Simulation software for DC/DC converter

of

LTSpice

For the

assignments

Then you should be getting better service. I certainly haven't had any problems. The companies with lousy service get rewarded in kind. My current job is advanced development, so my designs may be in the tens or hundreds of parts (that I rarely pay for).

There is a difference between shipping out a handful of $.05 parts and letting your design engineers field questions from every Tom, Dick, and Harry that comes begging.

Exactly.. there may be no dumb questions, but there are a heck of a lot of inquisitive idiots, and you don't want to waste good engineering talent on those jokers. Just hire a call centre (note spelling) in Bangalore to read the data sheets back to them. Anyone with a serious need will still be able to get through, albeit with a bit more trouble.

Best regards, Spehro Pefhany

buys

No, that's what you pay the sales staff (qualification), FAEs, and disties for. Your alternative is to put the design engineer in Bangalore.

when it

engineers. Big

LTC's

That's right, that's what I meant, except I expect that come with the DC/DC chip. And LTC delivers in that domain. Drivers that have 1ohm or less internally and can drive 10V or more. Many other DC/DC controller chips can't do that, you need separate drivers. Which adds cost, complexity and most of all it eats up precious board space.

Like the DrMOS from Fairchild? I might bite once there is a 2nd source. But only if the price is right. Same with eGaN, it's single-sourced and there is no real IC support I could see.

Yeah, but what would it have cost to do this right and have 25k memory like the DSO I bought? A buck fifty maybe?

A lot of lab equipment from Western companies is already made in Asia, meaning they can compete on cost. It's the engineering where they fell short and that has nothing to do with Chinese competition.

... and then a few years later Harry ends up being chief design engineer at XYZ Corporation. The fact that this one company stuffed me with tons of parts and databooks has later brought them millions in component sales. Because I knew most of their parts, and that's exactly the reason why they did it.

The same model works exquisitely for all those math software places. Students get it for next to nothing, with full support and all that. Then once they graduate and get tasked with automating this, that and the other thing they will resort to what they are familiar with. Because they don't want to screw up. But now it's the full price per license ... ka-ching.

the

As others have mentioned, "product differentiation".

But it does. US engineering isn't free, either.

buys

My point still stands. ...or are you saying that you wouldn't use parts from a company that didn't send you free databooks?

They *don't* get unfettered access to their developers.

who buys

to

of

it

letting

that

Ya know, ten years ago i would have given you a blast for that. The fabulous idiocy i have seen since has placed me on your side. Engineers that cannot connect lines on paper to things in the real world in their designs do give me a bit of a pip.

?-/

the

I'd call it a serious marketing blunder. You cannot do any decent pulse-echo stuff with a 2.5k or 4k sample memory scope. They've lost substantial ground with this sort of "differentiation" and it was totally unnecessary to lose that ground.

I need the exact same number on engineering minutes to design in a mamory chip for 4x the size. It had nothing to do with the cost of US engineering.

buys

letting

No, but naturally one will use what one already knows. And that's the point, this turns into major revenue Dollars for generous companies. The others are, as they say, penny-wise but pound-foolish.

I was told differently by researchers. Universities are often the places where people really push the envelope. Just like I did during my master's project where I got 60dB of dynamic out of a CCD array where the manufacturer never got more than abour 45dB. This cost them several EE hours of support on their side but boy did that pay off. Pretty much instantly because I presented at SPIE and the reaction of the audience was amazing.

buys

Now you're getting it!

I call them SE's "Shoulder Educated", what real world knowledge they have is mostly collected by watching those put their abortion together and seeing the failure rates, but also at the same time, collecting ideas from those that actually know what is wrong, while on sight watching this.

When the project is finalized and working, the Engineer will of course take all credit for it, but not until it is a fully working product. Mean while the failures are blamed on the inconsistent work ethics and errors the tech's are creating.

It's nice to work with SE engineers. We had our last one leave us not to long ago. It was a blessing in disguise.

Jamie

------------ Thank you Paul for your detailed reply. In our last prototype we have done just that. We have extend the protection for 500ns on the way that we added RC's. In serial with base we put 150Ohm resistor and parallel of the B-E transistor,we add 22nF capacitor. Tau is around 3 milliseconds and when you put some 3V on it ( from the "shunt" ) it look like that the 0.5V will rise to this value in cca 0.5 microseconds.

I dont know what else can i test before we hook up the transformer but i agree with Robert that EMI can be one of the problems.

I was also thinking that we need to add some extra look at transformer overcurrent, magnatising current,number of windings,voltage crash...

The prototype is now working perfectly and on weekend we will test it with transformer. If you have some additional suggestion please write down...

Can you post your LTSpice .ASC file? It is a text file.

I still wish I could have gotten a copy of PSpiceHDL. That was a cool program!

They combined the PSpice and NC-Sim simulators, combined Probe with an MS-Studio like debugger front end, and you got a full mixed mode simulator that could handle almost anything. Then the folks in San Jose got a look at it, realized that this was a $20K product that would put their $100K+ products out of business, and laid off the entire PSpice team... :-(

buys

Or, like MicroSim and PSpice. The free 'student' or 'evaluation' version was more than good enough to do small designs, so thousands of students used it in class, and then wanted it when they went out into the 'real' world. That, plus the reasonable cost for even the full version of the software, gave them like a 90% market share. Even now, with LTSpice and other alternatives, I think they have an 85% or so market share, and they only do the most minimal updates or improvements to it...

Charlie

buys

letting

I still have my old Microsim license. Including the nice cloth-clad Synthesis-Analysis binders in the cloth-clad pouches. Has this good old "IBM-feel" to it. But it's the DOS edition and I haven't used it in ages.

What does that yellow rising sun logo with the red bird wings in there represent?

Anyhow, I no longer belong to the 85% because the modern version is way too buggy for my taste.

-------- Yes i can, but there are bunch of parts missing. Some mosfets, some 4040's 4060's, 4081 and so on.