Simplest possible tube regulator -- will it work?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I made a few changes in Spice and was able to get a good simulation using a 6080, total output noise was 160nWs, which translates ripple voltage in the micro volt region. This one is set for 300VDC.
 

Attachments

  • E3821871-7E61-4F2A-A6B1-EBD6330E80FA.png
    E3821871-7E61-4F2A-A6B1-EBD6330E80FA.png
    96.3 KB · Views: 273
  • C27F6690-2779-423C-8186-B0CC4BBB2B15.png
    C27F6690-2779-423C-8186-B0CC4BBB2B15.png
    112.6 KB · Views: 270
The noise will be reduced by the suggestion I made - it's a usual technique with the tradeoff of somewhat reduced transient response.

LT3080 is a horse of a completely different color. PSRR at least 20dB better than LM317.

Yes it is a very good regulator, it can be implemented for high voltages
As well, I would love to find a standard Berkeley spice model, that would work
In B2!!
 
Yes, Art of Electronics has a design which works -- at least in LTpice. Tom was selling some HV LT3080 boards.

21st Century Maida Regulator

You can modulate the current adjust pin and get it to amplify.

The easiest way to model any regulator using the LM317 is to do a test of impedance. In LTSpice you can click on the left axis and select phase angle. Phase angle and Q are directly related to phase margin.

If using a bode plotter with the LM317 remember that the voltage reference is tied to the positive rail.
 
Yes, Art of Electronics has a design which works -- at least in LTpice. Tom was selling some HV LT3080 boards.

21st Century Maida Regulator

You can modulate the current adjust pin and get it to amplify.

The easiest way to model any regulator using the LM317 is to do a test of impedance. In LTSpice you can click on the left axis and select phase angle. Phase angle and Q are directly related to phase margin.

If using a bode plotter with the LM317 remember that the voltage reference is tied to the positive rail.

Yes I’ve seen Tom’s designs, implementing a tube pass device wouldn’t be hard.
Analog Devices used to post generic spice models for their products
Most are designed for LTSpice, although they still post models for
Their op amps.
 
I'm sorry to resurrect this thread.



The conceptual circuit works using the 317, i've verified this.



Most of these circuits call for the LM317HV because some tubes require rather large negative grid voltages if you want a widely adjustable supply.



The TL783 might be a good candidate if you want to use low Mu triodes that require large negative grid voltages. The 783 also has better noise rejection then the 317.



The TL783 requires a minimum load of max 15mA to maintain regulation, as opposed to the 317's minimum load requirement of around 3mA. In the 317's case this requirement is mostly met by the divider resistors.


Here's my implementation, ive arrived at virtually the same circuit as Denny.



I wanted to be able to use the boards i made for experimental variable supplies, not just fixed ones. So i added a current sink to meet the TL783's minimum load requirement.



If you have an extra HV winding, you can put the output tubes in pentode mode, in which case the open loop gain will be greatly improved.
 

Attachments

  • Schematic.png
    Schematic.png
    30.4 KB · Views: 197
  • Board.png
    Board.png
    53.4 KB · Views: 187
  • photo_2019-09-12_16-10-04.jpg
    photo_2019-09-12_16-10-04.jpg
    195.1 KB · Views: 189
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.