crazy NPN

hfe = 20. Am I missing something? Where's the bit deserving of the exclamation mark?

** High voltage NPN switches are like that - 1500V horizontal output transistors ( like the BU208 ) for colour TV show betas around 20 at 0.5A falling to less than 10 above 3 amps.

Simple Hfe testers often show zero readings with one cos the do not supply enough base drive.

... Phil

Yeah, but when the base circuit is doing 1V, 1 mA and if the collector is doing 1200V, 3 mA the power gain is around 70 dB.

It's like the old trick of spec-ing an op amp by its DC gain (20,000 V/V) and output-referenced offset (15 mV times 20,000)... would you believe offset voltage of plus or minus 300 V?

Still, the old CA3140 was kinda useful regardless. My main gripe was the quiescent current: so high, I had to heatsink the thing.

At the rated peak collector current of 400 mA, the beta isn't specified, but is probably around 2.

The data sheet is sort of skimpy.

--

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

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

Pedantically speaking, beta isn't mentioned? So the answer is, NaN?...

I would be a little more comfortable if they had left off pin 2 on that poor SOT-223. Or same with a DPAK (which usually has a nub there), if it were offered in one.

No graphs, no mention of fT or t_r. I'm sure t_stg is characteristically long, but a ballpark would've been nice.

A 6V6GT has much higher hFE, is almost certainly much faster, and is capable of dissipating more power, too. I'm not feeling won over, here. ;-)

Tim

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

"John Larkin"  wrote in message  
news:5fugia95gsbkt4e0ml5709t7q02p8k2777@4ax.com... 
> 
> http://www.st.com/web/en/resource/technical/document/datasheet/DM00047825.pdf 
> 
> Check the betas! 
> 
> 
> 
> --  
> 
> John Larkin         Highland Technology, Inc 
> picosecond timing   laser drivers and controllers 
> 
> jlarkin att highlandtechnology dott com 
> http://www.highlandtechnology.com 
>

HFE isn't beta? If typ is 3, I wonder what the minimum might be.

I might have trouble putting 16 of them in a box the size of a bagel.

--

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

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

yeah, but this part is all about voltage gain, not current gain.

--
umop apisdn

Seems for good reason. Phil said something about horizntal (line) outputs f or yoke driven TVs ? Those pins are to close together for a kilovolt. What' s more, a gain of 3 is not really impressive. The current, are you really g oing to drive a relay with a 1,200 volt coil ?

And this "haptic" - what does that mean ? A qwiki comes up wiht something t o do with touch. You want to shock people of they touch your stereo or some thing ? Well with a wimp like this they are not likely to convict you of mu rder...

But then again, remember the Watchman ? Hell this might even be better than the transistors they used in those.

What sort of high voltage can one safely generate on FR4 without the fear of tracking arcing ? Is it enough to have wide enough gaps or will it eventually track through top to bottom ( 1.6 mm) assuming there is a ground plane on the bottom layer ?

--
This email is free from viruses and malware because avast! Antivirus protection is active. 
http://www.avast.com

I hadn't even thought about that. Sounds like a circuit where if you leave it running and go to the fridge and get a drink it will arc over and burn the board.

Off the top of my head I think you got 1.6 mm.

Of course if you are worried about silly little things like getting fired and all that or going under, you could probably elevate the middle pin betweent the base and emitter and just use the one at the other end.

Pedantically, no. At least, I think there's some stupidly subtle difference in definition. For everything else, yeah it's the same though.

Not having a minimum or maximum is... well, just as skimpy as everything else in there, really.

On a related note, it seems to me, manufacturers have stopped controlling for Qg(tot) / Qdg / Qgs. Typical is all fine and good, but I haven't been seeing max/mins on those parameters, quite often...

Fine, then use a few 5687s, see if I care. :^)

(Which, by the way, has an hFE up in the 5-10 range, and is capable of more peak current than the present transistor, too!)

Tim

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

Short term: air gap minus 50%, roughly? Tracking/creepage is... kinda-sorta more of a contamination thing, so the spark still travels through a lot of air if it's jumping over clean board. At least for the first couple strikes.

So, it'll handle it no problem, at least for a little while.

After a few years sitting in a mildly dusty environment... who knows? I wouldn't bet a $20 on it, myself...

You can probably convince the board house to route slots that thin. Probably at some marginal extra cost, for the slower feedrate plus more fragile endmills.

Look at the slots in this fancy-ass bastard:

They've routed around SOIC pins, for god's sake! Must be 20 thou cutters,

30 max. Of course, it's a "cost no object" type of product, so they don't bat an eye at that sort of stuff.

I have to wonder just how much leakage you save, over board creepage (give or take guard traces), against the semiconductor junctions inside the thing. (@25:55 ish) It's just a common CMOS switch, and it's not like they've Peltier cooled it or anything. And 2N3904/6s in other places. It's either madness, or genius...

Tim

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

.pdf

Current gain of 3 is not unusual for that sort of voltage rating. Its the p rice of high voltage. Used to play with BU208A quite a lot, also specced at 3. It was a big improvement on the previous generation that was rated at a bout 700v IIRC, resulting in needing to use 2 in series, plus all sorts of horrid circuitry to handle that.

NT

Yap, not unusual for a HOT. The datasheet's claim is amusing:

Features ■ High gain

--
 Thanks, 
    - Win

Should I use the regular ones, or the cryogenically frozen ones? Bronze or black plates? Looks like I can get them for about $40 each.

--

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

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

My situation will need about 800 volts, no problem on FR4 with sensible clearances. I've seen bad oscilloscope circuit layouts that collected dust and arced at several kilovolts, cured with conformal coating.

This made about 900 volts, no problems.

Top to bottom, through an FR4 board, should be good for 10 KV or so.

--

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

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

High power gain, maybe.

There is only a typ beta, no minimum! I don't think that will be my transistor of choice.

I suppose I'll use mosfets, capacitance and all.

--

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

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

OK, that is a rather precipitous drop in gain. Thanks for pointing it out to me.

Glad you asked! Only authentic name-brand, Premium Quality, NOS, gold plated, degaussed, cryogenically treated, virgin-signed types will do.

...

$40 sounds like a list price; I should think you can find them cheaper (if not virgin-blessed), but they are kind of an oddball.

Kind of the 1950s equivalent of having a 60V, 2A "Low Vce(sat)" transistor in a SOT-23-6 (dual). Those were pretty neat when Zetex (and others) introduced 'em.

Ironically, 6V6s and whatnot are current production, but the modest market size and current demand tend to inflate the price over old-stock prices. If you wanted cheap tubes (it's a hypothetical, bear with me), the myriad oddball TV tubes are still the way to go.

Unbranded subminis can go for pretty cheap, but they aren't much use outside of nuclear attacks (ex-JAN stock turns up quite commonly), being that they have specs largely similar to a 2N3819 with the drain voltage multiplied by 10, plus that pesky heater.

Tim

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