Author Topic: High Voltage constant load design  (Read 336 times)

0 Members and 1 Guest are viewing this topic.

Offline 001

  • Super Contributor
  • ***
  • Posts: 1168
  • Country: aq
High Voltage constant load design
« on: October 10, 2019, 04:07:04 pm »
Hi

I`m gonna buid constant load to test homebrewered tube power supplies 0-600V 0.6A

Nothing special  exept voltage
so I  plane to use 807 tubes instead mosfets in nowday common schematic like this


But problem is that tube bias voltage is about -40volts
It is greater than opamp power rails
What the way?  :-//

Thanx
 

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1474
  • Country: de
Re: High Voltage constant load design
« Reply #1 on: October 10, 2019, 04:41:17 pm »
Paralleling some high voltage MOSFETs is going to be easier. 2SK1317 have a SOA of approx 150 mA @ 600 V, for example, so ~7 devices in parallel should be sufficiently robust for a .6 A load.

Also, narrowing the input voltage and current range allows you to relax pass element stresses a lot by burning power in fixed resistors instead.
,
 

Offline magic

  • Super Contributor
  • ***
  • Posts: 1616
  • Country: pl
Re: High Voltage constant load design
« Reply #2 on: October 10, 2019, 05:07:30 pm »
If you can accept that the input voltage must be 40V or more...

Use any small (not so small at 24W, actually) 40V transistor to regulate current and cascode it with the tube, tying the grid to ground.

edit
For voltages less than 40V the tube could be bypassed.
« Last Edit: October 10, 2019, 05:15:59 pm by magic »
 

Offline aheid

  • Regular Contributor
  • *
  • Posts: 146
  • Country: no
Re: High Voltage constant load design
« Reply #3 on: October 12, 2019, 12:52:55 pm »
I'm by no means an expert, so this might be silly.

You're asking the tubes to dissipate 360W, Wikipedia says a single tube can do 25W max, so that's gonna be a lot of tubes.

At 600V, the equivalent resistance of 600mA is 1k Ohm, which at 360W should be doable as a regular resistive load. Then if you hook your dynamic load in parallel to this 1k resistor, the pass element only has to do <100W according to my sim. Maybe use a bit higher resistance so you have some headroom, 1.1k would work up to 650V and keep pass element still <100W.

The 2SK1317 mentioned can barely do it within the specified SOA, though if I made this circuit I guess I'd put two in parallel to be on the safe side. It's also specified as a switching transistor, not sure how reliable the DC SOA is.

Then again, what do I know, maybe a dumb idea?

In any case, do you really need it to go all the way down to zero?
 
The following users thanked this post: 001

Offline dom0

  • Super Contributor
  • ***
  • Posts: 1474
  • Country: de
Re: High Voltage constant load design
« Reply #4 on: October 12, 2019, 04:11:06 pm »
MOSFET DC SOA generally doesn't account for or underestimates Spirito effect, so they should always be taken with a grain of caution. However, for a reliable design in practice you need 100 % or more safety margin on top of SOA anyway. That's why I recommended at least seven of these MOSFETs in parallel to handle this load earlier. Of course, if you use dumping resistor(s), the load imposed on the pass element is greatly reduced, so you can do with fewer FETs.
,
 
The following users thanked this post: 001


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf