Power transistor question...

Dave · 2011-05-28T13:51:01+00:00

Have a very small (5" screen) black and white TV that has lost its verticle hold. MIL would like it repaired if possible, and kept in storage in case of another hurricane knocking out power and services like the last one did. Upon opening it up I see a power transistor (2SB834) with signs of overheating. My meter tells me that the E/B and B/C pn junctions are switching at .574 and .572 volts, respectively. Am I missing the mark, or does this sound a little low to others as well, as if the pn junctions have been damaged? Its a few more minutes until my candy store opens, and several hours before I'll be able to get by there, but to me, this transistor souinds like a good candidate for replacement. Would appreciate hearing what others think, if anyone cares to comment. Otherwise I'll bide my time until I can prove the part one way or another. Just hope I don't have to order the damn thing and wait a week.Thanks for any comments...Dave

P

Phil Allison 15 years ago

"Jeff Liebermann" "Phil Allison"

** I see no such trend for the electros.

The ESR of most electros is almost constant down to about 500Hz, then rises to a few times its high frequency value at 50Hz. I believe mobility of the ions in the electrolyte is the cause.

Electros uniquely have very low Q factors - which means their impedance * stays low * over a very wide frequency range.

A typical 470uF electro has an impedance under 0.1 ohms from 4kHz to 1MHz as does the 330uF example in the graphs.

High Q capacitors ( film and ceramic) all have sharp impedance dips at self resonance.

.... Phil

Vote

D

Dave 15 years ago

Wanted to thank *everyone* for the info and encouragement I have received in this discussion. Bottom line for me, at least for now, is simply don't test in-circuit (pop one lead and then test after discharging). Have also found a ton of material on youtube and the net that convinces me to get a dedicated ESR meter (or possibly build one.) Anyway, thanks all for a very enlightening discussion.

Take it easy...

Dave

Vote

P

Phil Allison 15 years ago

"Dave"

** This is the one I would buy - if I did not already have two of Bob's earlier designs.

formatting link

BTW

Bob is an old mate of mine.

.... Phil

Vote

J

Jamie 15 years ago

I use a cheap B&K LCR, it does in circuit beautifully.

Jamie

Vote

M

Meat Plow 15 years ago

I'd like to see the meter described test at a higher frequency but other than that I can't find much else wrong. I've been relying on an old Tenma cap meter for years. It's gets you in the ball park. And there were reliable repairs before LCR and ESR meters. I know, been at it from high school back in 1970. These days I tend to shy away from anything I need an ESR meter for. Plenty of vintage stuff out there that people like to keep going.

Live Fast Die Young, Leave A Pretty Corpse

Vote

P

Phil Allison 15 years ago

"Meat Plow"

** ESR meters are MOST useful with older equipment, particularly valve amplifiers.

All those electros are old and have been running in a hot environment.

..... Phil

Vote

D

Dave 15 years ago

Hey Jamie, thanks for the note. So, how do you perform an in-circuit ESR test on a suspect cap? Do you bother to discharge it, or do you just let the circuitry around the cap do that? And how reliable are the numbers you come up with? I tried to do an in-circuit test today on a cap and thought I had gotten lucky with my first shot. A 10uF @ 100V cap gave me an ESR reading of 42.15 in-circuit, but when I popped it out and tested it again it dropped to 0.12. big difference! Still trying to figure out what I might have been doing wrong...

Have since discovered a different cap that may actually be the problem. A .22uF @ 50V cap with an ESR reading of 25+ ohms. Am thinking this component may be where the vertical hold sync pulse gets lost, as it feeds an oscillator/mixer transistor which feeds the output transistor of the vertical hold circuitry. All of this two days after I discovered I had an ESR meter and didn't know it.

Dave

Vote

P

Phil Allison 15 years ago

"Dave" "Jamie = Maynard A. Philbrook = Lunatic, code scribbler & radio ham: KA1LPA

** Blatant lie.

** Jaime is a congenital bullshit artist.

Just IGNORE him.

** All that means is there is some resistance in parallel with the 10uF cap.

Any parallel resistance is computed to a misleading ESR number by an LCR meter.

That's the problem.

With a dedicated ESR meter that measures the impedance of an electro at

100kHz, the effect of parallel resistors or diode junctions that may exist is ignored.

.... Phil

Vote

Power transistor question...

Join the Discussion

Didn't find your answer?