Ralph Mowery wrote:

--------------------

I have seen some pictures of the internals of the 2n3055 transistors
from China and some had a very small silicon area compaired to the
'good' ones. They would test good but probably only good for about 1/2
to 1 amp.


** 2N305s have been around since the mid 1960s and undergone number of chip redesigns in that time.

The JADEC specification does not determine the chip itself - only its overall performance. Long as the watts, volts amps, gain and Ft etc are OK you can label it as a 2N3055.

After discovering fake Motorola power BJTs in 1982, I began cutting the tops of dead and suspect TO3 examples to see what was inside.

Verrry interesting ....

In the case of 3055s, older examples use a large chip with solid metal bonding attachments. Later ones, particular from Motorola use a much smaller chip with aluminium wires. Nothing wrong with either.

Some of the fakes I came across were 2N3055s relabelled as the vasty better MJ15003 types. Others turned out to be high voltage switching devices with tiny chips and even power Darlingtons - visible by the presence of a smaller transistor on the same chip.

One particularly egregious fake from India had two chips inside, wired in parallel in a desperate attempt to meet the Motorola spec. But that idea does not work.

Another fact that I uncovered is that Motorola like to change the chips inside their power BJT devices without warning. Devices sold since the 1990s as the classic 2N3773 or MJ802 types are in fact MJ15003s relabelled.

Fine for new designs, which is all Motorola ever consider, but can cause all sorts of strange behaviours in older ones. Repairers beware.

Plus the price is double or triple that of the more recent part.

Anecdote:

A famous audio engineer from the UK called Doug Self got trapped by this.

He published a design in Wireless World magazine for a "Blameless Amplifier" using MJ802 and MJ4502s from a stock of old ones he had. It had vanishingly low THD.

When folk built his design using the parts and PCB as specified they could not achieve nearly the same result.

Lots of weeping and gnashing of teeth went on in the letters pages.

Nobody ( but me it seems) figured out the cause was Motorola being deceptive.


...... Phil