|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.127.166.64
If my math is correct, a LM-317 regulator provides a little less than 1mv AC ripple output (80dB rejection). Is there any regulators that can provide 1.5 amps that is quieter?
This PS is for a 26 tube filament & bias.
A single choke input design provides about 3mv ripple. I assume too high AC ripple for the sensitive 26 tube.
Follow Ups:
The LT1764 series is marketed as low noise.
Understand the "noise" in this context is different from ripple. A quick look at the data sheet shows ripple rejection at 65dB, if I read it correctly.
All integrated voltage regulators I've seen are noisy. If you want a quiet regulator, you'll have to construct it yourself. One good topology would be a well-filtered Zener driving a darlington power transistor (or driving a high-gain TO-92 to an emitter follower). Another is a standard opamp with a reference on its non-inverting input and feedback to its inverting input from an output transistor. Either of these will exhibit a noise level that is orders of magnitude less than what's produced in any device employing thermionic emission.
--------------------------
Buy Chinese. Bury freedom.
Here is a design. How much ripple allowed for the input voltage? Would it be better to increase the 100uF cap at transistor base? How mush capacitance on output?Also, most Darlington transistors have HFE of 500 to 1000. One has gain of (HFE) 25000. Is the higher gain better?
I also understand adding a 100 ohm resistor in-between the zener and ground reduces zener noise and output noise at the cost of voltage stability. That is not an issue as the 26 tube filament is constant current draw. However, I could use the forward drop from LEDs for most quiet performance.
Edits: 01/29/17 01/29/17 01/29/17
I would try the hfe=25,000 device first and add a resistor between the Zener and the 100uF cap. Depending on how fast you need it to start, something on the order of 22K-62K would be appropriate for reducing Zener noise. You could also use a TO-92 transistor like the MPS-A18 (gain = 1,000) to drive a standard bipolar. That would provide a very high total gain and allow for the possibility of a second RC filter between the two devices. There are lots of ways to skin this cat. Note that I'm not making any recommendations here about sonics, just how you can get clean DC.
--------------------------
Buy Chinese. Bury freedom.
you could use HI FREQ heating LOL jk look into filiment bias using Rod colman regulator there is no schematic but if you look hard enough you would be able to cobble together something similar
the 26 is a real bugger to get quiet
Lawrence
You can buy a Rod Coleman regulator for $40. How low noise is it? Is this the 'ticket' for 26 tubes?
What would happen, if you blew off all the complexity, and just used a simple second L/C section, as I do in all my B+ filters ? L1/C1/L2/C2.
How might that sound, if it were well implemented,m parts and design-wise ???
Jeff Medwin
Simple LC section using a 60mH choke and 44000uF capacitor results on .00 AC ripple in 9 volt bias. A 8 ohm resistor & 100uF cap at the filament is less than .001 AC ripple.
I do not have a AC ripple spec for a quiet running 26 tube, but .001 ripple is not good enough.
I have NOT used rod's regulators for the 26 tube but I will say when done right the output hum will be Uv level!
Lawrence
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: