part 1 1167 bytes content-type:TEXT/PLAIN; charset=US-ASCII
OK, I did some desoldering and hacking tonight to test one possible
configuration and see how it would work. I removed the old regulator
circuit and replaced it with my own. See attached schematic - an LM317T
driving the bases of the two 2N5885 power transistors. I omitted a couple
of 2200uF caps on the input side of the regulator, two more on the
collectors of Q1 & Q2.
When I unplug J3, disconnecting the regulator from the transistors, I get
a nice solid stable 28.5V from the LM317, and I can adjust it up or down.
As soon as I connect the regulator output to the transistor bases, I
measure 44.0V on the output of the LM317 and at the transistor bases. No
adjustment regardless of the pot setting. The collectors of Q1 & Q2 are
at 44.7V. This happens with or without a load on the emitters of Q1 & Q2
- I used a 100 Ohm resistor for a test load. No appreciable difference
with or without it. The LM317 stays cool the whole time.
I don't measure any obvious collector-base shorts on Q1 or Q2, but it sure
is acting like it. Or am I missing something?
Dale
---
We are Dyslexia of Borg.
Fusistance is retile.
Your ass will be laminated.
part 2 6085 bytes content-type:IMAGE/png; name="TS930pwr.png" (decode)
What does it do with a load (like even a 4.7K R)
attached?
(I think Q1 or Q2 may be 'leaky' though - impurities
have a way of 'migrating' within devices with the
passage of time and elevated temperatures don't
help either ...)
The only other thing that might possibly be a
problem - you might require a bit of a pull-down
resistor on the bases of Q1 and Q2 to gnd (but
with what you desribed - it looks like Q1 or
Q2 is bad).
I suspect a bad pass T. With an ohmmeter you should get somewhere around 13
ohms (give or take 10 ohms) between base and emitter, and base and collector,
infinite with reversed leads. Infinite both ways from emitter to collector.
Rick
> When I unplug J3, disconnecting the regulator from the transistors, I get
> a nice solid stable 28.5V from the LM317, and I can adjust it up or down.
> As soon as I connect the regulator output to the transistor bases, I
> measure 44.0V on the output of the LM317 and at the transistor bases.
> I don't measure any obvious collector-base shorts on Q1 or Q2, but it sure
> is acting like it. Or am I missing something?
>
> Dale
> ---
> We are Dyslexia of Borg.
> Fusistance is retile.
> Your ass will be laminated.
>
> ------------------------------------------------------------------------
> Name: TS930pwr.png
> TS930pwr.png Type: PNG image file (IMAGE/png)
> Encoding: BASE64
Jim, Rick, thanks to you both. I was hoping I wasn't going nuts. Q1
measures infinite resistance from E to C but not C to E. I'm ordering a
new pair of transistors tomorrow. Lot markings are 8239 - Darned things,
they're only 20 years old, what are they doing failing this soon?? 8-)
Dale
---
We are Dyslexia of Borg.
Fusistance is retile.
Your ass will be laminated.
On Mon, 4 Nov 2002, Rick C. wrote:
> I suspect a bad pass T. With an ohmmeter you should get somewhere around 13
> ohms (give or take 10 ohms) between base and emitter, and base and collector,
> infinite with reversed leads. Infinite both ways from emitter to collector.
> Rick
> When I unplug J3, disconnecting the regulator from the transistors, I
get
> a nice solid stable 28.5V from the LM317, and I can adjust it up or
down.
> As soon as I connect the regulator output to the transistor bases, I
> measure 44.0V on the output of the LM317 and at the transistor bases.
No
> adjustment regardless of the pot setting. The collectors of Q1 & Q2 are
> at 44.7V. This happens with or without a load on the emitters of Q1 &
Q2
> - I used a 100 Ohm resistor for a test load. No appreciable difference
> with or without it. The LM317 stays cool the whole time.
Are you really really sure the pass transistors are wired correctly? It
sure sounds like the 40V is being conducted right thru one of the
junctions to the LM317.
*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
> > When I unplug J3, disconnecting the regulator from the transistors, I
> get
> > a nice solid stable 28.5V from the LM317, and I can adjust it up or
> down.
> > As soon as I connect the regulator output to the transistor bases, I
> > measure 44.0V on the output of the LM317 and at the transistor bases.
> No
> > adjustment regardless of the pot setting. The collectors of Q1 & Q2 are
> > at 44.7V. This happens with or without a load on the emitters of Q1 &
> Q2
> > - I used a 100 Ohm resistor for a test load. No appreciable difference
> > with or without it. The LM317 stays cool the whole time.
>
{Original Message removed}
Olin sez:
> Are you really really sure the pass transistors are wired correctly? It
> sure sounds like the 40V is being conducted right thru one of the
> junctions to the LM317.
Yep, really really sure. Unless Kenwood built it wrong and nobody figured
it out for 15-20 years. And unless I was totally wrong when double
checked it just in case. But yes, it does seem the colelctor voltage is
being seen on the base, which probably induces the LM317 to just shut down
immediately.
Michael sez:
> Just wondering, they definately aren't PNP transistors are they? :o)
Not unless 2N5885s have had a major polarity shift I didn't hear about. 8)
"they're only 20 years old, what are they doing
failing this soon?? "
It all depends on a number of factors including the
original manufacturing 'proceses' of the fab (fabrication
facility) and methods used to make the devices (types
of band wires, methods used to seal the die in the
package, how clean their 'processes'/processing
facilities were ...).
Semiconductor manufacturing is a lot like politics
or sausage making - you may appreciate the outcome
BUT the process itself is messy and you REALLY don't
want to know all the intimate details and pitfalls
of what it takes to 'make product'.
And this is aside from any abuse they may seen through
the years!
RF Jim (who spent sev. years at TI's "GaAs Facility")
Once you get the new transistors are you going to re-try the origonal
circuit?
Richard P
Olin sez:
> Are you really really sure the pass transistors are wired correctly? It
> sure sounds like the 40V is being conducted right thru one of the
> junctions to the LM317.
Yep, really really sure. Unless Kenwood built it wrong and nobody figured
it out for 15-20 years. And unless I was totally wrong when double
checked it just in case. But yes, it does seem the colelctor voltage is
being seen on the base, which probably induces the LM317 to just shut down
immediately.
Michael sez:
> Just wondering, they definately aren't PNP transistors are they? :o)
Not unless 2N5885s have had a major polarity shift I didn't hear about. 8)
> "they're only 20 years old, what are they doing
> failing this soon?? "
I was being facetious - I'm not surprsed when 20 year old power
transistors fail, especially considerign these had a thermostatic fan
control that probably allowed some pretty big thermal swings. A new pair
is on the way, that should fix it.
> Once you get the new transistors are you going to re-try the origonal
> circuit?
Nope. I don't think it has any advantage over mine, and mine uses a whole
lot fewer parts. The power supply in these particular radios is widely
known to be of marginal design. I may add some additional protection but
the old parts aren't going back in. I'm also replacing the thermostatic
fan control (and probably the fan as well) and running the cooling fan
nonstop at a lower speed to reduce noise and keep things more consistently
cool.
Dale, I've got a solid-state 4-band Sears AM/FM/SW1/SW2
portable (it has a handle on top and uses 4- D cells!)
receiver I bought back in 1965 - I have only replaced
the FM RF Amp transistor (back sometime in the late
sixties - and this was after I probably fried it during
a lightning storm when I had it connected to an external
antenna) -
- and the AC power supply in that radio uses a series
regulator design and a push-pull audio stage that drives
the built-in 5" x 7" speaker - nothing but that one RF
transistor has failed in that portable in 37 years!
I've also got a 1965 Motorola Motrac (commercial 2-way
radio) that is all original as far as transistors - this
includes some beefy TO-3 cased transistors that perform
+12V conversion to +450 VDC for the 85% tubed
transmitter ... I have had those beefy transistors blow
on *other* Motrac radios (a 100W six meter 'trac conversion
- it tripped the 40 Amp breaker in-line to the battery
when it shorted).
So, I guess I took the "20 years" too literally, but, as
I point out that would only seeem to be part of the
answer - it gets more complicated with time - and
with "temperature cycling" as you also point out ...
To that end - both I and a friend had the switching PSUs
in our 83 channel (yes, 83 standard broadcast channels!)
Emerson 20" colors TV die in the mid-late 80's (we had
bought them about the same time mid 80's).
After installing my replacement PSU I proceeded to do some
temp tests - and discovered a 60 degree C rise above ambient
on the heatsink!
A fan installed inside the set pointed at that PSU
(with attached stock heartsink) has apparently 'solved'
the temperture induced failure mechanism and extended
it's life greatly (now into late 2002).
I agree, it's certainly not unusual for transistors to go a lot longer
than 20 years. I was just saying I'm not terribly *surprised* when they
fail any time past 15 or so years, especially on gear that I already know
has some well known design shortcomings.
I've had consumer gear quit after a couple of years; I'll be pleasantly
astonished if my current PC power supply and motherboard last another year
(though it will probably be a dried out cap that kills it). On the other
hand, I've seen audio gear and other stuff last well past what you'd
normally expect to be its useful service life. But when a TS-930 power
supply croaks... well, let's just say no one falls over from surprise!
Dale
---
We are Dyslexia of Borg.
Fusistance is retile.
Your ass will be laminated.
On Tue, 5 Nov 2002, Jim wrote:
> So, I guess I took the "20 years" too literally, but, as
> I point out that would only seeem to be part of the
> answer - it gets more complicated with time - and
> with "temperature cycling" as you also point out ...
Audio gear - Heath AR-1500A that still runs hours a
day (although it needs another change-out of pilot
lamps AGAIN! A set, it turnms out, lasts about 2 1/2
years in this service cycle with the last bulb left
standing then out at about the 3 year mark) ...
(AR-1500A a 70's era AM/FM integrated tuner/amp
that I picked up 'at the factory' ... the AR-1500A
was derived from the popular component set before
that - the Heath AR-15 series, output power: 60W
RMS/channel continuous.)
This same amp blew a PS bridge rectifier and a transistor
or two during a "loud passage" (I had it cranked open)
a number of years ago - owing to a pinhole punched through
a Mica TO-2 insulator caused by a small burr on the power
transistor heatsink ... it took that LAST loud sound
passage, though, to finally cause things to terminally
'breakdown' ...
Are the 2N5885s pnp or npn ? They sound like pnp from what you see. In
that case you need to connect the regulator input to them and a resistor
between B and Es.
Dale
---
We are Dyslexia of Borg.
Fusistance is retile.
Your ass will be laminated.
On Tue, 5 Nov 2002, Peter L. Peres wrote:
> Are the 2N5885s pnp or npn ? They sound like pnp from what you see. In
> that case you need to connect the regulator input to them and a resistor
> between B and Es.
Jim wrote:
>
> Audio gear - Heath AR-1500A that still runs hours a
> day (although it needs another change-out of pilot
> lamps AGAIN! A set, it turnms out, lasts about 2 1/2
> years in this service cycle with the last bulb left
> standing then out at about the 3 year mark) ...
Hi Jim, we used to fix a lot of audio bulbs,
try replacing them with 3v higher rated bulbs,
ie if driven by 9v (common) use 12v bulbs.
The axial bulbs are more reliable than the radial
type and if they are driven from AC try a series
diode and a cap to make it DC, and rate the bulbs
3v higher again than whatever DC voltage you got.
Most TV repair shops will have a wide range of
dial bulbs in varying voltages and sizes. :o)
-Roman
Hi Roman, I think I may take your advice ... also,
I may look at some of those socketed LED re-
placements (these are currently socketed
bayonet-based bulbs - type #44 I think). Most of
these bulbs first 'silverize' their interiors
(reducing their light output substantially)
then they fail a year later or so!
This will be the fourth 'set' (fully 'bulbing' this
dude takes 5 bulbs) that I have put in since the
late 1970's since I started keeping records on when
bulbs were changed out!
Jim wrote:
> Hi Roman, I think I may take your advice ... also,
> I may look at some of those socketed LED re-
> placements (these are currently socketed
> bayonet-based bulbs - type #44 I think). Most of
> these bulbs first 'silverize' their interiors
> (reducing their light output substantially)
> then they fail a year later or so!
>
> This will be the fourth 'set' (fully 'bulbing' this
> dude takes 5 bulbs) that I have put in since the
> late 1970's since I started keeping records on when
> bulbs were changed out!
>
> RF Jim
This is 100% correct, a simple LED can survive feeding light for 100 years.
No wonder many cars, traffic lights, bus, trucks, even flashlights are now
using LEDs.
I already contacted some companies who are producing all kinds of old lamps
with leds, old sockets and all. Googling can locate some of those
companies.
Thomas Edison would be proud to see how many years his invention survived
until finally replaced by something with a better life span.
mitvma.mit.edu with SET PICList DIGEST in the body