For my EE friends...

Sep 09, 2006 22:48

Damn I am sick of this already... but here is what I have come up with... and you'll probably find it interesting if you like this sort of thing ( Read more... )

Leave a comment

Comments 15

(The comment has been removed)

loriana September 10 2006, 03:29:21 UTC
Mmmhmmm.. schematics...
Doesnt it just get you all hot and bothered?

Reply

caleb_badger September 10 2006, 15:10:24 UTC

... )

Reply

fuzzytoedcollie September 10 2006, 21:50:17 UTC
Totally tubular, dude....

Reply


caleb_badger September 10 2006, 05:41:23 UTC
I just found a diagram in a book that described a similar situation. I only vaguely comprehend the details of the situation but the concept I know few things about.

Basically it describes using a resistor in series with D1 in the last diagram to speed up the turn off of the inductive current of the load. Probably makes more sense to you than me.

I'm also tired and just throwing out ideas. I'll keep looking tomorrow morning.=P

Reply


rcking September 10 2006, 11:46:39 UTC
Well, I'm glad you found something that works - now I wonder how your customer feels about adding that monster Cap. Does this blow the budget or make the unit impractical?

As for me - I just spent Saturday night writing Cell phone message protocol tests. While my wife and kids are off at an Amtgard event having fun. So I understand about life passing us by.

Reply

loriana September 12 2006, 09:51:12 UTC
Fortnuately the cap should cost about $0.40 or so in quantity. The space is the problem.. as in I have none. I dont know how this will play out in the end.

Test writing..... fun...... and yeah, considering what your workload has been like, you know what i am saying here.

*HUGS*

Reply


fuzzytoedcollie September 10 2006, 21:59:10 UTC
Of course, initially closing the switch (placing 12V directly across a discharged capacitor of several 1000uF) might create a turn-on current surge that overwhelms the system unless the turn-on rate of Q1 is controlled, which then stresses Q1.

And a low-ESR capacitor that can handle 30A probably isn't cheap, although they may be plentiful, being used in switching power supplies as they are.

Nice challenge you've got going there, Loriana! Be curious to hear what finally works out.

Reply

loriana September 12 2006, 09:55:52 UTC
The mosfet is rugged, and can handle turn on stresses. The driver transistor never gets close to its rated current either, fortunately.

The cap doesnt have an esr rated that high... its on the order of about one tenth of that. With higher capacitance than what the simulation would suggest is necessary. And it is my understanding that these two issues are related. Lower esr than necessesary translating into a need for a larger capacitance, by in fact about the same ratio.

Anyway, its not a small cap, but not huge. still even a medium size cap doesnt have a home here... there isnt any board space left. so I am kinda hosed :P

Reply


shockwave77598 September 10 2006, 23:05:50 UTC
1, provide additional capacitance to the drive control system so they fail last.

2, capacitors in parallel with inductors form an oscillator.

3, IGBTs are good for this kind of thing. Vbr > 600V is easily obtained.

4, Flyback diodes are your friends.

5, I've heard that you can put a neon lamp (v=90V) between the transistor and ground, so anything spiking above 90 gets shorted to ground through the bulb. Never tried it myself.

Reply

loriana September 12 2006, 10:05:20 UTC
You know, I have been very lucky with regard to the control system side of this. The 470uF or so of capacitance, fed through series diodes, and the regulator, and the good bypassing of the microcontroller seems to have saved it. I have NEVER lost a micro due to voltage transients on these - on the test bench and in the field. *knock on wood ( ... )

Reply


Leave a comment

Up