Beast with 1000 JFETs redux ?

[vitorio]

Woofertester, thank you for so long and detailed answer. Many things are clearer now.

I have to: classify jfets, check wattage/dissipation, put bigger resistors, change pinouts, make some boards and mount two amps.

I think it worths trying, even it's at my limit. We're confinated now in Spain, boredom is my friend.
Again thank you

[woofertester]

Originally Posted by vitorio

Woofertester, thank you for so long and detailed answer. Many things are clearer now.

I have to: classify jfets, check wattage/dissipation, put bigger resistors, change pinouts, make some boards and mount two amps.

I think it worths trying, even it's at my limit. We're confinated now in Spain, boredom is my friend.
Again thank you

I hope this helps relieve some boredom for you.

[woofertester]

Originally Posted by vitorio

Woofertester, thank you for so long and detailed answer. Many things are clearer now.

I have to: classify jfets, check wattage/dissipation, put bigger resistors, change pinouts, make some boards and mount two amps.

I think it worths trying, even it's at my limit. We're confinated now in Spain, boredom is my friend.
Again thank you

Bigger resistors will not solve a power dissipation problem. The circuit will have to be redesigned to add source resistors for local degeneration. This will lower bias current and may increase output resistance.

For now, just live with the J175 operating its power limit.

[Kris0603]

I wonder if anyone considered putting their beasts in a tank and filling it up with fluorinert and an aquarium pump to circulate, one of those neat waterfalls for cooling just like Cray did.

[Zen Mod]

HAM Geezers did that with their melting 807s and 6L6 ..... bucket of oil ......

why one should the same with eeny tiny weeny JFets?

if you push them in enormous current region , byebye any linearity

[woofertester]

Originally Posted by Zen Mod

if you push them in enormous current region , byebye any linearity

When I have a pcb populated, I will post some scope shots. If linearity is bad, it should be visible on the scope.

[Zen Mod]

enormous current region means - above Idss ....... gate going positive ref. to source

I can't see that your pcbs are made to accommodate that (and they shouldn't be)

[Biosonar]

Below is the basic schematic for a CCS beast cell. R3 is just a gate resistor. Any value from 200 to 1k ohms will fine. Nelson's micro-beast uses 470 ohms for R3.

For good linearity from the J113 jfet, Nelson's bias resistor, R2, is 47 ohms for 10 to 15 milliwatts dissipation per J113. Even at that dissipation, I can't keep my fingers on the jfets for more that 10 seconds. So heat dissipation needs to be taken into account.

Lastly, R1 is 47 ohms. One drawback of the J113 jfet is Idss values range from 7 to 30. So matching J113s is important.

[Biosonar]

The CCS micro-beast schematic:

[wattsman]

Thank you!for the schematic and the information.

Best regards, Frank