Hi! I have an HT power supply made by me in operation..on a 1000-hole base ... in my prehono with light bulbs (valves). 🙂
The scheme is not my work..but it works quite well.
I still have a little noise towards the low end of the spectrum..with a small peak at 100Hz ... a little at 200Hz ... and some varied noise below 100Hz ... 50Hz zone.
But taking a look online ... on similar schemes ... I notice that there are different capacities than mine.
I refer specifically to:
- In mine I have a mkp capacity of only 0.1uF at the output of the Mj13009 (I also see 22uF mkp on the net ... of which I attach the diagram) ... which is then also being at the input of the LM317.
On the net I see schemes with very different capacities. The LM317 datasheet speaks for the input ... up to 1uF.
- In mine I have a capacity of only 3.3uF electrolytic .. output from the LM317. The datasheet speaks up to 25uF ... I also see 33uF on the net.
Maybe with higher capacities there is to add a protection diode (booooooo.. I don'tknow).
- In mine I have a capacity between the pin adj and the ground from 0.47uF when on the datasheet of the LM ... it says up to 10uF ... for the best results.
It seems that my scheme..and therefore power supply is with the values reduced to a minimum. They made economy!
Since I am making a friend (a little too expensive $$$) unravel the scheme because I would like to make it decently on pcb..and definitively ...
I would be asking you if perhaps it would be better not to optimize these skills in such a way that made the pcb.
I don't think about it anymore.
But, in fact, I have to decide (best capacity) before doing the pcb ... breast 'I'm screwed!
The data ... compared to the capacities ... that I read on the datasheet of the LM but I am not able to understand them well .. I don't know how important it is that in any case my work at 250 and passes volts.
I would not, loading too much Mj13009 and Lm ...., fry them .. or make them work uncomfortable.
I have several electronic materials available and I could draw from them what I need.
Specifically, I also have beautiful vishay mkp caps from 22uF 400vdc ... which could be perfect for Mj13009 and Lm output ... but I ask you.
I also seem to have 10uF mkp 400vdc..I have a vague memory.
The preamp absorbs about 7 / 10mA maximum..I think.
From the Mj13009 I go out with about 275/7 vdc and I go to power the output stage V3 which is an ecc81 but in the future I could opt for ecc82.
From the LM317hv..out with 255/8 vdc and go to power stage V1 and V2 (both ecc83).
From the leveling I arrive to the stabilization with about 300vdc.
For the filaments, which I run at 12.6vdc ... as I have found better results in terms of noise compared to 6.3vdc, I have 2 pcb power supplies based on LT1083.
One for the filaments of V1 and V2 ... the other for the filaments of V3 (of which the relative divider is missing on the diagram ... but on the final pcb there will be ....)
Capacitors added in blue on the diagram, "around" the zener ... currently I have not mounted them but I am going to add them to the final pcb.
I copied these caps from internet schemes ... where they are defined as "equalization" ... perhaps more than equalization which maybe was a term error ... they are used to filter-clean any noises .. rf .. etc...
Thanks if you help me understand if I can do better than the diagram below. I am referring to the capacities intended as type and value.
Here is my scheme .. ugly..made by hand..but you understand:
PS: the R value of the divider for the filaments are incorrect. Those written were for when the filaments worked at 6.3vdc.
Datasheet 317hv: https://www.ti.com/lit/ds/symlink/lm317hv.pdf?ts=1588375219843
.. this below instead is a scheme taken from the internet that I know is an excellent scheme! If I'm not mistaken, it was a scheme made for a pre-amplifier like mine.
I report it only because I see that it has 22uf mkp output for the HT..and for coincidence .. those big and expensive capacitors I would have them already in my drawer...but I don't know if to use them.
They are just a little bulky .. but I would gladly use them if they are needed ... if they can filter better.
I'm sorry if I wrote too much. I also apologize for my bad English.
The scheme is not my work..but it works quite well.
I still have a little noise towards the low end of the spectrum..with a small peak at 100Hz ... a little at 200Hz ... and some varied noise below 100Hz ... 50Hz zone.
But taking a look online ... on similar schemes ... I notice that there are different capacities than mine.
I refer specifically to:
- In mine I have a mkp capacity of only 0.1uF at the output of the Mj13009 (I also see 22uF mkp on the net ... of which I attach the diagram) ... which is then also being at the input of the LM317.
On the net I see schemes with very different capacities. The LM317 datasheet speaks for the input ... up to 1uF.
- In mine I have a capacity of only 3.3uF electrolytic .. output from the LM317. The datasheet speaks up to 25uF ... I also see 33uF on the net.
Maybe with higher capacities there is to add a protection diode (booooooo.. I don'tknow).
- In mine I have a capacity between the pin adj and the ground from 0.47uF when on the datasheet of the LM ... it says up to 10uF ... for the best results.
It seems that my scheme..and therefore power supply is with the values reduced to a minimum. They made economy!
Since I am making a friend (a little too expensive $$$) unravel the scheme because I would like to make it decently on pcb..and definitively ...
I would be asking you if perhaps it would be better not to optimize these skills in such a way that made the pcb.
I don't think about it anymore.
But, in fact, I have to decide (best capacity) before doing the pcb ... breast 'I'm screwed!
The data ... compared to the capacities ... that I read on the datasheet of the LM but I am not able to understand them well .. I don't know how important it is that in any case my work at 250 and passes volts.
I would not, loading too much Mj13009 and Lm ...., fry them .. or make them work uncomfortable.
I have several electronic materials available and I could draw from them what I need.
Specifically, I also have beautiful vishay mkp caps from 22uF 400vdc ... which could be perfect for Mj13009 and Lm output ... but I ask you.
I also seem to have 10uF mkp 400vdc..I have a vague memory.
The preamp absorbs about 7 / 10mA maximum..I think.
From the Mj13009 I go out with about 275/7 vdc and I go to power the output stage V3 which is an ecc81 but in the future I could opt for ecc82.
From the LM317hv..out with 255/8 vdc and go to power stage V1 and V2 (both ecc83).
From the leveling I arrive to the stabilization with about 300vdc.
For the filaments, which I run at 12.6vdc ... as I have found better results in terms of noise compared to 6.3vdc, I have 2 pcb power supplies based on LT1083.
One for the filaments of V1 and V2 ... the other for the filaments of V3 (of which the relative divider is missing on the diagram ... but on the final pcb there will be ....)
Capacitors added in blue on the diagram, "around" the zener ... currently I have not mounted them but I am going to add them to the final pcb.
I copied these caps from internet schemes ... where they are defined as "equalization" ... perhaps more than equalization which maybe was a term error ... they are used to filter-clean any noises .. rf .. etc...
Thanks if you help me understand if I can do better than the diagram below. I am referring to the capacities intended as type and value.
Here is my scheme .. ugly..made by hand..but you understand:
PS: the R value of the divider for the filaments are incorrect. Those written were for when the filaments worked at 6.3vdc.
Datasheet 317hv: https://www.ti.com/lit/ds/symlink/lm317hv.pdf?ts=1588375219843
.. this below instead is a scheme taken from the internet that I know is an excellent scheme! If I'm not mistaken, it was a scheme made for a pre-amplifier like mine.
I report it only because I see that it has 22uf mkp output for the HT..and for coincidence .. those big and expensive capacitors I would have them already in my drawer...but I don't know if to use them.
They are just a little bulky .. but I would gladly use them if they are needed ... if they can filter better.
I'm sorry if I wrote too much. I also apologize for my bad English.
Last edited: