high current JFETs 1

[zappedagain]

Can anyone recommend a manufacturer of high current JFETs (0.5-5 amps)? I've started with the big firms (On, Fairchild) and so far I'm only finding smaller ones (< 100 mA). Does this beast exist?

I have a circuit where a PFET is charging a very large capacitor; the Rds of the PFET keeps the power supply from triggering its overcurrent limit. After it gets charged I'd like a lower Rds than the PFET has. If I can parallel the PFET with a JFET then the JFET would turn on after the initial charge and reduce the overall Rds.

 

[itsmoked]

How about instead something like this?

You kick in the bypass to dump the charge limiter. Set the resistor at your pre-charge current requirement.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[zappedagain]

Ineresting. That's a bit more complicated, but it can work. Thanks.

 

[itsmoked]

Hey Zapped; If your supply voltage is low enough you can ditch R3 and D2. Little simpler.

The PTC is to protect the limiting resistor. If your limiting resistor has got the stones you can ditch it to. Getting much simpler.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[zappedagain]

I like it. My original PFET was acting like a current limiting resistor. After I ran the heatsink calculations I realized I'm much better off with a real resistor and a shunt, just like your circuit. The heatsink for the PFET current limiter was huge!

 

[itsmoked]

Oh yeah, fore-shore.

One of those ceramic caps can probably take a monstrous blast since it is just a brief inrush current, once in a while. Maybe 10X the watt rating. I think many of them list the maximum current allowed that won't "pop" the resistive trace. And you are even protected from rapid power cycling because brief OFF/ONs will find you with the caps still charged.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[zappedagain]

"charging a very large capacitor" - Did I mention 20 Farads?!? With a 10 amp current limit I'll get it charged in about 30 seconds. All this because our laser vendor mentioned that "oh yeah, if you don't follow the proper shutdown sequence your $50K laser might pop". Now I get to retrofit against a mains fault!

Z

 

[itsmoked]

Yikes! What? One plate is the Pacific Ocean?

What kind of laser? Ion? CO₂?

Seems really lame that the laser guy expects you to handle shutdown sequencing. Or are you doing the 'Laser' and the laser guy is just a 'tube' salesman?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[MacGyverS2000]

Not to sidetrack, but what kind of laser? I'm putting together a largish (3'x3') gantry platform for a laser cutter/etcher, but I haven't searched for laser suppliers yet. Is this an RF unit, sealed CO2, flowing CO2?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[logbook]

Of course we are going back to basics here, but charging a capacitor from any resistive device dumps 0.5*C*V*V into the resistor. Sounds like you are charging to 15V which means 2.2kJ in the resistor.

Have you considered using a power inductor to charge the capacitor. This will improve the efficiency tremendously, thereby reducing the waste heat in the charging circuit.

 

[zappedagain]

400 mW RF Laser. No, I'm just supporting laser guy. We are rushing into production, so hopefully he'll improve it later. and I'll remove my band-aid circuit.

100F, 2.7V in 17.1 ml! NessCap ESHSR-0100C0–002R7. These supercaps are amazing. I'll put three together with some protection circuitry for 33F.

5V, so only 413 J in the resistor. With a several second time constant I'd think the power inductor would get much larger than my resistor. Luckily I only charge the cap at power up and then let it trickle-charge after that so I don't have to go through this sequence too often.

Z

 

[VisiGoth]

logbook has a good point. With global warming and all a switching power supply may be a good idea.

Your inductor should not be designed based on the time constant of the charge time but rather on the time constant of your switching time, which is anything you want. You might find an itty bitty transistor, a small inductor and one control IC will do almost everything you need.

jsolar

 

[zappedagain]

Thinking out loud here...

I'm charging a 33F cap with a current limit of 14A. A quick simulation shows that to stay below that current I need an 8H inductor that can handle 14A. Does such a creature exist? Or am I reading VisiGoth's post properly that this is the wrong way to think about this? I need to add a switch.

I'm presently planning to charge with a 0.4ohm resistor, this will take about 40 seconds. If I switch to a PWM controller could I get much faster? I'll have to dust off my switching power supply books.

I'm also attempting to charge this cap to 5V off a power supply that I can crank up to 5.5V. I don't have much headroom because this circuit is an afterthought/band-aid.

Z

 

[itsmoked]

I'd just run the resistor. A measly 5Ws.. Put in a 10W and call it a day. Simple is beautiful.(read reliable) Don't bother with a honk'in 20lb inductor.

Still not sure about this crazy scheme.. Why can you not just yank the power on the laser? I have never seen one you can't with the exception of flowing gas ones that need to cool down with flowing gas briefly.

RF laser? Is a Maser or are you talking about an RF excited gas laser? (I'm presuming.)

How is adding a truck load of caps going to fix this shortcoming?

BTW I believe you must put a resistor across those caps if you hook them in series or you can over stress the one with slightly smaller capacitance.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[VisiGoth]

The resistor is the simplest solution. And you can't get more reliable at only 10W.

But you do not need an inductor based on your charging time constant. You can create an intermediate time constant based on a clock. i.e. 100 kHz switcher current source would be a small low cost solution. But not less than a 10 W resistor.


jsolar

 

[zappedagain]

This is a seed laser followed by an amp laser. The manufacturer is concerned the amp will smoke itself if it is still running when the seed laser starts to shut down. So I need to provide three amps for three seconds and keep the 5V supply above 4.5V while the laser shuts itself down gracefully (3A = 18F * 0.5V / 3 sec). The worst part is most of that 3A is driving the TEC on the laser because they didn't put separate laser and TEC power pins on the connector!

Luckily though, as I've shown management how interesting this little band-aid is getting, they've decided to reopen discussions with the laser manufacturer...

 

[itsmoked]

Cool! er ah Hot! I mean. That explains it.

Thanks.

Sounds like a helluva lot of fun anyway you cut it.

Some people have all the fun...

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com