Honkin' big pass transistors.

I'm especially blond today -- several hundred volts and several tens of amps. I don't see that changing the switcher/linear equation, although it certainly impacts what specific devices one may wish to use!

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html
Reply to
Tim Wescott
Loading thread data ...

ms

ns

s
e

ort

e

d text -

Configuring a bunch power semis to handle it is the easy part- the real challenge is keep die temperature excursions, and thus die expansion, under control to avoid early failures due to that effect. This is a completely separate failure mechanism from heat related stress. You should be discussing heat sink capacity and thermal time constant designs that are larg compared to your duty, if that's possible.

Reply to
Fred Bloggs

As Jim said, it depends on the quantification of the two instances of the word "several" :-)

In the end it's all in the SOA, meaning the milliseconds, duty cycle, amps and volts. Figure 9:

formatting link

That'll tell you how many you need and how well the load needs to be balanced between them. You can also stack them in series but that can get iffy.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

Still got eyebrows?

Reply to
John S

The minimum delay in the feedback loop is cycle_time/number_of_legs, plus some margin for stability.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

formatting link

Reply to
Vladimir Vassilevsky

Well, copper wire, switches, fuses, relays and connectors are all linear pass devices. I get the impression you really want a solid state switch or device with gain, though? Hundreds of amps usually means thyristors, thyratrons, IGBTs. If you really need a linear range of gain, though, as opposed to a switch, this is devices-in-parallel territory.

Reply to
whit3rd

I tend to concur. For linear, I'd use mongo MOSFETs on a MONGO heatsink, maybe even liquid-cooled or something.

Cheers! Rich

Reply to
Rich Grise

1mS settling time isn't much of a challenge for a switcher, but the extreme pulsed duty means using parts sized for the peak. Only the heatsink size is reducible to the average loss.

It amazes me where you get these pulsed poer applications that are so regulation-critical. I bet if you looked closer, you'd find it was just an ignorant gandy-dance.

I'd be tempted to segregate those circuits that are actually supply voltage sensitive, and those that actually need power.

Where local regulation is critical, you might try a compound circuit - one to provide the juice and another to do the actual fine regulation and filtering, at the load.

RL

Reply to
legg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.