|
|
This gain provides feeds feedback that wraps around the MOSFET's already low output impedance and reduces it milliohms. Additionally, by purposely limiting the Op-Amp's power supply voltage, we build in a current limit for the pass device MOSFET of sorts. Adding a handful of diodes makes the regulator much more robust.
One liability, however, of the floating regulator, like all high bandwidth IC circuits, is that it requires careful attention to its physical layout, particularly if tweak, high slew, high bandwidth ICs are used. (I have seen perfectly reliable high-voltage regulators fail or break into oscillation when the slow LM741s and electrolytic capacitors were replaced with ultra-fast wide-bandwidth ICs and polypropylene capacitors.) It is common for the negative power supply pin of an Op-Amp to have a much poorer PSRR than its positive power supply pin. Consequently, a small high quality bypass capacitor should be attached right at the negative power supply pin.
In sum, I am puzzled by the results from your experiments as they are at odds with those I and other experimenters have undertaken. Possibly the values of the floating regulator from the GlassWare article are too small in value. Possibly your implementation was mis-wired. Something went wrong: BB guns do not outshoot 357s.
I do repeat the recommendation that you try the choke based pi filter by itself. Because your line stage draws a constant current, it makes very little demands on the power supply. Thus low noise is primary requirement. It might be a very ear-opening experiment. And I plan on re-measuring the floating regulators I have built.
Subject: current regulated heater CCT
I have built a line stage using a 12BH7 as a plate loaded voltage amplifier, which has an all tube shunt regulated HT supply that operates without feedback.
I have been using DC voltage regulated heater supply and have been entertaining the idea of
|
|
|
|
using a current regulated supply instead.
I have plans to use the circuit from your JULY 2000 archive on current regulated heaters. I like the simplicity here and actually require the same 6.3V @ 900 mA that you have set for the example.
I would very much like to ask you the following questions for clarification:
1) Within what range of values should the cap after the LM317 be?
2) are schottky diodes ok to use in this circuit?
3) will the supply require any additional filtering, or will rectified 12V winding with voltage dropping resistor to 10V be fine?
Shane C.
AUSTRALIA
The LM317 will work without any capacitor, but I recommend a value of about 1 to 10KµF across the heater element, if only to further decrease the inrush current into the heater at turn-on. The LM317 will oscillate within a certain range of values, but anything over 100µF is certainly safe.
Yes, Schottky diodes are an excellent choice because of the lower voltage drop across these rectifiers and their high speed.
I would worry about giving the regulator enough voltage to play with, as there is a dropout voltage that should always be exceeded. So the 6.3 volts of the heater must be added to the 1.25 volts of the LM317's internal reference and then added to the about 2 volts dropout voltage and then further added to a line voltage brownout safety margin, say 2 volts. The total is close to the full 12 volts.
Subject: Negative high voltage regulator
I am currently trying to build my first own tube amp. I still haven't decided yet if it would be the Leak TL/10 or a 807 PP amp I found in the Fi Primer. I have no questions about the circuits as I plan to build them exactly as they are. What I would like is to have a regulated power supply for my amplifier.
|
|
