Simplistic MosFET HV Shunt Regs

[vgeorge]

Not right at the input, say 15cm. First killed the oscillation with a 56ohm resistor and then tried a 47uF capacitor I had in hand. It was closer to pick up! Everything was with alligator clips and cables not twisted or anything.
It would be better at the input, but as I sorted it out, I did not investigate more.

 

[Salas]

How far is the original rectification + filtering? A resistor is a good inexpensive means also if there is much loop inductance meeting the input. But you will have to see if it needs anything at all in the final installation.

 

[merlin el mago]

Power transformer 2x220V, 5R4GY rectifier, choke input ll1673, ll1660PP for output transformer and AVC from Intact audio.
Also got some nice BIG 10ohm 100W Dale resistor for filament bias.

Where did you bought the resistors?

 

[vgeorge]

leads were 1m. Final layout will have caps near the regs.
Phillipe, I bought these from ebay some time ago. If you want will search and tell you tomorrow.

 

[king4joy]

Salas, and vgeorge....I will have long cables too...between the SSHV2 (in the power supply box) and the 2 CCS's (in the amp box) loading the VT25....where and how should I do the decoupling please??

 

[vgeorge]

Just put the last cap close to the input of the regulator.
As I understand, you have the regulator in another box from the pre, which is not optimal, but given that you use 4 wires output, it won' t matter much.

 

[king4joy]

Yes, I need to put the SSHV2 (only one to start) in the PSU box....as there is not much space left in the amp box....I am using a pair of Rod Coleman's board for the filament...

 

[vgeorge]

Are you going with one or two SSHV2?
One could fit easily, two if vertical maybe?

 

[king4joy]

I am going with only one SSHV2 for now to try....your boards seem quite small...are you using 10uF for C2??....and are you using heat sink on Q1 and Q6 as well...does Q6 need a sink as well??

 

[vgeorge]

Heatsink on Q1 and Q3 only. Yes, 10uF Solen 250V.

 

[king4joy]

I will try and see how big heat sinks I need first...if heat sinks not too big, I can fit it in the amp box....but then I need to put the last cap in the amp box as well.....

 

[vgeorge]

If you do not drop too much voltage a small sink for Q3 will do.
I think that a 0.1uf cap at the input will do also. Try and see.

 

[king4joy]

I think I just need to frop the minium for the board...say 20V...a 0.1uF 400V is not big....thankyou....you think it is not good to put the board in the PSU box??

 

[king4joy]

frop should read drop.....sorry

 

[vgeorge]

It' s always better to have the regulator close to the load.

 

[king4joy]

What are the chances that Q4 and Q5 will need heat sink?? if they will not need heat sinks at all...I can put them closer...and make the board smaller...I plan to have Vout = 270-290V max.....

 

[Salas]

There are good chances they will not need mini sinks for under 300V if the box ain't much hot inside.

 

[AndrewT]

What are the dissipations of Q4 & Q5?
Use that information to estimate the Tj from the datasheet temperature de-rating.

 

[Salas]

2mA @ ~150V each for 300V out.

 

[AndrewT]

King,
look at Salas posted pic.
2mA & 150Vce (or Vds) gives a dissipation of 300mW.
Find 300mW on the pic.
Draw a horizontal line at the 300mW height.
Where does the 300mW line cross the lower of the two plots? I'd guess about 115degC to 120degC.
The plot is telling you that at Ta ~115degC, the Tj = 150degC.
Now reduce Ta by 20C degrees to 95degC.
What is the new Tj? (answer, Tj=130degC)
Now reduce the Ta to what you estimate is the internal ambient temperature of the air around the regulator.
What is the estimated Tj?
If Tj < 100degC the device should survive long term.
If Tj < 70degC the device should never deteriorate in your lifetime.