|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
73.182.164.162
In Reply to: RE: I need to calculate the value of a voltage dropping resistor posted by DAK on October 12, 2016 at 01:18:52
The better way to do this is to use a 250V (2x 125V in series) zener from the screen grid to ground with a resistor from B+ to the grid. That gets you a basic shunt regulator.
At your operating point, the screen grid draws from 0 to 7mA, so size the resistor to run enough current through the zeners such that they keep regulating even at maximum draw - maybe 10mA - which means a ~5K (1W min) resistor.
To make it even better (lower noise and impedance), add a ~100u/300V cap to bypass the zeners.
A couple of glow tubes in series will work instead of the zeners, but they are noisier and have higher impedance and require more standing current.
Follow Ups:
Hi drsx, i have used zeners attached to the B+ to drop the voltage to the grids and they seem to work ok. Although i believe there is a sonic imprint to using it hence my search for different methods for a G2 supply. But i have not tried the method you describe. Others always mention that high voltage zeners are notoriously noisy so i did not pursue trying any zener more than 21v.
Have you implemented your method in a single ended amp and how does it compare to the dropping resistor, sound wise. cheers,
I have used both Zeners and glow tubes for the screens on both single ended and differential amps. It works well. It is not a perfect regulator, but it does not need to be. It only needs to be better than the ~18K series resistor you would use from B+ to drop 50V at 2.7mA.
And indeed, you'll probably want a ~100R screen grid stopper anyway which adds 100R of series resistance, reiterating that a few ohms of impedance in the shunt reg is not terrible.
That said, if I were building an amp to be as good as it can be, then the simple Zener shunt is probably inadequate. In that case, I'd include a pass transistor and feed it from a CCS.
I'll add that the zener noise has never really been a problem in this use, but a cap bypass will eliminate it.
Also, I like the series resistor for use on guitar amps for the added distortion.
I am almost finished building a SE KT120 amp and since it is "Deluxe" amp with expensive transformers and parts, i wanted to also give the G2 circuit an upgrade if possible. I thought about using the VR tubes but for my chassis , the big transformers have taken up half the space. So, i am pretty much forced to go underneath and compact.
The LR8 circuit with the pass transistor? It's great. I have used it myself. You'll want to make a PCB.
If B+ of 300V is supplying the 250V zener string, that's only 1.2 X the zener voltage. A larger margin is needed for the zener to operate properly. X 1.3 or more is the rule of thumb, 330-350V is a more suitable source.
If the zener is set to 10 mA and supplies a grid drawing up to 7 MA, it shares almost even current with the load and will not regulate well at all. The situation looks worse when the grid draw varies with signal. To give adequate regulation, a simple zener regulator should draw 10 X the max. load current, or 70 mA. This regulator works best under a very small, constant load.
http://www.antiquewireless.org/uploads/1/6/1/2/16129770/37-a_solid-state_filter_choke_or_field_coil_replacement.pdf
It is a little more involved than a string of zeners an supposedly very little noise in this circuit. Anyone try this?
This is another example of a Net-borne design that isn't fully vetted. I only had time for a quick scan of the article, but that's all it took to realize the author may not understand this topic sufficiently to rank as an authoritative source. Here's what I'm referring to:
"In most applications, the pass transistor will not require a heat sink. For example, in a circuit with 50 volts differential across the TIP50 and 100 mA load, the pass transistor is only dissipating 5 watts. The TIP50 is rated 40 watts dissipation."
This statement tells me the author doesn't understand semiconductor ratings, and that throws the validity of the entire article into question. The 40W rating of the TIP50 is based on maintaining the collector at a temperature of 25C. That requires a big heat sink. The datasheet further confirms this error with its 25C ambient (no heat sink) rating of 2W dissipation. Hook it up as the author suggests - 5W and no heat sink - and the part will fry.
That aside, I'm not saying the circuit won't work, or that it's not effective for screen regulation (although it's much more complex and costly than I would find appropriate). It's just worth pointing out that articles and projects like this need to be approached with your eyes open. It's not a commercial design, and it may be burdened with design parameters or advice/technical opinions that will cause failure in your specific application.
--------------------------
Buy Chinese. Bury freedom.
And obviously very familiar with tube circuits. So other than the faux pas with the heat sink, somewhat forgivable since he is a tube guy, the circuit should be good for a job like G2 supply voltage?
It seems that having an independent G2 supply would be beneficial to a power tube's performance. Dak
" the circuit should be good for a job like G2 supply voltage? "
Well, the concept seems OK in theory. Whether it will function as intended in practice won't be known unless you build it. If I were attempting to use it, I would first need to study the datasheets for the semiconductors to determine whether the design as presented is fundamentally correct. I would also need to consider the usual "what-ifs," such as a shorted output and other issues, then decide whether to include protection for those things. After that, I'd have to build it, attach it to the power supply I intend to use and search for any undesired behavior such as oscillations and noise. The final electrical test would be monitoring of the output with a scope while the amplifier is in use to be sure the circuit is regulating correctly and doesn't misbehave in some unforeseen way. Once all that was done, I would collect a group of people to listen to the amplifier, preferably with a second amp using a "standard" screen feed for comparison.
I know this isn't what you want to hear, but it's the general process that I find necessary in order to ensure my work is good. FWIW, you'll probably find numerous people around who will tell you that mixing SS with HV tube circuitry is asking for trouble. I share that perspective, especially when the circuit isn't a commercial design and the DIY user isn't a design engineer. Some very fine equipment has used nothing more than a resistor and cap to feed the screens. You should also consider VR tubes for this. In addition to everything else, they glow. :)
--------------------------
Buy Chinese. Bury freedom.
take a look at pages 3-56 to 3-58
.
Have Fun and Enjoy the Music
"Still Working the Problem"
The LR8 with a pass transistor is a great option, though it may be overkill, and the zeners are probably better than you think. This amp uses an LR8 with a pass FET. But yes, a heatsink is necessary.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: