cross-discipline job descriptions

I once interfaced an 029 to a PDP-11 so we could punch cards from it.

John

i might be one of them such people...

so what is it? that the filter passes *some* DC? right now, that's all i can think of. it ain't necessarily an LPF.

r b-j

Well, it'll have its peak response at DC, but beyond that, there's not a great deal you can say for sure. Most such filters that people use, e.g. boxcars, von Hann windows, truncated Gaussians, and so on, have bad passband slope, poor skirt selectivity, and (on the plus side) relatively low noise gain.

More generally, you can take any FIR filter and offset it by a constant equal to -1 times its largest negative coefficient, so apart from the effect of the offset, it can look like anything you want.

Did you have other properties in mind?

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

s t

What if you hit such a filter with a step?

Nah - 12AU7. ;-)

Cheers! Rich

A building, especially one built with tax money, should be functional, period. If it's private property, let them doll it up however they want; if it's a government office, let the government workers buy the artwork they want out of their paychecks.

But using taxpayer money to pay for frivolities is wrong.

Thanks, Rich

Maybe not, but it's supposed to be.

Thanks, Rich

i'm not even sure it would peak at DC. i guess i'll have to come up with a counter-example.

ah, monotonic (strictly increasing) step response.

okay, so i would flunk your test and get filtered out.

r b-j

Me, too, apparently, althought the pass criteria haven't been stated. I guess I'm not seeing the issues here.

Eric Jacobsen Minister of Algorithms Abineau Communications

Actually, not in my cases. I work a lot in this kind of mode and not once has this micro-team of me and a fresh hire on their side been the pacing item. In fact, right now we'll come up against the usual scenario where a design will be ready some time around April and we'll be tapping our fingers, waiting for the rest of the project to show up. We already are way ahead of the eight-ball, and at a point where our part has become like a freight train that can't be slowed down anymore :-)

--
Regards, Joerg

http://www.analogconsultants.com/

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

You actually spend a lot of time working with and directing super-bright junior engineers fresh out of school and with little experience? How do you do all the required hand-holding without being on-site? Most of them will happily go off and spend days doing unproductive things, just because they have not yet learned..

Yes, some of them were fresh out of college. The very first thing I get them into is learning LTSpice. That alone is usually an eye-popper to them. Now I can feed them sims and massage their files. Next I often tell them to find a Tektronix 2465 on Ebay and get it. Then comes the more hardcore stuff such as "no, even if da professah told ya about star grounding, it plain don't work", and so on. If they are bright it doesn't take all that long. However, we generally do prefer candidates with a solid hobby or ham radio background. Can't use people who aren't even able to solder. I know that flies against the philosophy of many hiring managers but for me that's what it is. For example, if someone has successfully built a half-kilowatt 3-30MHz amplifier there's a whole lot of stuff you do not have to explain to them, they just know.

Key is to always explain why something works or doesn't.

T'is the beauty of the world wide web :-)

Email exchanges can get down to interval cycles of very few minutes and if that ain't enough one can always schedule a video conference on the web. I also found the investment into a headset that fits my cordless phone to be a real benefit. I used to smirk when I saw sales guys running around like that, but not anymore.

Not if they have committed to a schedule with milestones and their boss tays on top of things.

--
Regards, Joerg

http://www.analogconsultants.com/

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

I can only say 'ditto that' from my own experience. I haven't done much remote work with bright young things, but I've worked with bright engineers with less than 5 years experience, and I've worked with bright young things on site.

Since I generally consider myself under the gun to get folks spun up and running fast, I don't let the dust gather too much.

--

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

Joerg, how in the world can they "commit to" (or even buy into) a schedule with milestones when they have virtually no idea how to do the job or how much total work is required?

The only time I've had problems with remote contractors is when they have embelished on their experience. Then things tend to fall apart and I have to get them to fix their work, correct their work myself, or re-assign it (if there is time). The same situation as with junior folks. Fortunately, most people who manage to work remotely are pretty good at what they do (a few are extraordinarily good), have a lot of empathy (bandwidth is low, so that skill is needed at both ends), AND learn quickly. For relatively straightforward stuff on a fairly loose schedule, it's not a big deal.

Mentoring a bunch of junior engineers, sure. Remotely? No, thank you!

I did that on my very first job. Actually, I later found out that my first boss took the liberty to sign me up for the commitment _before_ I had accepted the job. He just knew I would accept, or at least hoped I would. Oh, and it was engraved in the bonus objectives ...

Got to step out of one's comfort zone once in a while. When I interviewed for that first job I had no clue what they'd do with depth sounders on humans, found out that it was B-mode scanners, and that I'd have to take over the frontend design from an engineer who was going back to the US. So I had to get a clue, and pronto. This was in the

80's, before Wikipedia. I also didn't know how hybrids are designed but they told me I had to design two, as a side dish.

Many times during my consulting work I've had to commit to schedules and stuff before there was any semblance on how we'd go about it. Neither on my part nor at the client. We had to figure it out, somehow, and by a due date.

A loose schedule would be nice for a change. Haven't had the luxury to enjoy one in a long time. The tough part is when stuff spills over into the weekend and the honey-do schedule needs to be pushed out a bit :-)

So far I have enjoyed that. Actually, Usenet is a good training turf to hone that skill.

--
Regards, Joerg

http://www.analogconsultants.com/

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

I avoid mentoring altogether, unless it's simply in seminar (and paid :-)

Most young graduates, and even some older ones, can't find their asshole with both hands, a mirror and an assistant. And they can't keep up.

Even MIT grads, like some I met recently at Intel, are starting to bug me... robots have more personality and enthusiasm :-( ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

Typically the young folks underestimate drastically how much work is required and over-promise far beyond what is realistically possible. They need to be protected from themselves or they'll burn out and that's not good for anyone in the long term. They're used to being handed a different sort of problem with far different expectations in many areas.

Well, there's always the weekends to fix the other guy's foul ups, but it's not my preferred way of spending time. I'd rather watch my kid play baseball or hockey, read, go to a play or whatever.

Anyway, in the big picture , there are always a lot of work packages that are well off the critical path. That's a good place for jrs to be resource-leveled and learn the ropes. If they're late, it's a learning experience. Early? Get kudos and something else interesting to do.

For better or worse, there are rather few young engineers on usenet these days. Most are on mailing lists and web-based forums.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

If they'd work totally on their own, yes. That is why there needs to be a good manager for the group. Someone seasoned enough to have been through the paces. After all, you wouldn't hurl a guy with a fresh degree straight into a management function, would you?

That is also where the consultant comes in. When the project is rather special and the group manager or VP engineering might have never been through a project like this he or she is expected to find the proper outside expertise. Just like nobody in their right mind would show up for a court trial without proper counsel.

Unfortunately many do it anyhow in engineering, trying to go it alone with their team. And then stuff hits the fan. Major schedule overruns, the CEO all hollering and red in the face, and so on.

Then they have to start in large corporations. Small companies generally do not have any work off the critical path. Believe it or not, I actually liked it when they threw me into the hot seat rather soon at my first job. Initially the engineer from whom I took over was still there but eventually he chucked me the keys and I was on my own.

Over here, yes. But, for example, on European newsgroups there is a surprising number of under 30-year olds. Maybe that's because companies over there often consider an engineer above 45 a geezer :-(

--
Regards, Joerg

http://www.analogconsultants.com/

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

Hi Robert, I may have also flunked the test. miso-at-sushi's question is not a simple one. My first thought was,

[1] "Well, a boxcar (moving average) filter has all-postive coefficients. So Rick, what's "interesting" about a boxcar filter?" Not much can be said that's interesting say about a boxcar filter. [2] We know miso-at-sushi's filter can't be a highpass filter. [3] I thought of a passband FIR filter (whose impulse response I can't visulaize at the moment) could have a constant added to its coefficients so their all positive. Such a filter's freq mag response would still be a passband with a non-zero response at DC (zero Hz).

So I couldn't think of anything "interesting" (extraorddinary) to say about an FIR filter whose coefficients are all positive.

But if miso-at-sushi is looking for an answer related to applying a step input to the filter, then that makes the question a borderline "trick question." In practice I don't apply step inputs to filters. (Maybe the control systems guys do, but I don't.) In the practical filters I've built, we often go to great lengths to make sure a filter's input has no DC component. For example, "DC blocking" filters.

Maybe miso-at-sushi's question would be less "unreal" if he'd asked: "How does the behavior of an FIR filter change when you add a constant to its coefficients so that they're all postive? (I wish I'd thought of that question back when I was creating homework problems for the 3rd Edition of my book.)

Anyway, it looks like I would have also flunked the test.

See Ya', [-Rick-]