Simplistic MosFET HV Shunt Regs

[Salas]

Wow, thanks! Here is the PXTA92 which is a SOT89, probably similar performance?

http://www.nxp.com/documents/spice_model/PXTA92.prm

Many, many thanks!

Just a little smoother phase than the other one

 

[michip]

Q4 Ohmic sounds low. Make sure about Q4, Q5 orientation, because plastic backed, spot their front by the lettering

Hi Salas,
the printings on the case are facing towards C1 so assembly is correct. I have re-checked Q4 and found 1088 ohm.

 

[Salas]

Do Q4 & Q5 measure Vbe around 0.6V when in powered circuit?

 

[luvdunhill]

The 2sk880gr recommended in this thread looks good for my purposes. It has 100mW Pd figure, I assume that's okay? Is it 24mW here, based on the 2mA current from your previous Pd calculation you have above?

 

[Salas]

Yes, its alright

 

[michip]

Do Q4 & Q5 measure Vbe around 0.6V when in powered circuit?

Q5: 0,50v
Q4: 0,54v

 

[Salas]

They work but not much on. Try halving R9 & R10 by parallel piggy back ones. Maybe you just work at the bottom end voltage range of adjust now and the circuit is saturated.

 

[Salas]

P.S. You mentioned breadboard. You mean running a pre-test on a development board with flying wires? It can also be oscillation if in such a layout state. When the LED is rather dim its a tell tale sign.

 

[michip]

I mean that it is not the final assembly.
Just for pre-testing before putting everything on the final chassis. So it is a piece of wood where I put the power transformer, CRC filter then SSHV kit (as it is purchased from the GB). The wiring between parts is quite short.

 

[Salas]

So you feed it with how high a rectified DC?

 

[michip]

470uf - 1200+470ohm - 470uf

 

[Salas]

If that RC filter keeps with the CCS current draw for 120VDCin at the regulator's input connector and the output sense+force are properly wired the dummy load, then some semi must be faulty although not looking shorted or discontinued up to now. The relatively small output voltage you can now reach, is it almost equally divided across Q4 and Q5 (collector to emitter)?

 

[michip]

Vout: 76,5v
Q4: 34,8v
Q5: 39,8v
I use 4 wire mode to the load.
Force and sense wires have same lenght.

 

[Salas]

So it shares properly at the cascode. Remove Q6 make sure it does measure normal GR range idss.

 

[Salas]

But use 10k load first in case there is something wrong with the ccs/consumption ratio in the rig

 

[michip]

Hi Salas,

sorry for this delay.
I have removed Q6 and replaced it with a new one previously measured as 4,5mA idss. Issue is still there.
I am not able to increase output voltage over 78V.

 

[Salas]

See what happens if you will short the trimmer. Something is not letting it push enough current to the 68K ref resistors. Also confirm their 68k value.

 

[michip]

I have shorted the trimmer(R11).
Nothing has changed. Output around 79V.
Checked R9and R10 in circuit.
Value is ok (68k).

 

[Salas]

I can't make it out logically after all those positive tests. Something semi must be faking it. Do you have a couple BJTs to exchange, or can find MJE350s compatibles locally?

 

[michip]

At the time being I have no spare.
I will try to find locally but not sure to succeed.
Otherwise there is internet but you are speaking of faking semi. Do you have any suggestion where to get the right parts?