Interesting question/rant? from EEVblog Dave

If they have questions, let them ask. We learn from that, too.

Too many kids in engineering school these days, kids without a real aptitude for electronics. They mostly learn how to code.

Yup, that gets them a gold star.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

There's something special and magical about the 2N2222. It must have a built-in oscillator.

Tends to be roughly 100 MHz. TO-can and wirebond geometry or something.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

There you go, you blew the budget. Those two beads now made that circuit costly over simply getting something ready made! :)

Jamie

I once took to a research job interview a piece of kit that I had lashed up in the garden shed a couple of days earlier, which out-performed the results of their previous year of research by a factor of several hundred. The reason mine worked and theirs didn't was because they had fundamentally misunderstood the physics of the problem.

The project superviser's reaction was to tell me in no uncertain terms that my knowledge of the physics was WRONG, my ideas of electromagnetic design were out of date and they were going to press ahead with their super high-tech materials and their computer-controlled system.

During a break, I went to their library and checked the physics against a standard textbook, then slipped the reference to the post-grad who was lumbered with this idiot superviser. Needless to say, I didn't get the job - and after that interview I didn't want it.

--
~ Adrian Tuddenham ~ 
(Remove the ".invalid"s and add ".co.uk" to reply) 
www.poppyrecords.co.uk

No inductor required! Just putting a resistor in the emitter will make the real part of the base impedance negative. The extrinsic base resistance is often enough to keep that under control, but not always.

Cheers

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

What worries me is that you've labelled it 'advanced'

NT

Well, I start them on a 2-resistor voltage divider. 2 volts and a pair of 1K resistors; don't want to stress the poor kids too much, early in the interview.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

I was testing a little RF amplifier recently. Common base, moderately tuned collector. The emitter had 4.7 ohms degeneration, and was biased at

5mA. MPSH10. Trying to calculate the input impedance -- for proper input matching. I could, at best, determine the emitter was very close to 0 ohms, plus or minus a few. It's already got 4.7 ohms in series, so it's a little big negative already. Gee, great. ;-)

I do have a question for the noise specialists here (Phil?): how, exactly, would one match (for optimal noise and power gain) to a zero ohm port? Clearly, the answer must be: you don't. But then, what IS the correct answer, and how does one measure or calculate it, in the presence of parasitic feedback?

A related question: what's the best collector load impedance? Again, it's clearly not the conjugate AC impedance, because the collector is predominantly a capacitor, plus some Early effect. R ~= VCC / Iq seems the most reasonable (if nothing else, it's near the maximum power compression point, for obvious reasons), but that seems like a bad choice, not invoking any device characteristics to reach it.

On the plus side, BJTs are generally better than anything else in history (*cough* toobs), so relatively speaking, you can afford to waste a lot of gain, because they have a lot of gain to begin with. But by the same token, you can stand to win that much more dynamic range, too.

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

Well, unfortunately negative resistors don't have purely imaginary noise. ;)

At high frequency, the collector load impedance gets reflected back to the base, so the ideal noise match will depend on everything including what you had for breakfast. On the very rare occasions I've needed to do that, I've either done it by iteration in front of a noise figure meter (long ago when I was more of an RF guy) or pulled out Carson and drudged through the algebra.

Usually my stuff is made in small enough quantity that a bit of extra circuit complexity is a good trade for more easily predictable behaviour. Thus I tend to use a lot of cascodes for RF amps. As I've mentioned before, my fave ATM is an ATF38143 pHEMT cascoded with a BFP640 SiGe:C BJT, which has some gigantic Early voltage. Approximately infinite gain and zero noise, if your spherical cows are grazing above about 10 MHz. ;)

(The problem is keeping the BJT from oscillating without losing too much bandwidth in the process. The pHEMT is pretty stable.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

:) You're way too kind.

NT

Well, it makes me feel bad to interview a kid who worked 4 years to get an EE degree, and I have to dump them ASAP because they are obviously clueless about electricity.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

** OTOH, I bet he was a whiz with simulation software and on-line games.

Obvious truism: people cannot know much about things they have never seen or dealt with.

... Phil

In the olden days, they'd wind magnet wire over a carbon comp resistor. Nowadays, you can get ferrite beads that do a pretty reasonable job, though the impedances and cutoff frequencies aren't quite as selectable.

Something else I need to do, too. I have a GK71 (the Soviet's answer for the 811) hanging around. It would make a lovely Tesla Coil. Not sure what to power it with; an MOT seems handy, but rather aggressive on the peak voltage.

Or maybe a power amplifier for blasting circuits with RF, but I don't think you can really get much bandwidth from a tube, so it would be no end of coils and varicaps and grid/screen current tweaking...blah!

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

Have a nice day.

That BJT is a 45-GHz device, though. (Among its many other talents, it can easily oscillate at 12 GHz.)

See if you can make a quadcopter fall out of the sky. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

Which is getting to be a bigger and more legitimate concern as time goes on...

Others have their pistol or shotgun, me, I have magnetrons and tin cans handy. ;-)

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
Website: http://seventransistorlabs.com

much

for

end

Hunh? Tubes narrowband?? Why? What part of the physical process of their operation make you say so? They did microwave with tubes before they did it with transistors.

An old 6J6 is good from DC to 200 Mhz. Isn't that enough?

?-)

their operation make you say so? They did microwave with tubes before they did it with transistors.

An old 6J6 is good from DC to 200 Mhz. Isn't that enough?

Mmm?

Can you give me any examples where *a single circuit*, built with tubes, provided flat DC (or LF limit

:)

Also, the 454 front end was rated to 150 MHz with Nuvistors and was still usable at 200. But I don't think it qualifies under your "no tweaking needed" constraint. They drifted like Chris Harris.

-- john, KE5FX