Ltp Dual Case

[sakis]

in the past most japanese manufacturers designed amplifiers with ltp stage in a dual case and probably with a matched pair .

QUESTION:

is this done for thermal tracking reasons???
is there any other reasons????
if so what is the beneffit of thermal tracking in this particular LTP application ????

after all that
why today's designers no longer design with this kind of transistors????

if for a minute we presume that these transistors( dual cased ) are either obsolete or too expensive why not desingn with bc 559 for example but in such a layout design that both transistors can have mechanical joint with some glue for example.

thanks sakis

[Tube_Dude]

Originally posted by sakis


QUESTION:

is this done for thermal tracking reasons???

Yes!

if so what is the beneffit of thermal tracking in this particular LTP application ????

Less offset voltage, over the variable temperature range

why not desingn with bc 559 for example but in such a layout design that both transistors can have mechanical joint with some glue for example.

Yes, you can do that...

[lineup]

Originally posted by sakis
in the past most japanese manufacturers designed amplifiers
with ltp stage in a dual case and probably with a matched pair .
-------
if for a minute we presume that these transistors( dual cased ) are either obsolete or too expensive

why not desingn with bc 559 for example but in such a layout design
that both transistors can have mechanical joint with some glue for example.

Originally posted by Tube_Dude

Yes, you can do that...

BC550C, BC560C (45 V), works no problems in 2 x 22 VDC supply, for output stage.
BC549C, BC559C (30 V), works no problems in 2 x 15 VDC supply

We are many that uses these lownoise hifi transistors for LTP pairs.

It is recommended to use C versions,
as higher gain is a plus and also reduces base input current (bias).

Less bias current makes lower input noise from resistors
and gives less DC-offset problems at output.

As these transistors are very very cheap, you can buy many.
Then you test until you find good matching pairs.
Usually, if your buy is from same batch, same bigpack (made by a specific machine on a specific day),
should be easy to find matching pairs.

This is why is much much better to buy 100 once,
than buy 10 here and 10 then and next 20 of those BC560C / BC550C.

You can measure and match V b-e,
at the specific Current level
(usually range: 0.4-2.5 milliAmpere is ideal for these transistors)
and Collector-Emitter voltage of your circuit.
At least this is what I do.

Then, if you put each pair Body-to-Body contact,
you have made close to a 'perfect' dual transistor.

See my attachment, what I use instead of glue (some glue will attack plastic!).

If you wish, you can add silicon thermal grease compound
http://en.wikipedia.org/wiki/Thermal_grease
but I do not bother with it. Just good tight mechanical pressure works good enough for me.


Regards, lineup

[lineup]

Hi!

This is another of my original audio ideas.
When I use this way to attach and mount one LTP, long tailed pair,
I will get the pcb layout to correspond to most amp schematics.

You put one BC550C with legs to LEFT and the other with legs to RIGHT.
C, B and E will be logically situated.
A matched pair in body contact, for good thermal tracking, is supposed.


Regards Lineup

[DRC]

I think many dual transitors use a common die and so their characteristics are very closely matched in addition to the thermal tracking advantage.

[roender]

Originally posted by lineup


Hi!

This is another of my original audio ideas.
When I use this way to attach and mount one LTP, long tailed pair,
I will get the pcb layout to correspond to most amp schematics.

Regards Lineup

Not so good idea ... you will get an asymmetrically temperature gradient

[lineup]

Originally posted by roender

Not so good idea ... you will get an asymmetrically temperature gradient

yes, true, a bit

But if you test it in real life, you will see it is no big problem.
There are many theoretical assumptions, ungrounded fears
.. that are of no deal when it comes to practical reality.
( just read some of atom level, cryonic resistors and 0.01 nV noise
academical discussions in blowtorch, for example.
Between these "who's best boy in town & knows it all" -posts,
the John Curl thread gives some good useful knowledge )

One other feature I did not mention:
you get both collectors at V+ side and both emittors at V- side.
So most of the time you can easily make short tracks for supply.


regards, lineup

[roender]

Originally posted by lineup


One other feature I did not mention:
you get both collectors at V+ side and both emittors at V- side.
So most of the time you can easily make short tracks for supply.


regards, lineup

Yes and no ... one bjt will have longer pin conections

[unclejed613]

there are companies that make thermal clips that hold TO-92 transistors together for diff pairs. i think yamaha is one of the amp manufacturers that used the 5-pin dualies in their amps.

nat-semi also makes this device:

[lineup]

Originally posted by roender

Yes and no ... one bjt will have longer pin conections

This is no problem for me.
As I use hardwiring, when doing 'critical' circuits.

for normal audio, roender, I can asure you,
you won't be able to measure or hear anything
by using the construction in my last picture.


Regards having no fears & anxiety, when doing audio, Practically