@Osvaldo de Banfield
Thanks!
Do you build with triodes?
This regulator suits only with some lower currents. Say max like 50mA.
It is nothing for heavy tubes.
Here is an update.
I have added protection diodes at the gate of MOSFETs.
Thanks!
Do you build with triodes?
This regulator suits only with some lower currents. Say max like 50mA.
It is nothing for heavy tubes.
Here is an update.
I have added protection diodes at the gate of MOSFETs.
Attachments
I built several regulator with tubes, linear and one of the SMPS kind using only tubes as an exercise:
http://www.diytube.com/phpBB2/viewtopic.php?f=4&t=6408
A professor of the university I graduated as engineer come to home to see it by his own because he couldn't believe me I did that. He went convinced and happy.
http://www.diytube.com/phpBB2/viewtopic.php?f=4&t=6408
A professor of the university I graduated as engineer come to home to see it by his own because he couldn't believe me I did that. He went convinced and happy.
I build nothing 🙁
But I design very many circuits using SPICE.
I have models for many transistors and some tubes. Mostly triodes.
This regulator I use for a hybrid ECC99 + MOSFETs. At 242 Volt.
It is only in my SPICE and I test so the circuits function.
Some circuits I publish. Only those with good performance.
Ok. LTSpice is fine, but keep in mind that simulations are quite different to real world. I used it sometimes ago.
For example THD gives fantasy figures in SPICE.
But in SPICE you can compare parts and see which is best.
If you have accurate models.
This gives a hint about what will function best in real world.
But the result will be different in a real circuit. This is true.
But in SPICE you can compare parts and see which is best.
If you have accurate models.
This gives a hint about what will function best in real world.
But the result will be different in a real circuit. This is true.
For series regulator, the best tubes I used was the 6FM7 in compactron 12 pin socket and her daughter, the 6FD7 in 9 pin. The same tube in different sockets.. also used 6DE7 for smaller current drains. Neon bulbs as reference.
If you like to play around voltage regulators/stabilizer, there is a pair of things you want to know:
1.- Convert the voltage or error amplifier into a cascode, tube, JFET or MOSFET (BJT has the base current limitation and I personally dislike them as voltage amplifiers). Use the grid/gate of the bottom to feed back a sample of the output voltage to close the regulation loop as normal. Then feed forward a sample of input voltage to the top grid/gate. There will be an optimal top input value for what the ripple rejection is almost perfect (infinite), above which ripple has negative polarity repect to the input. I made it with real tubes and this behaviour is predicted as old as 1939, 8 decades ago and still in use. The two bulb regulator depicted use this principle, with some refinents.
2.- In a normal series regulated supply there is a tendency of it to become inductive looking backwards from output terminals into the regulator. As both the series pass and the error amplifier's gain decline with increasing frequency, its output impedance increases as result. Thus it's like an inductor. Thus, there is always an output capacitor value in which the PSU breaks into oscillations stimulated by load currents or the intrinsic noiser elements in the PSU itself. I discovered it in an article from Philips Technical Review about the 40's of the past century.
As you can see, it is an appasionating item per se.
1.- Convert the voltage or error amplifier into a cascode, tube, JFET or MOSFET (BJT has the base current limitation and I personally dislike them as voltage amplifiers). Use the grid/gate of the bottom to feed back a sample of the output voltage to close the regulation loop as normal. Then feed forward a sample of input voltage to the top grid/gate. There will be an optimal top input value for what the ripple rejection is almost perfect (infinite), above which ripple has negative polarity repect to the input. I made it with real tubes and this behaviour is predicted as old as 1939, 8 decades ago and still in use. The two bulb regulator depicted use this principle, with some refinents.
2.- In a normal series regulated supply there is a tendency of it to become inductive looking backwards from output terminals into the regulator. As both the series pass and the error amplifier's gain decline with increasing frequency, its output impedance increases as result. Thus it's like an inductor. Thus, there is always an output capacitor value in which the PSU breaks into oscillations stimulated by load currents or the intrinsic noiser elements in the PSU itself. I discovered it in an article from Philips Technical Review about the 40's of the past century.
As you can see, it is an appasionating item per se.
Attachments
Last edited: