engineering graduate school question

Joel Kolstad wrote:


> I think you're guilty of making some generalizations as well here, Vladimir.

This is my personal experience with hiring freshmen from school. As 
pointed by many, the grown up engineer is judged by his merits. One of 
the not so many merits of a freshman is the diligence demonstrated in 
obtaining the degree.

> Yes, many BSEEs aren't able to "run a project" (even a very small one) on 
> their own, but some are.  Similarly, I've had MSEEs and PhDs whose practical 
> skills were so poor 

I know :) That sort of people can be good at paperwork though. Somebody 
has to do the boring part, too :)


> I honestly think I could have done better standing in the 
> an engineering school's student union, talked to a couple dozen students as 
> they walked by, and returned with the best candidate.

That is exactly what we tried. It appeared that the percentage of the 
good guys with BSEEs is 1% vs somewhat 10% in MSEEs. It is just more 
economical to draw from MSEEs.

> P.S.: Since this is related to your area... we did have a guy come in to 
> interview about doing DSP work, and I asked him how he might go about taking 
> some digitized signal he had in the memory of a DSP and reversing its 
> spectrum -- preferably as efficiently as possible.  His answer was that he'd 
> take the FFT, reverse it, and then perform an iFFT. :-(   That's the sort of
> answer I might expect from someone right out of school, but not from someone 
> who'd been in industry for many years as he had.

There is a zillion of possible methods for the spectral inversion 
depending on what exactly is required. Perhaps, the simplest is 
multiplying every second sample by -1.

> P.P.S.: I clicked on your web page.  Shouldn't that circuit board photo you 
> have at least be of a DSP rather than a microcontroller? :-)  Or do you do a 
> lot of "hard core" DSP in microcontrollers?

For DSP, ADI BlackFin is the main workhorse. We also work with TMS 28xx 
and 55xx. As for the small MCUs, 68HC12 is the preference, and, indeed, 
we do some DSP work with it as well as with Atmel AVR.


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

Joel Kolstad wrote:
> "Salmon Egg" <salmonegg@sbcglobal.net> wrote in message 
> news:C29C7795.8088F%salmonegg@sbcglobal.net...
> 
>>On 6/18/07 3:55 PM, in article pan.2007.06.18.22.55.31.843507@example.net,
>>"Rich Grise" <rich@example.net> wrote:
>>
>>>It depends what you want. If you love going to school, then do that. If
>>>you want to actually do something useful, then get some practical
>>>experience.
>>
>>Of the responses I have seen so far, this is the best advice.
> 
> 
> I have to disagree a little here.  If your goal is to be, e.g., an analog IC 
> designer, unless you're Jim Thompson's kid it's very difficult to do these 
> days without the formal education.  The problem is that building ICs requires 
> big bucks, which implies big companies, and big companies are notoriously bad 
> about using academic credentials as the first "gatekeepers" to employment.
> 
> On the other hand, for someone who wants to write software, do some digital 
> design, perhaps some power supply stuff, etc. (i.e., decent chance of getting 
> a job in a smaller company as a "general-purpose useful person), it is 
> reasonable to just learn on the job.  Perhaps not the most secure career 
> strategy, but I certainly know several people who've taken this route and it 
> works just fine for them.
> 
> 
He didn't really say what his main interest was, so it is hard to tell...

20 years ago, when I went for my Masters, I picked UCSB for two reasons 
- one, the great location, and two - they were doing free electron laser 
research there, and I had an interest in laser launch systems for space 
travel.

Now, when I got there, I found out that one - you couldn't afford to 
live there, and two - the laser research was all going on in the Physics 
department, not the EE department, and the only faculty member in EE 
that had worked with them retired last year... ;-)

Charlie

"Vladimir Vassilevsky" <antispam_bogus@hotmail.com> wrote in message 
news:sdXdi.5171$bP5.559@newssvr19.news.prodigy.net...
> There is a zillion of possible methods for the spectral inversion depending 
> on what exactly is required. Perhaps, the simplest is multiplying every 
> second sample by -1.

Yes, exactly -- inverting every other sample is a very good answer.  "FFT, 
reverse, iFFT" is probably about the worst you could do and shows no 
"practical" experience.

> For DSP, ADI BlackFin is the main workhorse. We also work with TMS 28xx and 
> 55xx. As for the small MCUs, 68HC12 is the preference, and, indeed, we do 
> some DSP work with it as well as with Atmel AVR.

I only have experience with TI 55x DSPs, but I've always been told the ADI 
chips are somewhat friendlier to program, if a little slower.

---Joel

"Charlie Edmondson" <edmondson@ieee.org> wrote in message 
news:4678435a$1@news.cadence.com...
> Now, when I got there, I found out that one - you couldn't afford to live 
> there, and two - the laser research was all going on in the Physics 
> department, not the EE department, and the only faculty member in EE that 
> had worked with them retired last year... ;-)

Nice.  I was in the microwaves department (part of the "school of electrical 
engineering and computer sciences"), and ended up doing circuit modeling.  It 
happened to involve components at GHz frequencies, but at least in the 
research I did the frequencies were pretty irrelevant -- they could have been 
1-3Hz or 1-3THz rather than the 1-3GHz they were and my thesis would have been 
the same.  (Much of the time, for numerical accuracy reasons, you ended up 
normalizing a lot of the data to, e.g., 1Hz or 1 rad/s anyway...)

There had been some early discussion of doing some cool RFID stuff, but 
unfortunately we weren't able to get funding for it.  Circuit modeling was 
funded, so there I was...

What *did* you end up researching?

Vladimir Vassilevsky wrote:

   ...

> I found that when hiring a freshman just from school, the BSEEs are 
> worseless. They don't know anything, they can do nothing and, what is 
> much worse, they don't want to do anything about that. It will be years 
> and years till they reach the level of apprentice.

You over generalize. My boss in a small development lab threw me out 
(well,strongly urged me to leave). I was one of two technicians there -- 
the other had been with the firm since before I was born! The boss said 
that I was already doing the work of an engineer, I might as well bet 
the diploma and get paid like one. (One of the engineers there had no 
diploma beyond high school, but he was a real genius. George Gauthrin, 
AKA Crazy George, was the guy who invented the 60-dB low-distortion AGC 
that got us the Mercury capsule audio contract.)

> The fresh MSEEs and PhDs are lot better in the general; it takes only 6 
> month or so to make them productive. The advanced degree is an indicator 
> of diligence, discipline and ambition; this is good.

Some fresh Ph.D.s I worked with were inclined to ignore the importance 
of the curved side on a 'lytic symbol, thereby making loud noises. One 
inserted a wood screw with my chisel-point soldering iron, then asked me 
what the plug and cord were for.

   ...

Jerry

P.S. "Worseless"; a good word. I like it!
-- 
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Mark Walsh wrote:

   ...

> I have no beef with credentials, but they aren't an acceptable substitute 
> nor necessarily an indicator of competence.

When I was am MTS at RCA Labs, the division president, when he was 
escorting a visitor, would stop me in the corridor and introduce me as 
"The only staff member we have with a steam-driven pencil sharpener." 
(He neglected to add that I was one of the very few staff members 
without an advanced degree.) I then gave the obligatory explanation that 
in order to avoid tending a boiler, I ran it off the compressed-air 
supply. The visitor and I would then shake hands and go on our ways.

If I decide to abandon retirement, I'll ask you for a job.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

On 6/19/07 7:06 AM, in article 137fokc4ab1kac9@corp.supernews.com, "msg"
<msg@_cybertheque.org_> wrote:

> (repost attempt, first post failed)
> 
> Salmon Egg wrote:
> <snip>
>> Modern electronics and radio is not really feasible or
>> economical for home construction. I can get an FM stereo radio with
>> earphones at the 99&#4294967295; store. You cannot buy the parts for a transceiver for
>> what it costs for a much better piece of equipment commercially.
>> 
> 
> FWIW, please try my homebrew HF receiver (online/interactive)
> at <snip>
> 
> I hope to have measurements of its characteristics available
> on the site soon also.
> 
> Regards,
> 
> Michael

I do not know if this poster is malicious, but I ran into problems trying
his site. I had to force quit my browser to continue.

Bill

-- Support the troops. Impeach Bush. Oh, I forgot about Cheney.

On Tue, 19 Jun 2007 14:57:54 -0700, Salmon Egg wrote:
> On 6/19/07 7:06 AM, in article 137fokc4ab1kac9@corp.supernews.com, "msg"
> 
>> (repost attempt, first post failed)
>> 
>> Salmon Egg wrote:
>> <snip>
>>> Modern electronics and radio is not really feasible or
>>> economical for home construction. I can get an FM stereo radio with
>>> earphones at the 99&#4294967295; store. You cannot buy the parts for a transceiver for
>>> what it costs for a much better piece of equipment commercially.
>> 
>> FWIW, please try my homebrew HF receiver (online/interactive)
>> at <snip>
>> 
>> I hope to have measurements of its characteristics available
>> on the site soon also.
> 
> I do not know if this poster is malicious, but I ran into problems trying
> his site. I had to force quit my browser to continue.

I get a bunch of java crap that doesn't want to talk to me unless I go
through some hoops, which I don't do on websites.

Cheers!
Rich

Salmon Egg wrote:

>>
>>FWIW, please try my homebrew HF receiver (online/interactive)
>>at <snip>

>  I ran into problems trying
> his site. I had to force quit my browser to continue.
> 

Apologies for any problems you may have experienced. The
main page uses a java applet to handle the remote-control
session with the receiver.  It is known to work well
on Netscape/Mozilla and IE on Win32 using JRE1.4.x and
on Netscape/Mozilla on Solaris; I don't have many reports
of how well it works on MacOS/OS X, Linux or other X-windows
OS implementations.

What browser/OS were you using? Did you look at the Java
Console messages in your browser?

Your newsgroup posting headers include domains 'prodigy.com'
and 'yahoo.sbc.com'; are you using some kind of bundled
browser/Internet access software that may be conflicting
with Java on your machine? I also see reference to
Microsoft Entourage (I am not familiar with that client).

It is difficult to provide a client-side realtime browser-
embedded control method without using Java or undesirably
Microsoft-specific active controls (which I'm not using).

If anyone has problems, I'd appreciate a report.

Regards,

Michael
(msg _at_ cybertheque _dot_ org)

On 6/19/07 2:57 PM, in article C29D9F72.80C89%salmonegg@sbcglobal.net,
"Salmon Egg" <salmonegg@sbcglobal.net> wrote:

> On 6/19/07 7:06 AM, in article 137fokc4ab1kac9@corp.supernews.com, "msg"
> <msg@_cybertheque.org_> wrote:
> 
>> (repost attempt, first post failed)
>> 
>> Salmon Egg wrote:
>> <snip>
>>> Modern electronics and radio is not really feasible or
>>> economical for home construction. I can get an FM stereo radio with
>>> earphones at the 99&#4294967295; store. You cannot buy the parts for a transceiver for
>>> what it costs for a much better piece of equipment commercially.
>>> 
>> 
>> FWIW, please try my homebrew HF receiver (online/interactive)
>> at <snip>
>> 
>> I hope to have measurements of its characteristics available
>> on the site soon also.
>> 
>> Regards,
>> 
>> Michael
> 
> I do not know if this poster is malicious, but I ran into problems trying
> his site. I had to force quit my browser to continue.
> 
> Bill
> 
> -- Support the troops. Impeach Bush. Oh, I forgot about Cheney.
> 
> 

It worked for me, as far as I went, but it is klunky.

Don