Errata to Single Super Stage (7-fix)
Was just re-dusting some old texts, and found this small issue. So here comes an Errata.
Things got a bit confused in that Super Stage of mine, a way time back. The more I pondered about the Positive vs. Negative Feedback phase issue, the more confused I was. And there was a good reason for that confusion. The culprits name is CTRL-C, CTRL-V, or copy-paste.
The reason for this errata is that due to the fact that the two CASCODES, sitting on the anodes of the second pair of tubes, are also based on Mosfet transistors, … albeit DIFFERENT ones.
Although we use DEPLETED mode N-Type Mosfets (DN2540) within the cascaded Constant Current Sources (CCS) way up above, the actual CASCODES should, logically thinking, be of the ENHANCED mode N-Type Mosfets (such as the “normal” stuff).
Please note that there is a slight difference in the actual graphical representations of those two kinds of mosfets. the vertical line inside of the mosfet symbol is “continuous” for the depletion mode device, but “dashed” within the enhanced mode device (the case pin is just a side effect of the symbol library lacking a 3-terminal device).
The purpose of the cascode is to hold a steady “voltage” on the top of the anode, as defined by the voltage divider and the preprogrammed voltage upon its gate. The Cascode serves as a “current forwarder”. But in order to do that, it obviously needs to INCREASE current in sync with an INCREASE of Gate-Source Voltage (think of a typical NPN transistor, when used as a cascode, per simple analogy). For a cascode to work properly, a higher positive gate to source voltage differential should actually result in bigger conduction of current. This is the case with N-Type enhanced mode mosfets. You have to polarize the gate more positively so as to “open the mosfet up”. The enhanced mode N-Mosfet is the most “natural” type of semiconductor that fits into this role. Or maybe even a bipolar junction transistor, something from the HV series, like the MJE… series, or similar.
Using the DN2540 as a cascode was a controversial and unintended idea. Actually, it was not even an idea, but a misfortunate side effect of some heavy cut-copy-paste activity.
Now, when we analyse all the inversions of the signal phase, we come to the conclusion that the grids of the second set of tube should not be criss-crossed, but normally fed from the anodes of the first stage. So, here is the modified / fixed version (hopefully a better one):
If you prefer a straightforward HV BJT NPN transistor, with a well defined base-emitter voltage drop, in the whereabouts of 0,7V – no problem, be my guest:
Please also note that after these changes, the outputs became non-inverting.
The Output+ is on the same side as the Input+, and vice versa, the
The Output- is on the same side as the Input-.