High voltage current source as anode load...

[HVfanatic]

Hi,

I would like to build a very high voltage current source as anode load (5kV). Due to the fact, that silicon has max. values or around 1,5kV, I should sprit to serialized devices. An idea would be opto coupling??? Are there any proven schematics around???

The Idea behind is:

ESL DD, an I dislike the SRPP... (Avoid elevated heaters...)

Thanks for your proposals,

Dirk

[HVfanatic]

No one has any idea???

[jcx]

mosfets can be cascaded with MegOhm value gate bias resistor string which won't substanially hurt output Z

adding C to the divider can give better voltage division balance at higher frequencies

if you start with a depletion device you can make the whole circuit "2-terminal" and not require high V feedback isolation

[HVfanatic]

Hi JVX,

could you give me a schematic????

Regs, Dirk

[jcx]

http://img116.imageshack.us/img116/4928/amprm7.gif

output Q18-23 on right are cascaded - driven by voltage dividers that share the output V across the devices


I show a "2-terminal" regulated ccs with a 900 V depletion mosfet model in this thread

CCS transistor choices


I haven't played with multi KV circuits - I understand failure to evenly share the V can give spectacular failures - eye protection recommended

[HVfanatic]

Thanks. But for me it looks like that the string resistances sum up, so that the more silicon in the string, the worse the gate or base drive... Seems a lot inferior compared to an penthode design...

Thanks, Dirk

[jcx]

yes the devices in the middle see high drive impedance - why I mentioned adding a parallel C divider string to better drive the mosfet gates for better dynamic voltage division

with the regulated ccs reference at the bottom you can terminate the bias string to the regulator input/ depletion mode source and the regulator will compensate for divider current - allowing for lower impedance and better dynamic drive

this technique will probably still have a practical slew rate limit determined by device mismatch and dynamic gate drive requirements

but it can be good enough for audio - Koss uses cascades in its electrostatic headphone amp with Supertex 500 V mosfets and 1200 V output swing


its a big help to properly size your pass devices - don't try to drive nF of 15A power Mosfet gate C when you just need 100mA of current - look at IXYS devices

[HVfanatic]

Hi,

unfortunately I don't get the point of your explanation. Do you have the Koss schematic? (or alternatively, post a handdrawing???)

Regs, Dirk

[CBS240]

I haven't played with multi KV circuits - I understand failure to evenly share the V can give spectacular failures - eye protection recommended

Eye injury is not the worst thing that can happen to you naively playing with 5Kv.

That kind of voltage might not suck the paint off of your house, but it can give you a permanent orange afro......