Proprietary SMPS design tools, what for?
in [Design]
|
From: Joerg on 7 Sep 2009 20:34 miso(a)sushi.com wrote: > On Sep 2, 11:06 am, o...(a)uakron.edu wrote: >> LTs field engineering staff is saving my new job. We had a state of >> the art special buck switcher being custom made, and the supplier's >> marketing people dropped the ball and the project. It seems 3000 >> pieces per month is not worth a run to them. Two conference calls and >> a couple weeks later a LTspice model arrives with demo boards . LT >> didn't hesitate to sign our ND, either. I'm too busy doing the custom >> optics and control design to learn switchers from scratch right now. >> >> A bit of wimping out on my part, but I have spent three months just >> trying to source parts for the rest of the system, and the boss wants >> a field worthy trial system in 30 days to replace my huge bench >> model. >> >> Don't even try to manufacture anything in America any more, its a >> tough mess, and I need 60% US content for the contract. >> >> Thank You, LT! >> >> Steve > > I would say never design your own SMPS but use a well flogged chip run > past apps. There are so many gotchas in SMPS design that you will > eventually be burnt. But I've done it, and even current-controlled ones. The highest level of integration used in those designs is usually 1/6th of a hex Schmitt such as the CD40106 or 74HC14 :-) If I feel really generous I might spring for parts of a LM324 in the feedback paths, but usually not. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Jim Thompson on 7 Sep 2009 20:49 On Mon, 07 Sep 2009 17:34:24 -0700, Joerg <invalid(a)invalid.invalid> wrote: >miso(a)sushi.com wrote: >> On Sep 2, 11:06 am, o...(a)uakron.edu wrote: >>> LTs field engineering staff is saving my new job. We had a state of >>> the art special buck switcher being custom made, and the supplier's >>> marketing people dropped the ball and the project. It seems 3000 >>> pieces per month is not worth a run to them. Two conference calls and >>> a couple weeks later a LTspice model arrives with demo boards . LT >>> didn't hesitate to sign our ND, either. I'm too busy doing the custom >>> optics and control design to learn switchers from scratch right now. >>> >>> A bit of wimping out on my part, but I have spent three months just >>> trying to source parts for the rest of the system, and the boss wants >>> a field worthy trial system in 30 days to replace my huge bench >>> model. >>> >>> Don't even try to manufacture anything in America any more, its a >>> tough mess, and I need 60% US content for the contract. >>> >>> Thank You, LT! >>> >>> Steve >> >> I would say never design your own SMPS but use a well flogged chip run >> past apps. There are so many gotchas in SMPS design that you will >> eventually be burnt. > > >But I've done it, and even current-controlled ones. The highest level of >integration used in those designs is usually 1/6th of a hex Schmitt such >as the CD40106 or 74HC14 :-) > >If I feel really generous I might spring for parts of a LM324 in the >feedback paths, but usually not. From '77 to '87 I designed (off-line) SMPS's for GenRad portable devices... never used anything more exotic than combinations of LM324/LM339/TL431 in the loop... take that back, had some 4N35's in some of them. Power devices were Motorola high-voltage bipolar's (MJE... ?) with load line control so good I didn't need heat sinks for 250W outputs... to wit, my infamous, "Wonder how hot this is", grabbing a flag... it was cool thermally, the 400V P-P got me, much to the delight of the technicians ;-) ...Jim Thompson -- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona Voice:(480)460-2350 | | | E-mail Address at Website Fax:(480)460-2142 | Brass Rat | | http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
From: Tim Williams on 7 Sep 2009 22:14 On Sep 7, 7:49 pm, Jim Thompson <To-Email-Use-The-Envelope-I...(a)My-Web- Site.com> wrote: > From '77 to '87 I designed (off-line) SMPS's for GenRad portable > devices... never used anything more exotic than combinations of > LM324/LM339/TL431 in the loop... Why would you go and do it discrete when a TL494 has all the chip stuff onboard already (or UC3842 for flyback)? Yes, it's just as easy to do, but it takes four times more ICs and resistors, and what do you get? Pretty much the same thing. > Power devices were Motorola high-voltage bipolar's > (MJE... ?) with load line control so good I didn't need heat sinks for > 250W outputs... to wit, my infamous, "Wonder how hot this is", > grabbing a flag... it was cool thermally, the 400V P-P got me, much to > the delight of the technicians ;-) Funny, I was touching an MJE18008 the other day that had a good 140kHz squarewave on it, easily 300Vp-p. Didn't feel a thing. Nothing electrical (and nothing thermal either ;-) ). Matter of fact, the transformer's ferrite core, which is nearest to the 2kV supply's secondary (which was making about 3kVp-p at this time) didn't feel like anything at all, though the 0.05" spark makes a nasty smell from your finger's outer layer of dead skin (or worse if you hold it there..). Was yours really so bad that you got shocked by it? How did that happen? ;-) Tim
From: Joerg on 8 Sep 2009 09:46 Jim Thompson wrote: > On Mon, 07 Sep 2009 17:34:24 -0700, Joerg <invalid(a)invalid.invalid> > wrote: > >> miso(a)sushi.com wrote: >>> On Sep 2, 11:06 am, o...(a)uakron.edu wrote: >>>> LTs field engineering staff is saving my new job. We had a state of >>>> the art special buck switcher being custom made, and the supplier's >>>> marketing people dropped the ball and the project. It seems 3000 >>>> pieces per month is not worth a run to them. Two conference calls and >>>> a couple weeks later a LTspice model arrives with demo boards . LT >>>> didn't hesitate to sign our ND, either. I'm too busy doing the custom >>>> optics and control design to learn switchers from scratch right now. >>>> >>>> A bit of wimping out on my part, but I have spent three months just >>>> trying to source parts for the rest of the system, and the boss wants >>>> a field worthy trial system in 30 days to replace my huge bench >>>> model. >>>> >>>> Don't even try to manufacture anything in America any more, its a >>>> tough mess, and I need 60% US content for the contract. >>>> >>>> Thank You, LT! >>>> >>>> Steve >>> I would say never design your own SMPS but use a well flogged chip run >>> past apps. There are so many gotchas in SMPS design that you will >>> eventually be burnt. >> >> But I've done it, and even current-controlled ones. The highest level of >> integration used in those designs is usually 1/6th of a hex Schmitt such >> as the CD40106 or 74HC14 :-) >> >> If I feel really generous I might spring for parts of a LM324 in the >> feedback paths, but usually not. > > From '77 to '87 I designed (off-line) SMPS's for GenRad portable > devices... never used anything more exotic than combinations of > LM324/LM339/TL431 in the loop... take that back, had some 4N35's in > some of them. Power devices were Motorola high-voltage bipolar's > (MJE... ?) with load line control so good I didn't need heat sinks for > 250W outputs... to wit, my infamous, "Wonder how hot this is", > grabbing a flag... it was cool thermally, the 400V P-P got me, much to > the delight of the technicians ;-) > That happened to me last year. Finished a design, had the layouter put big fat high voltage warnings on the silk screen. In English and in Spanish. Had to because the circuit is small and innocently fed from a 12V rail yet packs quite a wallop. Guess who was bitten by it first? -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Michael A. Terrell on 9 Sep 2009 03:21 krw wrote: > > It's not *that* big of a hose. I don't even want to know how you keep your hose up. ;-) -- You can't have a sense of humor, if you have no sense!
First
|
Prev
|
Next
|
Last
Pages: 1 2 3 4 Prev: MIT rad lab series Next: PIC / MPLAB PICC new user prob.... |