|
|
|  Theory of Operation Basic tube operation. The tube operates at a nominal quiescent voltage of +150 VDC. The swing range is +300 to 0V. and the idling point is set by the servo feedback op-amps that drive the bottom current source. The tubes are operating pure "Class-A" at 8.6 ma. This operating point is established by the resistor on the bottom current source that is approximately 475 Ohms. The traditional cathode biasing resistor and bypass capacitor have been replace by a triple cascode connected constant current source.
Gain Settings The gain of the three gain stages is determined by the Mu of the tube minus the degeneration of the coupling resistor between the cathodes. When the cathodes are shorted together the tube gain is almost the idealized Mu converted to db! When the unit runs out of range using degeneration, ( it is a somewhat noisy gain setting method at the lower ranges), pads are inserted and gain stages are eliminated to maximize signal to noise ratio and minimize total harmonic distortion.
System Gain Structure Control Because of the humongous gain available, and the response to near DC, the gain switching, stage switching and transformer removal are accomplished via gas filled gold plated fast relays. These relays, along with special timing circuits, enables the unit to delay the change, mute the output, make the change, allow the output to settle, and then remove the mute. This logic and timing is all accomplished with Q46, Q17, Q10 Q29, Q56 and Q23 on sheet two. The left side of the drawing delays the switch action for approximately 10 ms while the right side enables the mute for approximately 500 ms. Because of this muting action, the input selector nor the preset switch should be used after initial setup when employing this unit for live broadcast !
Insert Sends and Returns The signal level appearing at the insert send is nominally -10 when the meter reads "0". This is done because the 2nd stage output is buffered to this point with 2 op-amps. The maximum level at these devices is +26, the maximum level out of the 2nd stage is +42. Therefore, in order to not make the insert send buffer the limiting element of the headroom of the box, the level is lowered at the send, so that +40dB can be accommodated internally. The insert return is nominally +4 (this gain differential must be made up via the insert loop) and goes directly into the gain cell through 2 small isolation resistors, and 2 100k dividing resistors. Any unbalance that occurs will result in signal degradation, so care must be exercised in this regard. The silicon is bypassed when there is no insert connection. This is accomplished via relay K7. The output of this relay is also the point that the side chain signal is developed. The nominal signal for the side chain is also therefor +4.
Level Meter The 10 segment level meter employs a unique combination of 2 separate audio level meter circuits. The output of a VU driver and a Peak driver are hard wire Or'ed to obtain the dual metering function. In order to distinguish between the ballistics of the two metering circuits the visual intensity of each scale is varied according to signal level. The peak scale is dim for the first 7 LED's and then is bright, whereas the Vu scale is Bright for the first 7 LED's and dim for the last 3 segments. This is set so that a wide dynamic range signals can be simultaneously displayed along with subtle 1 dB changes in signal level. The signal is selected at the insert return point and just prior to the mute relay at the output. signal is attenuated by 18 db to obtain the dynamic range necessary for operation. The signal is converted to single ended at the metering detection circuitry. The signal conditioning is performed on the gain cell control PCB, and the DC meter drive is presented to the front panel board independently.
Gain Cell The gain cell employs the same medium Mu twin triode as the gain stages. The tube is used as a transconductance shunt element. What this means in layman's terms is that the conductance of the vacuum tube, (i.e. resistance) is a function of the ratio of the Cathode current to the grid bias. If you will look at sheet 7, and note R12 and R13, these are build out resistors for the tube acting as a shunt resistor. The cathodes are shorted together, so the more the tube conducts, the closer the two plates are to each other the more attenuation through the circuit. The transistors above and below are there to control the bias and no audio flows through them! The plate circuit and the cathode circuit are forced by the gain control line which appears on the right side of R6 to increase the current through the tube, in equal and opposite directions. The two tube sections are part of the same tube, therefore will track one another. The transistors are part of a super-matched array, so they will track one another, and U1A sets the quiescent bias point with no limiting. The zener Z13 clamps the cathodes to +6 v above the grid so that the be doesn't become a diode. Whenever the control line goes above .75V the gain cell clip indicator comes on and shortly thereafter distortion ensues. The gain reduction bypass is a relay that simply cuts the tube off by putting a large negative bias on its grids. We also insert a 2.5db loss via shunt resistor elements to accommodate the tip in characteristic of the tube. The tip in adjustment is necessary because the point that the tube conducts initially is both noisy and abrupt. This point varies greatly from one tube to another, which is why the offset voltage between the output of the gain control system and the point of conduction is adjustable. This point is set so that with no limiting, the tube conducts approximately 2 db. The reason for this is that without it, the transition to limiting is sonically abrupt and thumpy. More than 3 db of limiting doesn't really seem to help and it comes right off the top of the maximum available limiting which is tube bound to approximately 20 db.
Gain Cell Control The audio signal is converted into a DC voltage that is an average of the LOG of the input signal. This is then presented to a logarithmic charge pump, whose charge time is adjustable linearly in time/db. This results in an attack time that tracks the average level of the input signal and is not affected by its crest factor. The parameters of ratio, attack and threshold are set with linear dc voltages. The release time is also determined via a log converted charge pump, which resets the past memory of the average signal level at a fixed recovery time. The objective was to provide transparent gain control without removing the life from the performance.
PRESETS A linear DC voltage controls all of the parameters of the signal dynamics, making remote and preset operation easy to implement. The control elements are voltage dependent, not current, so that they can have some variability in their switch elements. This enables simple FET gates to provide the matrix selections. The manner in which the preset voltages were obtained is a trade secret but for some good PR we will tailor them for YOU; consult your dealer or the factory for details.
Output Stage The same circuitry from the previous gain stages is employed for the output driver. The output stage itself is something else however. The power supply winding that feeds the 60 VDC to the output stage is balanced and floating in relation to ground. The circuit can operate in balanced or unbalanced configuration like an output transformer and allows either polarity of the output to be grounded. An unbalancing switch is provided on the rear panel to select the desired signal output to be grounded via an internal relay. 
| |
|