OT: Any electronic pros here, need help testing transistors

[Steve Walker]

I tore apart an old Denon amp.(I was bored) It has 2 heat sinks, with 4
transistors attached to each. Must be pairs, I guess, as each one has,
from left to right, Sanken transistors, C2837, A1186, C2837, A1186.
I don't know squat about transistors, except when measuring resistance,
if two the same have different resistance, one is bad, usually the one
that has continuity between leads. I found two (c2837 on the same heat
sink) that have continuity between the left lead and the heat sink,
looking at it from the front with the numbers facing you. How can I
check these out? I have a good digital Protek multimeter, and time. <G>
Do they have to be replaced in complementary pairs? I found a cross
reference to NTE36(C2837) and NTE37(A1186), wow, not cheap. I'm gonna
look like a rocket scientist to SWMBO if I can fix it. Not quite a
theoretical physicist, just a rocket scientist.

Leonard? Leonard? Leonard? (somebody will get this, I hope)

--
Steve Walker
Fusi...@frontierbrain.com (remove brain when replying)

[Bruce L. Bergman (munged human readable)]

Go look at the spec sheets, some transistors have the mounting tab
connected to the Emitter pin, so the heat sink has the power rail +V
on it. Therefore it isn't a problem if both the left and right
channels are tied together, they run on the same line.

If you want to function test them like that, they have to be out of
the circuit. Then you can check continuity between leads.

But I'd start by powering it up and checking all the Power Supply rail
voltages, that's usually a good place to start if it's Totally Dead.
Even if it lights up and displays and puts out Line Level at the back,
the PA power rail could be out.

And if that's good you start tracing audio signals through, see what
works and what doesn't. If you have a good Right output and no Left
Side, it's likely one of those big transistors. If certain functions
don't work you will get the incoming signal at the tuner output or the
Tape input and can follow it through the pre-amp, switching and power
amp stages.

--<< Bruce >>--

[Dave__67]

> Fusion...@frontierbrain.com (remove brain when replying)

If you can find a data sheet and sort out what type they are, yopu may
be able to accomplish a bit with the diode-tester on the meter, if it
has one.
I have heard some can be snarfed with a diode-test, not sure.

Dave

[Cydrome Leader]

first question in troubleshooting is "what's the problem?"

what's wrong with the amp?

[Steve Walker]

Oops. One side doesn't work. If you crank it up that side will produce
sound, but not loud, and very staticy. This is what led me to believe
the output transistors were bad

I do have the two heat sinks out of the amp, with transistors still
mounted. Anyone got any information on how to check these out, with a
multimeter? Mine, Protek D-910F, has transistor (NPN and PNP) sockets on
it. But I don't have a manual, nor do I understand all the information
on the transistor datasheets.

[Wild_Bill]

Poking around in electronic circuits, without knowledge of what one's doing,
can result in more damage and/or safety issues.
Use the amp for learning something new, while reading basic electronics
texts, or just throw it away.

There may already be some repairs for that model, to be found with an
archive search of:
sci.electronics.repair

Bazinga

--
WB
.........


"Steve Walker" <fusi...@frontier.com> wrote in message
news:bOKdne8S9KWxrTHT...@giganews.com...

[Carl]

A diode test will identify most bad bipolar transistors (PNP, NPN are
bipolar, diode test not real effective on FETS)

Set the meter to the diode test settings, most meters then measure
the forward voltage drop on the junction.

Isolated from surrounding circuits elements,
Red on any "P" and black on any "N" should read 0.600 to 0.800 ( this
can vary some, older germanium transistors are about half that)
Red on any "N" and black on any "P" should read open circuit
Between the 2 "Ps" or the 2 "s" should read open circuit.
for any transistors that makes 6 measurements, 4 open circuit, and 2
0.600 to 0.800

I made $15K over a ten year period fixing electronics for a former
employer, and never used more than a multi meter. Not the most
efficient method sometimes, but you can do a lot.

As far as safety or damage goes, it's broken now and can't get much
worse.
If you are a real neophyte don't work on it with the power on.

CarlBoyd

CarlBoyd

[Steve Walker]

On 10/29/2011 20:31, Wild_Bill wrote:
> Poking around in electronic circuits, without knowledge of what one's
> doing, can result in more damage and/or safety issues.
> Use the amp for learning something new, while reading basic electronics
> texts, or just throw it away.
>
> There may already be some repairs for that model, to be found with an
> archive search of:
> sci.electronics.repair
>
> Bazinga
>

Thanks. I had forgotten about that newsgroup. I didn't mean to sound
like I was an electronic doofus, I'm familiar with a lot, but not
testing of discrete components. Normally I'd just replace them all, but
with the price per transistor, I'd like to test and replace only the
ones that are bad.

The amp was given to me, and if fixable, will rock out in my garage. I
need something that can be heard over the Bridgeport.

I'm trying harder. <G> (see below)

--
Steve Walker
Fusi...@frontierbrain.com (remove brain when replying)

Sheldon: I'm not saying that all senior citizens that can't master
technology should be publicly flogged. But, if we made an example of one
or two it might give the others incentive to try harder.

[Spehro Pefhany]

Better than Sheldon.

The C2837 is short for 2SC2837 (NPN) and the other is the
complementary PNP 2SA1186

Here are the original Sanken datasheets:-
http://www.sanken-ele.co.jp/en/prod/semicon/pdf/2sc2837e.pdf
http://www.sanken-ele.co.jp/en/prod/semicon/pdf/2sa1186e.pdf

When they're removed from the circuit, you should be able to use the
diode test function most multimeters have and check E-B and B-C both
ways. It should be open one way and read 500 or 600 or 700 the other
way. If it reads a dead short or open both ways then it's probably
blown. If both look like diodes, it's more than likely okay.

Typically the collector (middle pin of both types) is common with the
tab (which may be insulated from the heat sink).

Most multimeters these days have a transistor gain test function, but
it's not all that handy or useful in this case. If you can find a
schematic.. there might be little stuff gone at the same time.


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

[Artemus]

"Steve Walker" <fusi...@frontier.com> wrote in message

news:1LednYdiFO4XCjHT...@giganews.com...

>
> Oops. One side doesn't work. If you crank it up that side will produce
> sound, but not loud, and very staticy. This is what led me to believe
> the output transistors were bad
>
> I do have the two heat sinks out of the amp, with transistors still
> mounted. Anyone got any information on how to check these out, with a
> multimeter? Mine, Protek D-910F, has transistor (NPN and PNP) sockets on
> it. But I don't have a manual, nor do I understand all the information
> on the transistor datasheets.
> --
> Steve Walker

From that description is sounds like the bad channel has blown one of
the output transistors. Since you have a scratchy sound it is likely that
it has blown open and not shorted. Can you put everything back together
and measure the DC voltage on the bad output? It should measure close
to zero when good. A positive voltage indicates that the transistor to
the negative supply is bad(open). A negative voltage indicates the
transistor to the positive supply is bad. A blown output transistor will often
take out the driver transistor too, or at least overstress it and shorten it's
remaining life.

Got a model # for the Denon so we can see about getting a schematic to
know for sure which transistors to go for?
Art

[Steve Walker]

On 10/29/2011 22:05, Spehro Pefhany wrote:
<SNIP>

>
> When they're removed from the circuit, you should be able to use the
> diode test function most multimeters have and check E-B and B-C both
> ways. It should be open one way and read 500 or 600 or 700 the other
> way. If it reads a dead short or open both ways then it's probably
> blown. If both look like diodes, it's more than likely okay.
>
> Typically the collector (middle pin of both types) is common with the
> tab (which may be insulated from the heat sink).
>
> Most multimeters these days have a transistor gain test function, but
> it's not all that handy or useful in this case. If you can find a
> schematic.. there might be little stuff gone at the same time.
>
>
> Best regards,
> Spehro Pefhany

Thanks. But, drat.

More research needed.

Denon PMA-770 from the early 80's


diode test (20K) removed from circuit


C2837 +E-B +B-E +B-C +C-B +E-C +C-E

#1 open 6.9k 6.6k open open open
#2 Open 7.8k 7.7k open open open
#3 open 6.9k 6.6k open open open
#4 open 6.9k 6.7k open open open


A1186
#1 7.1k open open 6.8k open open
#2 7.4k open open 7.3k open open
#3 7.6k open open 7.5k open open
#4 7.2k open open 6.8k open open


Looks like they are all okay.

[sta...@prolynx.com]

> Fusion...@frontierbrain.com (remove brain when replying)

It's been a looong time since I did consumer electronics repair for
tuition, but I used to use the hell out of a Radio Shack transistor
tester. Had several transistor sockets for the small-signal types and
three leads with clips for the big stuff. The way it worked was that
there was the makings of a simple audio oscillator on-board, minus the
transistor. The oscillator fed a garden-variety audio output
transformer which was connected to a neon light on the front panel.
Good ones lit the bulb, bad ones didn't. You could get a rough idea
of how good the amplifying parameters of the particular part were by
turning a pot. The farther around the pot went before the light went
out, the better the transistor was. Still have it somewhere, worked
off a C-cell, usually a Battery-of-the-Month in my case. All the
parts inside and out used to hang on the pegs at the stores, Radio
Shack ate their own dogfood on this one. Don't know if it's still
available. Was pretty cheap at the time. Strictly an out-of-circuit
tester.

Static testing on transistors is usually iffy. You can get leaky ones
that may look good with an ohmmeter, but don't work. You can also
blow the really sensitive types using an old-style analog meter.

Some circuits require matched transistors, usually not consumer
equipment, matching is expensive. Only gizmo I know of that would do
curve matching was a very expensive Tektronix dual curve tracer, multi
K-bucks. Derived from one of their high-end scopes. Usually circuit
design parameters are such that transistors can vary 20-50% on
internal parameters and the circuit would still work. That's the way
the "universal" replacement part biz got started.

As far as how-to, the other poster got it right. Power supply first,
preferably with an O-scope to see if the filter caps are open, leaky
or shorted, followed by voltage checks around the circuit to see if
everything is getting supplied correctly. Then you hook up a signal
source and start tracing. They used to sell an audio signal tracer
for the guys in the radio repair biz that didn't have a scope, just
was a small isolated amp with a speaker and a probe. A circuit
diagram is a really good thing to have for this stuff, I've done
without, but a beginner is going to be lost. It takes a whole lot
longer, too.

Stan