On 17 Sep 2014 13:36:52 -0700, you wrote:
Just some thoughts. The Match pots have most effect at the high currents and nearly no effect at the low currents. As I understand it, originally more readouts were shown than there were active channels selected. Was that perhaps due to a shifted 0 mA current, and missing a SKIP that way? After fiddling around with socketing the selected channels and the 0 became properly displayed. The +15V# and +15V* at the decoders are supplied, it seems, from the +15V at the data switch side (main interface board) via the braid of Peltola cables. These voltages at docoders pin 7 seem to serve as zero reference for the input voltage at pin 10. What would happen with the column decoder when pin 7 (+15V*) receives higher voltage than the supply voltage at pin 8 (+15V)?I do not think this is a problem and I have an idea as to how the decoders work
in this circuit.
Pin 7 and either pin 1 or 16 are at either end of the ADC resistive divider
chain for what is a simple 1-of-10 flash converter. Notice that 15 volts at pin
7 *also* connects to the top of the shunt resistors R3420 or R3409. The current
drawn through R3420 or R3409 which are about 5 kohms produces a voltage at pin
10 which is referenced to pin 7. For a current of 0 to 1 milliamp and a pin 7
voltage of +15 volts, that voltage will be +15 to +10 volts with increasing
currents lowering the voltage. (*)
So the ADC input is referenced to the positive supply or positive reference in
this case. In most ADCs, the signal is referenced to ground which is either the
midpoint of the scale or the negative reference voltage but in this case, ground
is the positive reference. The reason for doing it this way is that like an
LM301/LM301A operational amplifier, an all NPN differential input stage which is
typical of a comparator may have an input common mode range equal to the
The changes in the 7834 readout circuit from the earlier ones with the separate
connection to pin 7 will yield a more accurate +15 volt reference voltage. The
+15 volt sense line from the regulator goes to the main interface board so it is
tightly regulated at that point and provided to the readout board via the
coaxial shields carrying the readout signal currents. I assume Tektronix did
this to improve the accuracy of decoding. It is worth noting that the +15
supply used as a reference and the -50 volt supply used to generate the readout
signal currents through fixed resistors track so absolute accuracy of the supply
voltages is not required.
If there was a problem with the +15 volt reference connection being higher than
the +15 volt supply connection, then I would expect it to only affect the
decoding for 0.0 or 0.1 milliamps. Another possible problem could be if it is
more than 0.5 volts higher than the supply connection and there is a diode
between then internally but flash converters often lack that diode.
(*) This points to either Q3264 or Q3406 being the problem because Q3419 can
only increase the current.