly warm biased, and the low cutoff os done through cutting abruptly the lows (coupling caps are in the hundreds-of-pF range), but I personally don't like the sound of high gainers with all warm stages.
The choice has been the CAE 3+ because it's simple, has a partial bypass on the cold stage, has snubber on all stages (that is what the LIO should help to eliminate or at least reduce) and needs voltage dividers before 3rd and 4th stages, that can come handy.
Its schematic is in attachment.
I would avoid to use the LIO between the first and the second stage because it's where you need to shape the fequencies of the OD to make it cut better through the mix (at any level of gain), while most of the time after that it's just a matter of thaming highs to avoid unpleasant IMD and harshness, improving the definition of the overdriven sound, but I have to say that I never like the feed of the stages with snubbers, because they are always a compromise between low gain settings (where you don't need them so much) and really highest gain settings (when they are more needed indeed).
I see three options mainly:
- use split load stages instead of dividers to drive the following stage;
- use a source followers (IRF820?) to drive the following stage;
- use a source follower with increased dynamics by using the raw bias (on an head) or a negative voltage obtained by the main transformer (for the preamp) as ground to give some more useable swing.
Of course there's also the option to combine the first option with one of the other two, that is probably the best choice.
So the simplest option is of course to split the 100k on 3rd stage into two 47k, to get the same attenuation of the 470k 470k voltage divider, with a lower impedance. The partial bypass on the cathode helps to keep the impedance lower at high frequencies.
A more complex option would be to use a split load (to keep the source follower in its linear amplification region) to then an AC coupled source follower (IMHO a DC coupled one would have its working point too much moved towards B+) with CCS to drive the 4th stage.
Then apply the LIO approach to warm stages as shown in the link at the beginning of the thread:
What do you guys think?
