Simplistic MosFET HV Shunt Regs

[merlin el mago]

I have received IXYS, but noticed that perform worse current capability.
At same R1 R2 = 100R:
DN2540 can reach 134mA.
IXTP02N50P reach 100mA.
IXTP01N100D far less, 42mA.

Is Ok. 360R resistor value in the transiator gates? I can measure the performance in my oscilloscope, but very limited experience using it, I am afraid I will need assistance. By now the multimeter reading is stable. I wants also implement IXYS ones.

If the voltage is jumping up & down (oscillation), if it's solid as rock surely OK

 

[merlin el mago]

As I said 1.000.000 times the gate resistors value it's ony to avoid oscillation....

 

[Salas]

Hi, it takes lab work to confirm what is happening but upping the gate resistors value indicates that more current setting increases transconductance and it takes more gate damping to stabilize.

In any case 360Ω R1 R2 is permissible. There are examples of 1kΩ even. The downside of using more than needed gate stoppers value is it unnecessarily truncates the cascode's PSRR curve in very high frequencies. But in this case it looks like that higher value is needed indeed.

 

[Salas]

If the voltage is jumping up & down (oscillation), if it's solid as rock surely OK

Oscillation is confirmed if the oscilloscope can pick AC waveform riding the DC line. Sometimes jumps sometimes just brings heat sometimes hum and noise sometimes it does not respond to adjustment.

 

[JOIMONF]

Thanks, just the heat on D2 conceans me, is strong, so may I place a heatsing and done?

 

[Salas]

D2 is a Zener. Q2 you mean? Its Vds drain to source voltage times the mA it passes must mathematically explain the heat. Power=V*I. Else its oscillating. If Vds Q2 is high almost HV then Q1 does not work, its cold, and does not shield it.

 

[Salas]

Keep in mind that when you will put the Mosfets and gate resistors back into place without test pins the behavior will probably be more stable. Because gate resistors and gate pins work best at minimal distance.

 

[JOIMONF]

Yes the heat is in Q2 sorry. I will put some kind of headsink if in need then, thanks Salas!

 

[Salas]

Check if its Vds is few volts as it should. Read #6,586 again.

 

[JOIMONF]

I read it, thanks. I remember having Q1 cold, tomorrow I will measure Q2 Vds accordingly

 

[JOIMONF]

Yesterday I had a solder bridge, on Q1 drain to source pins. My bad at placing the three pin row as a socket.
Since fixed is no more heat on Q2.
Now I am happy to get finally 172mA. @ 77Vdc.
That's only by using DN2540 both on Q1 Q2.
I must note that it is consistent no matter value is on R1 R2 gate stoppers. Same results 100R, 360R and 510R.
If I use Ixys on both, then can't reach the specifications Load. I am going to try different combinations, maybe one only Ixys is still possible.
EDIT:
did'nt mention, R4 is set to maximum strengh= 0ohm

 

[merlin el mago]

Finally used SSHV2 schematic or euro21 mod Jordi?

 

[JOIMONF]

Hola Felipe, no need to mod the schematic, for some reason now performs Ok. within my load specifications by just lowering R4 down to almost 0.
Another rule is mantain DN2540 as Q1, then I even can use IXTP02N50D as Q2 without any drawback.
I tried IXTP01N100D also as Q2 only, but fail to reach my load, by small margin.

 

[merlin el mago]

R4 fixed resistor?

 

[JOIMONF]

R4 could end as a fixed resistor, but not before replace the variac for a properly mains audio transformer. Current on R4 is a bit related to diferent Vin.
I wil replace now the R4 trim by a low ohm value rheostat.

 

[merlin el mago]

Always current is related to Vin Jordi.

 

[JOIMONF]

Good

 

[merlin el mago]

Be care for the rheostat resistor will be properly power rated.

 

[JOIMONF]

I will, thanks Felipe hey, those gate stoppers R1 R1 did'nt deserve to upgrade to Carbon Film? R7also, as well?

 

[merlin el mago]

Good lucky Jordi